HEP Assessment Master Report Final

National
Assessment of
The Ethiopian
Health Extension Program
Final Master Report
May 2020
National
Assessment of
The Ethiopian
Final Master Report
May 2020
MERQ Consultancy PLC has received funding for the National Assessment of the
Ethiopian Health Extension program from the Bill & Melinda Gates Foundation.
CONTRIBUTORS
Consulting Firm MERQ Consultancy PLC
Tadesse Chekol Commercial Center, 8th Floor, Addis
Ababa, Ethiopia
Principal Dr. Alula Meressa Teklu (MD, MPH)

Investigator
Technical Lead Yibeltal Kiflie Alemayehu (BSc, MSc, PhD Candidate)
Technical Team Contributors to the overall assessment
Dr. Girmay Medhin, Fasil Walelign Fentaye, Tegene

Legese Dadi, Yohannes Ejigu Tsehay, Dr. Esayas
Haregot Hilawe, Wasihun Andualem Gobezie, Dr.
Girma Azene Chere, Teklemichael Gebru Tesfay, Melaku
Gebremedhin, Merhawi Gebremedihin Tekle, Daniel
Tadesse Assefa, Hajira Mohammed Amin, Dr. Abebaw
Minaye Gezie, Dr. Kirsten Senturia, Mekdes Demissie
Challa, Alemu Tesfahun Fida, Dr. Kesetebirhan Delele,
Dr. Trhas Tadesse, Werissaw Haileselassie, Tefera
Mulugeta, Bisrat Dinberu, Elias Geremew, Dr. Mulusew
Gerbaba Jebena, Dr. Million Tesfaye Eshete, Dr. Araya
Abraha Medhaniyie.
Contributors to supplementary studies
Urban HEP Assessment

Biniyam Tadesse Haile, Israel Mitiku Hatau
Assessment of the quality of Health Extension

Workers’ Training institutions
Dr. Seleshi Zeleke Teketel, Tadele Zebrea Shikur
Attrition among Health Extension Workers

Merhawi Gebremedihin Tekle, Habtamu Milkias
Wolde, Wondimye Ashenafi
page-II
National Assessment of
The Ethiopian Health Extension Program
The role of the Health Extension Program in
Public Health Emergency Management
Abduilhafiz Assen Endris, Dr. Musse Tadesse
Chekol, Tariku Takele, Dr. Negusse Yohannes Sebro,
Sileshi Demelash
Cost-Effectiveness Analysis of the Health

Extension Program
Lelisa Fekadu Assebe, Wondesen Nigatu Belete,
Amanuel Lulu Yigezu, Dr. Senait Alemayehu
Beshah, Dr. Elias Asfaw Zegeye
Analyses of HEP financing landscape

Ermias Dessie Buli, Ageazit Teka Gebreslassie,
Lulesged Nigussie Assefa
Qualitative data coding and synthesis

Helen Ali, Aneni Tesfa, Getinet Tesfaw, Emenet
Mesefin, Shumye Molla, Tesfay Birhan
Database Management and Quality Assurance

Nadew Tademe, Setegn Tigabu, Binyam Tedla,
Alehegn Engdaw, Redeat Workneh, Lemlem
Yehyes, Tsehay Tekle, Misgana Haileselassie
Reviewers Dr. Nejmudin Kedir Billal

Dr. Aida Abashawel
Prof. David Hotchkiss
Dr. William Weiss
Editors ProofreadingServices.com
Dr. Kindalem Damtie
Programs and Frehiwot Bekele, Daniel Tadesse, Mame Gurumu

Operations
Document Dimitri and Brehanie

Design
page- III
MERQ (Monitoring, Evaluation, Research and Quality Improvement)
Consultancy PLC: is a consulting firm established by a team of qualified
professionals with extensive expertise and experience in areas of public health,
monitoring and evaluation, research, organizational capacity assessment,
situational analysis, survey, health informatics, database management, data
analysis, development of guidelines, SOPs and manuals, and capacity building.
We envision to be a pragmatic choice for research and scientific inquiries in

Ethiopia and beyond. Our focus on rigor, responsiveness and utilization has
brought us a long way and has allowed us to establish strong collaborations
with multiple local and international academic and research institutions. We
have a sister company under the same name in the United States, which
solidifies our international collaboration solid.
Please contact us at: info@merqconsultancy.com or visit us at www.

merqconsutlancy.org or at our office: 8th floor, Tadesse Chekol Building, P.O.
Box 54023 Tel.+25111854754, Arada sub city Addis Ababa, Ethiopia for your
research, scientific inquiries and training needs.
RECOMMENDED CITATION
Alula M. Teklu1, Yibeltal K. Alemayehu1,2,3, Girmay Medhin1,4, et al (2020).
National Assessment of the Ethiopian Health Extension Program: Final Master
Report. Addis Ababa, Ethiopia: MERQ Consultancy PLC.
1 MERQ Consultancy PLC, Ethiopia

2 Jimma University, Ethiopia
3 Tulane University, USA
4. Addis Ababa University, Ethiopia
page-IV
ACKNOWLEDGEMENT
The study was conducted as a result of the Ministry of Health of
Ethiopia’s interest to have a comprehensive assessment of its flagship
initiative – the Health Extension Program, by an independent research
team for the purpose of evidence-based decision making going forward.
The MERQ team expresses its gratitude to the Ministry of Health for
giving us the chance to carryout this historic assessment. We would
like tospecifically express our heartfelt appreciation to Dr. Amir Aman
and Dr. Lia Tadesse for their leadership, and genuine quest for evidence
and for allowing us to do this assessment with complete independence.
The Bill and Melinda Gates Foundation’s willingness to respond to the

quest for evidence is a good example of responsive and pragmatic
program support. The MERQ team would like to express its acclaim to the
foundation for the exemplary decision it took to support the assessment
and for the trust it bestowed on us.
The MERQ team would also like to express its deepest appreciation to the
more than 15,000 women, men and adolescent respondents from the more
than 8,000 households for providing valuable information by dedicating
their precious time. We are thankful to the women development army
members, health extension workers, health extension professionals,Woreda
Health Office staff members, Woreda Health office managers of our study
woredas, nurses, midwives and managers working in thehealth centers that
we assessed as well as community leaders.
This study would not have been possible without the ultimate dedication
and unwavering focus to the quality of the data that we collected, given
by our data collectors. MERQ is grateful to all the 450+ data collectors,
supervisors, and coordinators who overcame the challenges to meet the
deadline while maintaining high standards.
Our appreciation also extends to the regional health bureaus of Tigray,

Amhara, Oromia, Southern Nations Nationalities and Peoples, Gambella,
Benishangul Gumuz, Harari, Afar, and Somali regions and to the city
administration health bureaus of Addis Ababa and Dire Dawa, for
facilitating access, providing data and dedicating time for the multiple
interviews we held with them.
page- V
The design, analysis, and interpretation of the assessment has benefited
from input solicited from most Ethiopian Universities, various development
partners and primary health care experts. We thank all the experts who
partook in the 26 consultative meetings we have held with them.
Given the challenges faced during the actual data collection which called
for flexibility and revisions to the original grant, the pragmatic and timely
response we received from the BMGF team - Yabsera Marcos, Melon
Feleke and Dr. Solomon Zewdu has made this study a reality. We are
grateful to this team.
Last, but not least, our heartfelt appreciation goes to the directorate of
the health extension program and primary health care at the ministry,
especially Temesgen Ayehu, Israel Ataro and their team members
for ensuring our independence in the conduct of the study and the
steering committee which comprised of members from the ministry, the
international institute of primary health care, the BMGF and EPHI for
helping make critical decisions in the prioritization of the activities
which ensured rigor and comprehensiveness.
Alula M. Teklu, MD, MPH

Principal Investigator
page-VI
EXECUTIVE
SUMMARY
The major health problems in Ethiopia remain preventable communicable
diseases; maternal, neonatal and child health conditions; and nutritional
disorders, while non-communicable diseases (NCDs) are also on the rise,
predicting a double burden of disease. Poor socio-economic conditions,
food insecurity and inadequate nutrition, a low level of awareness about
health, and inadequate health service delivery contribute to the poor
health status of the population. The Health Extension Program (HEP),
launched in 2003, is one of the Ethiopian government’s major health
investments designed to improve the health of communities.
The HEP is a community-based health program that was initially designed

to deliver 16 packages of health services to agrarian communities. Later,
the program evolved in various ways, including being adapted for
pastoralist and urban communities, making changes to existing packages,
and including 2 additional packages and upgrading training of Health
Extension Workers (HEWs). In 2019, the program involved more than
39,878 HEWs and 17,587 community health posts (HPs).
The program has been acknowledged as a major contributor to recent

gains in health outcomes in the country. In recent years, however, signs
of deterioration in the performance of the program have increased
concerns among stakeholders. In response, the Ministry of Health
(MoH), in collaboration with the Bill and Melinda Gates Foundation,
commissioned MERQ Consultancy PLC to conduct a comprehensive
national assessment of the HEP. The objective of the assessment was to
assess the status, determinants, and prospects of the HEP, and to identify
challenges and areas of intervention for program and policy decisions in
the Ethiopian health sector.
Methods
A national assessment of the agrarian, pastoralist, and urban HEP was

conducted from October 2018 to September 2019. All 9 regional states
and both city administrations were included. The assessment was guided
by the Primary Health Care Performance Initiative (PHCPI) framework,
page- VII
which has 5 domains: system, inputs, service delivery, outputs, and outcome.
Data were collected at all levels of the health system, including the MoH,
Regional Health Bureaus (RHBs), Woreda Health Offices (WorHOs),
Health Centers (HCs), HPs, and communities. Comprehensiveness and a
focus on level of use, independence, and responsiveness were the principles
that guided the assessment. This ensured that the processes and out-
puts of the assessment could accurately and adequately inform the
health sector in making HEP-related decisions.
The assessment had 3 components: collection and analysis of primary

data using mixed methods, a systematic review of peer-reviewed and
gray literature on the HEP, and the synthesis and formulation of
recommendations. Primary data were collected through a household
survey, a survey of HEWs, a health facility assessment, key informant
interviews, and focus group discussions (FGDs). Data were collected
from 6 430 households from the general population, 618 Women’s
Development Army (WDA) leaders’ households, 343 HPs, 179 HCs, and
62 WorHOs from 62 woredas randomly selected from all regions. The
household survey involved women, men, and youth girls; a total of 11 746
respondents were interviewed from the sample households in the general
population and 1 122 from WDA households, respectively. Qualitative
data were collected through 172 interviews and 109 FGDs.
The urban HEP was also assessed by collecting data from 1 912 households
(1 287 from Addis Ababa and 625 from Dire Dawa). More than 404
urban HEWs from Addis Ababa and 113 from other cities were included
in the study. A total of 132 qualitative interviews were conducted with
community members, policy advisors, partners, and urban HEWs. All
HCs in Dire Dawa and Addis Ababa were included in the study.
In addition to the primary assessment, in-depth studies were also

conducted on specific topics, including attrition among HEWs, the
quality of pre-service HEWs’ training institutions, the role of the HEP in
public health emergency management, and the cost-effectiveness of HEP
interventions. Stand-alone reports are included in Part 4 of the report.
page-VIII
Results
RELEVANCE OF HEP PACKAGES AND SERVICE DELIVERY
MODALITIES
Communicable diseases and maternal, neonatal, and nutritional

disorders still constitute 60% of the total disability-adjusted life years
lost in Ethiopia, and NCDs and injuries have been rising in recent years.
No HEP package has yet been fully implemented to a level calling for
the retirement of a package. The initial 16 HEP packages, as well as the
newly added packages on NCDs and mental health, remain relevant
in addressing the disease burden of the largely rural communities of
Ethiopia. The packages, however, lack the comprehensiveness needed
to meet the rapidly changing disease epidemiology and ever-growing
community expectations. Among the respondents who had ever visited an
HP, 43.9% of women, 51.5% of men, and 49.3% of youth girls recommended
either expanding the current packages or adding new packages of
services. More comprehensive maternal and child health services and the
treatment of sick adults were the most frequently suggested changes to
the HEP packages.
The current service delivery arrangement of the HEP through home-

, HP-, and community-based services was reported to be appropriate
by a large proportion of community members. Home visits by HEWs
are believed appropriate by 82.2% of women, 81.7% of men, and 76.6%
of youth girls. Mixing male and female HEWs in both agrarian and
pastoralist settings, however, has been suggested to address problems
arising from the assignment of either male or female HEWs. Communities
perceive that HEWs are generally trustworthy and friendly, considered
models of good behavior, and seen as helpful in empowering women and
communities to solve their own health problems. Trust is limited, however,
when it comes to more clinical or curative services.
INPUTS OF THE HEALTH EXTENSION PROGRAM
Human resources
Compared to the volume of work and skill sets required for the effective
delivery of HEP packages, the staffing standard for HPs—2 HEWs per
HP—was found to be inadequate. A high proportion (86%) of HPs meet
the minimum standard of 2 HEWs per HP. Almost all HEWs (98%) have
page- IX
attended level III or level IV education and training programs, and at
least 1 level IV HEW was available in 62.4% of HPs. In 5% of HPs, there
was at least 1 nurse or midwife. Absenteeism, however, was very common;
21% of the staff of the study HPs were absent, and 48% of the study HPs
had at least 1 HEW absent on the date of data collection.
Incompetence is a major gap among the current workforce of the HEP.

Despite the high number of level III or level IV qualifications held by
HEWs, only 57% of HEWs were certified. One quarter of HEWs were
unable to earn certification even though they had taken the competence
test at least once. Only 51.1% of HEWs could correctly describe the current
national guideline on the schedule of child vaccinations. On the other
hand, the self-reported competence was high across different service
areas. Challenges in pre-service education at HEWs’ training institutions,
including the inadequate recruitment of candidates, insufficient practical
sessions, language gaps in the medium of instruction, and irregularities
in the provision of refresher trainings were major sources of competence
gaps among HEWs.
The level of motivation among HEWs was sub-optimal. Nearly one third
(32%) reported having the intention to leave their jobs. The actual rate
of attrition was, however, only 21% from the start of the program to
2019. The rate of attrition was 2.9 per 100 person-years. Overall, 51%
of HEWs were dissatisfied with their jobs. HEWs were relatively more
satisfied in the dimensions of autonomy, work environment, relationships
with co-workers, and recognition for their contributions. On the other
hand, job security, salary and benefits, perceived alternative employment
opportunities, and opportunities for promotion were the areas in which
HEWs were the least satisfied.
Facilities and infrastructure

Almost all kebeles (98%) have at least 1 HP. Overall, the HP-to-population
ratio is 1:5 760. Most HPs’ buildings are substandard, however; only 37%
of HPs meet the required building standard. Most HPs also lack basic
amenities. An electric power supply, improved water source, and latrine
were available in 26.5%, 27.1%, and 87.4% of HPs, respectively. The lack
of basic equipment is a major gap in the majority of HPs as only
5%had all basic tracer equipment (blood pressure apparatus, stetho-
scope, thermometer, adult scale, child scale, and artificial light sources).
page-X
Drugs and medical supplies
Stockout of tracer drugs and medical supplies was widespread. For 6 of
the 14 items assessed, more than 50% of HPs had stockout on the date
of data collection. Oral Rehydration Solution (ORS), Depo-Provera, zinc
acetate, and Vitamin A capsules were the most commonly available items,
being available in more than 80% of HPs. Paracetamol suspension or
suppositories, tetracycline eye ointment, paracetamol tablets, amoxicillin
suspension, and pentavalent vaccine were available in fewer than 50% of
HPs. Implementation of drug supply chain management was sub-optimal.
Only 18% of HPs had a bin card for stock management, and, of these,
70.4% of the HPs’ bin-card data were complete and accurate. A majority,
81%, of HPs were using HPMRR, but only 25% of the HPs’ HPMRR
data were complete and accurate. Out of the HPs we surveyed, only 40%
had sufficient storage space and only 32% had a functional refrigerator
for pharmaceuticals. Only 57% of HPs had implemented first expiry,
first out (FEFO) stock rotation; damaged and expired products were
available in 75% of HPs.
Financing of the HEP

Between 2010/11 and 2016/17, HEP spending increased from 2.4 billion
ETB (USD 0.52 billion in terms of PPP) to 5.1 billion ETB (USD 0.58
billion in terms of PPP). Over the same period, the share of the total
PHCU-level spending represented by HEP spending declined from 25% to
22%. Similarly, the HEP’s share of the Total Health Expenditure declined
from 8.9% in 2010/11 to 7.1% in 2016/17. Despite recent increases in the
share of government spending on the HEP, the program still depends
heavily on external sources of finance; the government’s share of HEP
spending increased from 20.8% in 2010/11 to 40.3% in 2016/17. Child
health represents the largest portion of HEP spending, accounting for
46% of the program’s total expenditure.
HEALTH SERVICE DELIVERY THROUGH THE HEP
Of the 352 kebeles included in the assessment, 97.4% had at least 1 HP.
Communities did have other major barriers to access, however, including
the services’ lack of comprehensiveness and interruption, communities’
limited awareness of the services that were reportedly available, and a
lack of trust in HEWs regarding some services. Community members’
awareness of the availability of HEP services was only 58.8%. During
the 1-year period preceding the study, the women, men, and youth girls
page- XI
reporting at least 1 exposure to the HEP through any modality were
54.8%, 32.1%, and 21.9%, respectively. HPs were the predominant source
of exposure, compared to home visits and other settings, in both agrarian
and pastoralist communities, with even higher exposure in pastoralist
settings. On average, the implementation of the HEP packages at the
household level was 50.8% among households in the general population
and 60.6% among the households of Women’s Development Army
(WDA) leaders. The intensity of HEP implementation at the kebele level
was linked primarily to human resources. An HP’s professional mix and its
staff’s level of education, rather than its number of HEWs, are associated
with better implementation of the HEP through home and HP visits.
Community engagement
Model family training in the past and community mobilization through
WDA and Social Mobilization Committee (SMC) members have been
the HEP’s major strategies for community engagement. Despite official
reports indicating high coverage of model family training, only 14.9% of
women in agrarian settings and 8.0% of women in pastoralist settings
reported being aware of the model family training. Enrollment and
graduation rates were very low, with only 2.9% of agrarian and 2.1% of
pastoralist households reporting having ever been enrolled in the training.
Awareness of model family training was universal among WDA leaders
and SMC members, but enrollment and training completion rates were
very low compared to what would be expected from community leaders.
Community structures supporting the HEP were available almost in all
study kebeles. WDA structures were reported to be available in 97.0%
of agrarian kebeles. Similarly, 92.5% of pastoralist HPs reported the
availability of either a WDA, a 1-to-5 network, or an SMC structure that
supported the HEP in their respective kebeles. The inadequate density
and functionality of these structures, however, are major challenges;
only 25.9% kebeles have a WDA density of more than 30 per 1 000
households. The functionality of the existing structures is also very low.
The limited capacity of WDA leaders and SMC members was the other
major challenge to community engagement with the HEP. Compared
to the general population, WDA leaders are more likely to have better
health behaviors, higher educational status, and higher socio-economic
status. They did not have service use rates as high as were expected,
however, from a group of women meant to serve as role models. A
substantial proportion of WDA leaders had an incomplete course of
antenatal care (ANC) visits, delivered their last child at home, or failed
page-XII
to vaccinate their children fully, suggesting the need to revise how these
volunteers are nominated as leaders.
Information system and M&E

The Community Health Information System (CHIS) captures only a small
portion of the functions of the HEP. The currently reportable indicators of
the HEP focus only on outputs, with little opportunity for HEWs to track
their service processes. Among the 6 areas of the health information
system assessed, no single component achieved a score of 75% or higher.
M&E structure, data management, and the availability of inputs were
relatively better implemented. The data verification factor for the 10
selected indicators showed enormous over-reporting. Nearly one third of
HPs over-reported all 10 tracer indicators. Under-reporting was observed
in about 10% of HPs. Information use at the HP level was minimal in all
regions.
Coverage of HEP-related services

Hygiene and sanitation: The coverage of HEP-related services showed
incremental improvement from previous levels. The coverage gap, however,
remains very wide. The coverage of improved sanitation facilities and an
improved source of drinking water was 20.0% and 71.4%, respectively.
Appropriate waste-disposal practices were observed in only 10.7% and
10.8% of households for solid and liquid wastes, respectively.
Disease prevention and control: Half of the households in malarious

areas own at least 1 insecticide-treated bed net (ITN). The prevalence of
sleeping under an ITN was 27% among household members in malarious
areas and 47% among those who own at least 1 ITN. Only half of children
and one third of pregnant women in malarious areas slept under an ITN
the night before data collection. Large proportions of women, men, and
youth girls were aware of HIV and TB, but comprehensive knowledge
about their transmission and prevention methods was very low. HEWs
served as major sources of information about communicable diseases,
but they provided services to a low percentage of households. Patients
very often bypass HPs, even for services they expect to be offered at
the HP level. The HEP is currently playing a very limited role in the
prevention and control of NCDs and in addressing the mental health
needs of community members.
page- XIII
Family health services: Coverage of family health services has improved
significantly, particularly in recent years, but remains very low compared
to national targets. The contraceptive prevalence rate reached 44.6%,
and attendance of at least 4 ANC visits and health facility delivery was
48.3% and 54.9%, respectively. Coverage of postnatal care remained low,
with only 25.5% women who delivered during the 2 years preceding the
study having had a postnatal checkup. Coverage of full basic vaccinations
among children 12-23 months of age was 35.7%. Treatment-seeking for
children with diarrhea, pneumonia, or fever was inadequate, with fewer
than half of those children being brought to the attention of a health
professional.
HEWs and HPs served as both information sources and service providers
for the majority of family health services. The quality of these services,
however, is compromised for a number of reasons. ANC visits with missing
components, the increased probability of disruptions in the continuum of
care for maternal health services when ANC services are provided by
HPs, and high vaccination dropout rates were among the indications of
the compromised quality of care at HPs.
HEP service delivery outcomes

Trends in the health outcomes in Ethiopia show that there have been
substantial improvements in the health status of Ethiopians in the areas of
communicable disease prevention and control, maternal and child health,
and hygiene and environmental sanitation. Meta-analysis of studies that
investigated the effectiveness of the HEP in improving health outcomes
showed that both the HEP as a package and its specific components are
associated with improved maternal and child health outcomes. Regression
analyses of data from the household survey performed in this study also
showed strong associations between the implementation intensity of the
HEP and the adoption of healthy behaviors at the household level. A
10% increase in the proportion of households reached through home
visits was associated with a 19% increase in the adoption of HEP-related
behaviors among households. Similarly, a 10% increase in the proportion
of pastoralist households who had interactions with HEWs through HP
visits was associated with a 16% increase in the adoption of HEP-related
behaviors at the household level.
page-XIV
Governance, leadership and management of the HEP
The HEP structure that extends from the federal to the community level is
well designed and accepted. Governance and leadership roles, however,
are not functioning well, particularly at the woreda and kebele levels.
The rapid and massive construction of HPs, which now cover 98% of
kebeles, & the deployment of around 40,000 full-time, salaried HEWs
are signs of strong government commitment to the HEP. The HEP is
integrated into primary health care units (PHCUs), but attention to the
HEP by the government has been declining in recent years. Currently,
woreda and kebele administrations pay minimal attention to the HEP;
they lack the technical and administrative capacity to lead and follow
up on HEP activities. HEP coordinators at HCs and in some woredas
are non-technical, and this has profound implications for the depth and
breadth of the technical support that HPs and HEWs are receiving.
There is a limited multi-sectoral approach to the implementation of the

HEP at the kebele and woreda levels. HEWs’ dual accountability and
overlapping responsibilities from HCs and kebele administrations has
been a major challenge to the program. Assigning HEWs to non-health-
related activities (e.g., tax collection) has had a destructive effect on
both HEWs’ provision of health service and the acceptability of the HEP
by community members. The roles, responsibilities, and links between
WDAs and HEWs were designed properly, but their actual functioning is
in many places unclear. The WDA approach is found to be less feasible
in pastoralist settings. Community acceptance of WDAs and HDAs is
very low for a number of reasons, including community perceptions of the
organization, which is seen as having a political agenda.
The 15 packages included in the urban HEP are being delivered using
model family trainings, WDAs, and model trainings at schools and youth
centers, with the HC being the dispatch center. Although more than
three quarters of respondents agree that the packages are relevant, the
packages are not adequate to solve the multidimensional challenges
of the urban community. NCDs are not adequately addressed, and
water, sanitation, and hygiene (WaSH) is poorly implemented because
the challenges in this area are beyond the urban HEP’s capacity.
Overcrowding, poor housing conditions, and a lack of water and sanitary
facilities are major challenges. The home visit is the most challenging
page- XV
service delivery modality, with up to 36% of households having never been
visited, but the WDA seems to work well in urban settings. More than
28% of UHEWs are actively looking for another job, and the attrition
rate is 21%.
Conclusion and recommendations
The HEP has been a major contributor to Ethiopia’s recent gains

in health outcomes. The program’s original 16 packages of health
interventions, as well as the 2 packages added recently, are relevant in
addressing the priority health needs of communities. The packages lack
the comprehensiveness needed, however, to meet the rapidly evolving
needs of communities and the changing epidemiologic situations.
The program is currently exhibiting sub-optimal quality and coverage as

a result of inadequate inputs, particularly in the area of human resources,
as well as implementation challenges related to community-engagement
strategies, the information system, and leadership and governance. Men
and the youth have largely been excluded by the HEP in both service
delivery and community-mobilization strategies.
The comprehensive optimization of the HEP is necessary to improve

the quality and coverage of existing HEP interventions and expand
packages to ensure access to more comprehensive services. Developing a
document that can serve as a long-term guide to ensure the design and
implementation of responsive HEP that envisions the evolving needs of
the populations is vital.
Changes in service packages should be guided by long-term plans

to ensure universal health coverage and consider the availability of
alternative sources of services.
Service delivery modalities should also be expanded to reach all segments

of the population, including women, men, and the youth. The HEP’s
inputsshould be revised to ensure an adequate supply of human and
materialresources required for the effective delivery of more comprehen-
sive servicepackages. Stratifying HPs based on the relative distance of
kebeles fromtheir nearest respective HCs is a priority to encourage fur-
ther investment in the HEP. Introducing a professional mix at the HP
level, accompanied by the creation of an enabling environment, is key
to the provision of more comprehensive services.
page-XVI
The urban HEP requires rethinking in light of its contextual factors,
including access, control over key interventions, and demographic factors.
Contextualization of the packages and implementation strategies is
necessary for pastoralist communities.
Given the vital role expected to be played by community volunteers, it is

necessary to redesign community engagement strategies with a focus on
identifying and engaging community members who can influence their
community members both directly and indirectly.
page- XVII
CONTENTS
Contributors ------------------------------------------------------------------------II
Acknowledgment --------------------------------------------------------------------V
Executive Summary --------------------------------------------------------------------------VII
Contents ---------------------------------------------------------------------------------XVIII
Figures ---------------------------------------------------------------------------------XXVI
Tables -------------------------------------------------------------------------------------XXXIV
Acronyms -----------------------------------------------------------------------XLVI
INTRODUCTION ----------------------------------------------------------------------------3
PART 1: CONTEXT AND OBJECTIVES OF THE ASSESSMENT -------9
1 Context of the Ethiopian Health System ----------------------9
1.1 Health status -------------------------------------------------9
1.2 Health policy --------------------------------------------10
1.3 Organization of the health system ----------------------12
1.4 The Health Extension Program ----------------------13
1.5 Simplified logic model of the

Health Extension Program --------------------------------25
2 Objectives of the Assessment -----------------------------------27
2.1 General objective -------------------------------------------27
2.2 Specific objectives --------------------------------------------27
page-XVIII
PART 2: ASSESSMENT OF THE RURAL HEALTH EXTENSION
PROGRAM -------------------------------------------------------------29
SECTION I: METHODS OF THE RURAL HEALTH EXTENSION

PROGRAM ASSESSMENT ---------------------------------------29
1 General Approach -------------------------------------29
1.1 Study area and study period -------------------------29
1.2 Framework and general approach -------------------------29
2 Population and Sampling -------------------------------------31
2.1 Study population -------------------------------------31
2.2 Sample size determination ----------------------------------31
2.3 Sampling strategy -------------------------------------34
2.4 Inclusion and exclusion criteria -------------------------35
3 Data collection -------------------------------------35
3.1 Development of data collection tools ---------------35
3.2 Data collection, supervision, and

coordination team ---------------------------------36
3.3 Training of data collectors and supervisors -----37
3.4 Data collection fieldwork --------------------------------37
4 Data Management -------------------------------------39
4.1 Quantitative data management and analyses -----39
4.2 Qualitative data management and analyses -----40
5 Methods for systematic review of the

Health Extension Program literature -------------------------41
page- XIX
6 Methods of synthesis of evidence and
formulation of recommendations -----------------------------42
7 Limitations of the National Assessment Health Extension

Program -------------------------------------------------------43
7.1 Assessment of resource adequacy ----------------------44
7.2 Assessment of the availability of services at HPs -----44
7.3 Measurement of exposure to the HEP ---------------45
7.4 Assessment of the HEP’s contribution ---------------45
SECTION II: RESULTS OF THE RURAL HEALTH EXTENSION

PROGRAM ASSESSMENT -------------------------------------------46
1 Sample and Background Characteristics of Study

Participants --------------------------------------------------------47
1.1 Socio-demographic characteristics of household

survey respondents -----------------------------------49
1.2 Characteristics of key informants and FGD

participants --------------------------------------------50
2 Relevance of Health Extension Program

Packages and Service Delivery Strategies -------------------------52
Summary of key findings --------------------------------------------53
Introduction -------------------------------------------------------54
2.1 Epidemiological relevance of HEP packages -----55
2.2 Relevance of HEP service delivery modalities and

approaches ----------------------------------------------------62
2.3 Adaptability of HEP packages and service delivery

modalities ----------------------------------------------------75
page-XX
3 Inputs of the Health Extension Program ------------------------78
3.1 Human resources in the HEP ------------------------79
3.2 Facilities and infrastructure ----------------------------122
3.3 Management of drugs and medical supplies

at health posts --------------------------140
3.4 Financing of the Health Extension Program ----160
4 Health Service Delivery through the Health

Extension Program -------------------------------------------168
Introduction ------------------------------------------------------169
Summary of key findings ----------------------------------170
4.1 Physical accessibility of HPs and

availability of services --------------------------171
4.2 Community awareness of and familiarity with

HEP services --------------------------------------------173
4.3 Community members’ level of

exposure to the HEP ----------------------------------176
5 Community Engagement and Ownership in the Health

Extension Program --------------------------------------------198
Introduction ------------------------------------------------------199
5.1 Awareness of and enrollment in

model family training ---------------------------------200
5.2 Providers of model family training ---------------------203
5.3 Participation of community structures in Health

Extension Program implementation -------------204
5.4 Characteristics of Women’s Development

Army leaders -------------------------------------------210
page- XXI
6 Information System and Monitoring &
Evaluation in the Health Extension Program --------------214
Introduction ----------------------------------------------------215
Summary of key findings --------------------------------------------216
6.1 Information system ----------------------------------216
6.2 Data quality check ----------------------------------236
6.3 Monitoring and evaluation --------------------------244
7 Coverage of Health Extension Program-Related

Services ----------------------------------------------262
7.1 Coverage of hygiene and sanitation services ----263
7.2 Coverage of disease prevention and

control interventions ------------------------------287
7.3 Coverage of Family Health Services --------------339
8 Health Extension Program Service Delivery Outcomes ----398
8.1 Trends in health indicators in Ethiopia --------------399
8.2 Effectiveness of the Health Extension Program:

Evidence from a systematic review of
previous studies -------------------------------------------408
8.3 Association between intensity of Health Extension

Program implementation and household-level adoption
of health behaviors ------------------------------------414
9 Governance, Leadership, and Management

of the Health Extension Program --------------------------418
Introductions ------------------------------------------------------------419
Key findings --------------------------------------------------------------420
page-XXII
9.1 Overall governance and evolution
of the health sector ----------------------------421
9.2 Existence of an enabling policy environment ----422
9.3 Planning in the Health Extension Program ----------430
9.4 Decentralization of the Health Extension

Program --------------------------------------------------434
9.5 Coordination of Health Extension

Program activities -------------------------------------------440
9.6 Collaboration with other sectors ------------------------441
SECTION III: CONCLUSIONS AND RECOMMENDATIONS ----448
1 Health Extension Program service packages --------------448
2 Service delivery modalities --------------------------------------450
3 Implementation of the Health Extension Program --------------452
4 Human resources for the Health Extension Program ----454
5 Physical facilities, infrastructure, and basic utilities -------------456
6 Equipment, drugs, and other medical supplies --------------458
7 Financing the Health Extension Program -------------------460
8 Community engagement and ownership -------------------461
9 Information system and monitoring & evaluation -------------463
10 Governance and leadership --------------------------------------465
Reference -----------------------------------------------------------------------468
Annex: Methods for analyses of HEP

financing landscape -----------------------------------------------------476
page- XXIII
Part 3: Assessment of the Urban Health Extension Program ----484
1: Methods of the Urban Health

Extension Program Assessment ------------------------------485
2: Findings: Relevance of the Urban HEP ------------------------494
3: Findings: Availability and adequacy of

resources for UHEP --------------------------------------------539
4: Findings: Urban HEP workforce analyses ----------------------557
5: Findings: Implementation of the UHEP -----------------------579
6: Conclusions and Recommendations --------------------------------655
References ----------------------------------------------------------------663
Part 4: In-Depth Exploration of Specific Topics -----------------------666
Study 1: Assessment of the quality of Health Extension Workers’

Training institutions -------------------------------667
Study 2: Attrition among Health Extension Workers ----------751
Study 3: The role of the Health Extension Program in

Public Health Emergency Management --------------835
Study 4: Cost-Effectiveness Analysis of the

Health Extension Program ---------------------------------921
page-XXIV
page- XXV
FIGURES
Part 1
Figure 1-1. Ethiopian Health Service delivery tier system --------------13
Figure 1-2. Health Extension Program structure integrated into

primary healthcare units --------------------------------------------------------20
Figure 1-3. Administrative structure of the Health

Extension Program ------------------------------------------------------------21
Figure 1-4. Simplified program logic model of the Health Extension

Program at the kebele level --------------------------------------------26
Part 2
SECTION I
Figure 1-1. PHCPI framework adapted for assessment of the Health

Extension Program in Ethiopia -----------------------------------------------31
Figure 1-2. Components of the National

Assessment of the HEP --------------------------------------------------------32
Figure 1-3. Administrative map of Ethiopia -------------------------------34
SECTION II
Figure 2-1. Major causes of premature deaths in

Ethiopia, 2007-2017 ------------------------------------------------------------55
Figure 2-2. Major causes of disability in Ethiopia, 2007-2017 -----56
Figure 2-3. Risk factors driving deaths and disability

in Ethiopia, 2007-2017 ----------------------------------------------------------56
Figure 2-4. Households with at least one sick member in

the past 1 month -------------------------------------------------------------------58
page-XXVI
Figure 2-5. Reported causes of most recent illnesses among
households with at least 1 sick member during the
month period preceding the survey -------------------------------------58
Figure 2-6. Percentage of household members who

recommended additional services at HPs in response
to their health needs ---------------------------------------------------------62
Figure 3-1. Reasons for HEWs’ absence during the time of visit -----92
Figure 3-2. HEWs’ CoC certification status, by the

number of times taking the CoC test -------------------------------------94
Figure 3-3. HEWs’ knowledge of EPI schedule, by

level of education ----------------------------------------------------------------96
Figure 3-4. HEWs’ recommending IRT to other HEWs, by region --102
Figure 3-5. Reasons for intending to leave one’s job

among HEWs affirming that they had a plan to change
their jobs ------------------------------------------------------------------------112
Figure 3-6. Mean availability of basic amenities, by region ---------136
Figure 3-7. Percentage of HPs out of stock of tracer drugs

on the day of visit ---------------------------------------------------------------149
Figure 3-8. Total health center and health post spending,

2010/11 to 2016/17, in billion ETB --------------------------------------------162
Figure 3-9. Total health center and health post

expenditure, by source, 2010/11 to 2016/17 ----------------------------163
Figure 3-10. Magnitude of HEP spending (In billion ETB)

and share of the HEP from total PHCU (both health center
and posts) level expenditure 2010/11 to 2016/17 -------------------164
Figure 3-11: Magnitude and share of HEP the HEP from

the total health spending, 2010/11-2016/17 ----------------------------165
Figure 3-12. Total HEP spending trend, by source, 2010/11-2016/17 -166
Figure 3-13. Average total HEP spending share, by economic

classifications, 2010/11 to 2016/17 ----------------------------------------------167
page- XXVII
Figure 3-14. Average total HEP spending, by input type,
2010/11 to 2016/17 -----------------------------------------------------------------164
Figure 4-1. Service availability mean score, by service

category or package --------------------------------173
Figure 4-2. Women’s awareness of and familiarity

with HEP services ------------------------------------------------174
Figure 4-3. Home visits by HEWs --------------------------176
Figure 4 4. Exposure of household members to the HEP,

by region --------------------------------------------------------------------182
Figure 4-5. Source of exposure to the HEP among women, men,

and youth girls ----------------------------------------------183
Figure 4-6. Household-level implementation of the HEP among

ordinary and WDA households -----------------------------189
Figure 5-1. Model family training among ordinary and Women’s

Development Army households --------------------201
Figure 5-2. Trainers of model families ---------------------------203
Figure 5-3. Role of women who completed model family training ---204
Figure 5-4. Types of community structures supporting the Health

Extension Program, by livelihood ---------------------------------------205
Figure 5-5. WDA density, by livelihood ------------------------------206
Figure 5-6. Functionality of WDA structures in agrarian settings ---207
Figure 5-7. Composition of social mobilization committees

in pastoralist kebeles -------------------------------208
Figure 5-8. Educational status of WDA leaders compared with women

from the general population ------------------------------210
Figure 5-9. Wealth quintiles of WDA leaders compared with women

from the general population ----------------------------------------211
page-XXVIII
Figure 6-1. Number of indicators at national and HP level and
indicators with data sources, by program category ------------------218
Figure 6-2 Percent of HPs implemented CHIS (FF) -------------------222

Figure 6-3. Percentage of Health Extension Workers receiving
CHIS training --------------------------------------------222
Figure 6-4. M&E Structure and functions index, by region ----------223
Figure 6-5. CHIS Input or Resource index, by region ---------------225
Figure 6-6. CHIS recording and reporting tools and forms

index, by region ----------------------------------------------------------229
Figure 6-7. Data-management process index, by region ------------232
Figure 6-8. Information use index, by region ------------------233
Figure 6-9. Spider graph, CHIS functional areas, national level ---235
Figure 6-10. Spider graph, CHIS functional areas, by region ---236
Figure 6-11. Result verification ratio at national and sub-national

levels, by indicator --------------------------------------------------239
Figure 6-12. Result verification ratio of child health indicators at

national and sub-national levels, by indicator -----------------239
Figure 6-13. Result verification ratio at regional level, by

maternal health indicator and region --------------------240
Figure 6-14. Result verification ratio at regional level, by

child health indicators ---------240
Figure 6-15. HPs’ level of accuracy, sub-national level, by

indicator ----------------------------------------------------------242
Figure 6-16. Record or data consistency between tally sheet and health
cards or registers, by indicator ---------------------------------------242
Figure 6-17. Health post annual plan preparation process ----------246
Figure 6-18. Community involvement in health post Planning ---248
page- XXIX
Figure 6-19. Community involvement in
Health Extension Program performance review ---------------------251
Figure 6-20. Number of supervisions to health posts in 6 months, by

supervisor type ------------------------------------------254
Figure 6-21. Feedback type, by supervisor type and livelihood ---255
Figure 6-22. Challenges in communicating feedback between

supervisory institution and HP --------260
Figure 7-1. Sanitation facilities, by wealth quintile --------------272
Figure 7-2. Place of cooking among households, by region ----------282
Figure 7-3. Households receiving health education on

handwashing by HEWs, by region ----------------------------------------286
Figure 7-4: Providers of HIV counseling at ANC, by livelihood ---316
Figure 7-5: Household occurrence of illness and health seeking, by

wealth index --------------------------------------------------------------327
Figure 7-6. Method-based FP knowledge of women, men and

youth girls ----------------------------------------------------340
Figure 7-7. Share of long-acting and short-acting contraceptive

methods among current contraceptive users ------------------345
Figure 7-8. Reasons for not using any contraceptive methods among
married women who do not want to get pregnant soon ---------346
Figure 7-9. Place of delivery for most recent birth during the
last 5 years -----------------------------------------350
Figure 7-10. Treatments applied to the umbilical

cord after delivery -------------------------------------------------------------350
Figure 7-11. Content of postnatal care for the mother ------------353
Figure 7-12. Content of postnatal care for the newborn ------------354
Figure 7-13. Mothers’ reasons for not getting their

children vaccinated --------------------------------------------363
page-XXX
Figure 7-14: Percentage distribution of children aged 6- to 23 months
with minimum acceptable diet, by region --------------------373
Figure 7-15. Proportion of selected services provided by HEWs at
health posts or in home visits among those who got these services from
any healthcare provider --------------------------------------------------381
Figure 7-16. Place of FP service delivery for women and

youth girls -------------------------------------382
Figure 7-17: Primary sources of information on family planning ---383
Figure 7-18: Primary source of information about danger signs of

pregnancy among mothers who delivered during the last 5 years ---385
Figure 7-19: Source of advice or treatment for children with

symptoms of ARI ----------------------------------------------------------------392
Figure 7-20: Source of advice or treatment for children with

diarrhea -------------------------------------------------------------------393
Figure 7-21: Components of ANC received for the last pregnancy,

by provider category -------------------------------------------------394
Figure 7-22: Place of delivery, by place of ANC visit ---------------395
Figure 8-1: Trends in Maternal Mortality, 1990-2015, Ethiopia ------------400
Figure 8-2: Trends in childhood mortality indicators,

2000-2016, Ethiopia ------------------------------------401
Figure 8-3: Malaria deaths, 2002-2015, Ethiopia ------------------402
Figure 8-4: Nutritional status, children under 5, 2000-2016 -----------403
Figure 8-5: Trends in number of confirmed

malaria cases, 2002-2015 --------------------------------------------404
Figure 8-6: Coverage of key maternal health services ------------405
Figure 8-7: Coverage of WaSH Facilities -----------------------407
Figure 8-8: Trends in Tuberculosis treatment coverage ----------407
page- XXXI
Figure 8-9. A random effect model-analysis of the effects of the
HEP on maternal healthcare service use ----------------------411
Figure 8-10. Mean progress toward full implementation of the
HEP at the household level --------------------------------------416
PART 3
Figure 1-1. Schematic representation of sampling strategies --------489
Figure 2-1: Community Perception about Relevance of UHEP ---499
Figure 2-2: UHEPr’s Perceptions about Acceptability

of the Program --------------------------------------------------------528
Figure 2-3: Community Perception of UHEPr’s Friendliness ------------533
Figure 2-4: Community Perceptions of UHEW’s Courtesy ----------534
Figure 2-5: Overall Satisfaction of the Community

with the UHEP -------------------------------------------------536
Figure 3-1: Training Status of UHEP Coordinators

at Health Center level --------------------------------------------------542
Figure 3-2: Integrated Refresher Training Status

of UHEPr ----------------------------------------------------------------------545
Figure 3-3: Urban Health Extension

Professional-to-Household Ratio ---------------------------------------546
Figure 3-4: Availability of Pharmaceuticals, Medical

Equipment, and Supplies -----------------------------548
Figure 3-5: Availability of UHEP Implementation Guidelines among

Health Centers --------------------------------------552
Figure 4-1: Community Perceptions about UHEPrs’

Knowledge and Skills ---------------------------------------558
Figure 4-2: General Satisfaction of UHEPr Working in UHEP ---565
page-XXXII
Figure 4-3: UHEPr’ Perceived Trends in their
Level of Satisfaction -------------------------------------------------567
Figure 4-4. Attrition per 10 000 HEWs since the Implementation of the
HEP from 2007 to 2018 ------------------------------568
Figure 4-5: Attrition per 10, 000 since the Implementation of the
HEP from 2007 to 2018 ------------------------------------568
Figure 4-6: Percentage of UHEPr Currently Looking

for Another Job -----------------------------------------------571
Figure 4-7: Major Reasons for Intention To Leave Job ------------572
Figure 4-8: UHEPrs’ Tendency to Recommend

Others to Work as UHEPr ------------------------------------574
Figure 5-1: Performance of UHEPrs in Conducting

Routine Household Visits ------------------------------------------581
Figure 5-2. Households’ Model Family Training Status --------586
Figure 5-3. Households` Reasons for Not Taking

Model Family Training ----------------------------------------------587
Figure 5-4: UHEPrs’ Feelings about the FHT -------------------602
Figure 5-5: Recommendation of UHEPr to Scale Up the

FHT Approach to Other Settings ---------------------------------------602
Figure 5-6: Implementation of UHEP as Perceived by the

Community, Response from Household Survey -------------608
Figure 5-7: Supportive Supervision System and Practice by

HCs on the Implementation of UHEP -----------------------------------634
Figure 5-8: Major Challenges of UHEP Implementation

Reported by Heads of HCs -----------------------------------------648
PART 4
List of figures are presented within each specific study
page- XXXIII
TABLES
Part 2
SECTION I
Table 1-1. Eligibility criteria for study participants -----------------------35
Table 1-2. Health Extension Program assessment recommendation

formulation workshops and consultations --------------------------------43
SECTION II
Table 1-1. Number of sample woredas, health posts, health centers, and
Health Extension Workers -------------------------------------------------------47
Table 1-2. Number of households, women, men, and youth girls

among households from the general population and from Women’s
Development Army/SMC households ----------------------------------------48
Table 1-3. Socio-demographic characteristics of women, men, and youth

girl respondents, by type of household --------------------------------------49
Table 1-4. Characteristics of key informants and focus

group discussion participants --------------------------------------------------51
Table 2-1. Perceived epidemiological relevance of the HEP -------------60
Table 2-2. Community members recommending the addition of HP

services to current offerings, by region and livelihood --------------61
Table 2-3. Percentage of household members who agreed on the

appropriateness of home visits as an HEP service delivery
modality, by region and livelihood ----------------------------------63
Table 2-4. Percentage of household members who agreed with

statement about HEWs’ gender, by region and livelihood ------------66
Table 2-5. Perceived trustworthiness and friendliness of HE -----------69
page-XXXIV
Table 2-6. Household members’ perceived
acceptability of the HEP -------------------------------------------------------73
Table 3-1. Trainees’ and instructors’ ratings of the relevance

of the HEP curriculum ---------------------------------------------------------84
Table 3-2. State of HEP training facilities in Ethiopia, 2019 ---------------85
Table 3-3. Provision of orientation and a job description to HEWs ----87
Table 3-4. Staffing level of HPs, by region and livelihood ----------90
Table 3-5. Qualification and CoC certification of HEWs,

by livelihood (N=584) ---------------------------------------------------------94
Table 3-6. HEWs’ knowledge of EPI schedule and malaria treatment

guidelines (N=584) ---------------------------------------------------------96
Table 3-7. HEWs’ self-reported competence in attending labor,

detecting signs of danger in delivery, and providing
long-term FP, by background characteristics --------------------------99
Table 3-8. HEWs’ participation in IRT ----------------------------------101
Table 3-9. Short-term trainings attended by HEWs, by

training topic and livelihood (N=584) ------------------------------------103
Table 3-10. Job-related satisfaction among HEWs ------------------107
Table 3-11. Distribution of level of satisfaction of HEWs, by

background characteristics ---------------------------------------------108
Table 3-12. HEWs burnout status, by background characteristics -----111
Table 3-13. HEWs’ intention to leave their job, by region

and livelihood -----------------------------------------------------------------112
Table 3-14. Prevalence of probable symptom of depression,

by background characteristics ------------------------------------------------116
Table 3-15. Community members’ involvement in the

performance assessment of HEWs -----------------------------------------118
page- XXXV
Table 3-16. Performance appraisal of HEWs, by
region and livelihood ----------------------------------------------------------120
Table 3-17. Availability of written feedback for the

most recent performance assessment -------------------------------------121
Table 3-18. Community members’ involvement in

the performance assessment of HEWs --------------------------------------122
Table 3-19. HP and HEWs to population ratio, by

region and livelihood --------------------------------------------------------124
Table 3-20. Road connectivity and accessibility of HPs,

by region and livelihood ---------------------------------------------------------125
Table 3-21. Means of transportation used to conduct

outreach activities ----------------------------------------------------------------127
Table 3-22. Availability of basic facilities and

infrastructure at health posts --------------------------------------------------129
Table 3-23. Building materials used to construct health post, by

livelihood (N=343) --------------------------------------------------------------130
Table 3-24. Health posts meeting the building standards -------------------132
Table 3-25. Availability of incinerators, placenta pit, and

open pit, by region and livelihood --------------------------------------133
Table 3-26. Availability of basic utilities in the HPs --------------------135
Table 3-27. Availability of functional basic

equipment at HPs (N=343) ----------------------------------------------138
Table 3-28. Availability of basic equipment at health posts ------------139
Table 3-29. Availability of IP equipment, by

livelihood -------------------------------------------------------140
Table 3-30. Drugs and supplies recording and reporting formats ----145
Table 3-31. Availability of tracer drugs in the HP

at the time of visit -------------------------------------------------------------147
page-XXXVI
Table 3-32. Availability of tracer drugs in HPs at time of visit -----------148
Table 3-33. Percentage of HPs experiencing stockout of

tracer drugs in the past 6 months ---------------------------------151
Table 3-34. Longest duration of stockout of drugs in the

last 6 months ---------------------------------------------------------------153
Table 3-35. Inventory management implementation among HPs -------155
Table 3-36. Pharmaceutical storage conditions among study HPs ----157
Table 3-37. ETB to PPP and USD Conversion Factors

for Ethiopia, 2010/11 - 2016/17 -----------------------------------------------161
Table 4-1. Availability of selected HEP services --------------------------------172
Table 4-2. Awareness of and familiarity with services

provided by HEWs ----------------------------------------------------------------175
Table 4-3. Proportion of households receiving visits by HEWs, by

background characteristics of households --------------------------------177
Table 4-4. Characteristics of HEW visits to households among those

having at least 1 visit during the 1-year period preceding
the survey, by livelihood ----------------------------------------------------------178
Table 4-5. Health education topics covered among households that

received health education during the most recent HEW visit during
the 1-year period preceding the study --------------------------------------------------179
Table 4-6. HP visits during the last one year ---------------------------------180
Table 4-7. Health education or other services received in settings other

than home or HP, by respondent category and livelihood ---------------181
Table 4-8. Exposure of respondents to the HEP – Women ------------184
Table 4-9. Exposure of respondents to the HEP – Men ---------------------185
Table 4-10. Exposure of respondents to the HEP – Youth girls -------------187
Table 4-11. Criteria for assessment of the HEP implementation at

household level --------------------------------------------------------------188
page-XXXVII
Table 4-12. Household-level implementation of the HEP --------------190
Table 4-13. Factors associated with level of implementation of

the HEP through home visits, HP visits, and outreach sessions --------------193
Table 5-1. Proportion of households’ awareness of and enrollment

status in model family training, as reported by their female
representatives -----------------------------------------------------------------202
Table 5-2. Comparison of health behavior indicators between

women in the general population and WDA Leaders -----------------------211
Table 6-1. M&E structure, functions and capabilities of

CHIS for the HP survey ---------------------------------------------------------221
Table 6-2. Input or resources for CHIS of the health post survey --------224
Table 6-3. Availability of data collection and reporting forms,

tools and guidelines for CHIS, health post survey ---------------------227
Table 6-4. Data management process in CHIS, health post survey ---231
Table 6-5. Verification factor – standard -----------------------------------------234
Table 6-6. Health posts’ level of data accuracy, by indicators ------------237
Table 6-7. Information use at HPs -------------------------------------------------241
Table 6-8. Availability of health post plan (2011 EFY) -----------------------245
Table 6-9. Community involvement in health post planning --------------247
Table 6-10. Reported challenges in health post planning --------------249
Table 6-11. Community involvement in Health Extension Program

performance review ---------------------------------------------------------------250
Table 6-12. Availability and characterization of supervision of HPs ---255
Table 6-13. Knowledge of supervisors and usefulness

of supervision of HPs ---------------------------------------------------------256
Table 7-1. Key indicators in hygiene andenvironmental health --------------265
page-XXXVIII
Table 7-2. Basic and limited drinking water sources ------------------------267
Table 7-3. Drinking water, by livelihood, region, and

household wealth quantile -------------------------------------------------------268
Table 7-4. Water treatment methods ------------------------------------------270
Table 7-5. Availability of toilet or latrine facility ----------------------------271
Table 7-6. Sanitation facilities, by background characteristics -----------273
Table 7-7. Availability and adequacy of handwashing facilities -----------274
Table 7-8. Handwashing education and practice ----------------------------277
Table 7-9. Liquid and solid waste disposal ----------------------------------279
Table 7-10. Sleeping rooms and separating livestock

inside the house ---------------------------------------------------------------------281
Table 7-11. Type of fuel used and indoor air pollution ----------------------283
Table 7-12. Protection of food from flies and abnormal presence of

insects and rodents ------------------------------------------------------------------285
Table 7-13. Household possession of ITNs ------------------------------------289
Table 7-14. Access to an insecticide-treated net (ITN) -------------------290
Table 7-15. Use of ITNs and existing ITNs used in the household ----291
Table 7-16. Use of ITNs by children and pregnant women --------------292
Table 7-17. Awareness of HIV: Women, men, and youth girls -------------293
Table 7-18. Knowledge of HIV prevention methods: Women,

men, and youth girls ---------------------------------------------------------------296
Table 7-19. Comprehensive knowledge of HIV: Women,

men, and youth girls -------------------------------------------------------------298
Table 7-20. Knowledge of prevention of mother-to-child

transmission of HIV -----------------------------------------------------301
page- XXXIX
Table 7-21. Counseling on HIV during ANC for most
recent pregnancy ---------------------------------------------------------------304
Table 7-22. Providers of information and counseling on

HIV during ANC -------------------------------------------------------------------------306
Table 7-23. Pregnant women counseled and tested for HIV -------------308
Table 7-24. Knowledge of tuberculosis prevention methods -------------311
Table 7-25. Knowledge of women, men, and youth girls about modes of
transmission and prevention of tuberculosis, by
background characteristics -----------------------------------------------------313
Table 7-26. ITN demonstration, by source of education and

background characteristics -----------------------------------------------------315
Table 7-27. Source of information about tuberculosis (TB) --------------318
Table 7-28. The role of HEWs in TB diagnosis and management ---320
Table 7-29. Tuberculosis (TB) symptoms and health seeking --------------321
Table 7-30. Place of first contact for tuberculosis (TB)-related

care and reasons for bypassing health posts ----------------------------------322
Table 7-31. Places of tuberculosis (TB) treatment --------------------------------323
Table 7-32. Morbidity among household members ---------------------325
Table 7-33. Injury and type of accident -------------------------------------326
Table 7-34. Place of treatment and referral from health facilities

among household members -----------------------------------------------------329
Table 7-35. Information received from HEWs on NCDs and

cervical cancer screening among women ------------------------------------------331
Table 7-36. Information received from HEWs on NCDs among men ----333
Table 7-37. Occurrence of any NCD in households ------------------------335
Table 7-38. Place of diagnosis for NCD and reason for

bypassing health posts ---------------------------------------------------------------336
page-XL
Table 7-39. Suicidality among women in Ethiopia -------------------------------338
Table 7-40. Contraceptive prevalence rate among women

and youth girls -------------------------------------------------------------------------343
Table 7-41. Percentage of need and demand for FP among

reproductive-age (15-49) women ----------------------------------------------------344
Table 7-42. Coverage of ANC visits ----------------------------------------347
Table 7-43. Timing of first ANC visit -------------------------------------------348
Table 7-44. Women’s reason for not having ANC visits for their
most recent birth -------------------------------------------------------------------------349
Table 7-45. Place of delivery for the most recent birth in

the last 5 years -------------------------------------------------------------------------351
Table 7-46. Delivery attendance for the most recent live birth
in the last 5 years -------------------------------------------------------------------------352
Table 7-47. Postnatal visits for the youngest child in the last 2 years ----354
Table 7-48. Prevalence of ARI and treatment of ARI symptoms ---357
Table 7-49. Prevalence and treatment of fever for youngest

children under 5 -----------------------------------------------------------------358
Table 7-50. Prevalence and treatment of diarrhea for

youngest children under 5 ------------------------------------------------------359
Table 7-51. Oral rehydration therapy, zinc, and other

treatments for diarrhea --------------------------------------------------------360
Table 7-52. Percentage of children aged 12-23 months who

received basic vaccines ------------------------------------------------------364
Table 7-53. Percentage of children aged 12-23 months who received

PCV and Rota vaccines --------------------------------------------------------366
Table 7-54. Percentage of children aged 12-23 months who received

complete, partial, and no vaccinations ------------------------------------367
Table 7-55. Mothers’ reasons for not vaccinating their children --------------369
page- XLI
Table 7-56. Breastfeeding status of youngest children under age 2 ----374
Table 7-57. Breastfeeding frequency, initiation of complementary

feeding, and bottle feeding among children aged 6-23 months -------------376
Table 7-58. Minimum meal frequency -------------------------------------------377
Table 7-59. Food groups consumed by children under age 2 in the day or
night preceding the interview -----------------------------------------------------379
Table 7-60. Percentage of youngest children aged 6-23 months who were fed
a minimum dietary diversity during the day or night before the survey ----379
Table 7-61. Percentage of youngest children aged 6-23 months who are fed
iron-rich or iron-fortified food, HEW visits for IYCF, and
knowledge of complementary feeding -------------------------------------------380
Table 7-62. Primary source of information about family planning,

by livelihood -------------------------------------------------------------------------383
Table 7-63. Place of first antenatal care visit among mothers who had at
least 1 visit during their most recent pregnancy --------------------------------386
Table 7-64. Place of fourth antenatal care visit among mothers

who had at least 4 visits during their most recent pregnancy -------------387
Table 7-65. Information about birth preparedness plan, by

source of information ---------------------------------------------------------------388
Table 7-66. Service providers for postnatal visits for most recent
delivery during the last 2 years -----------------------------------------------------389
Table 7-67. Place of PNC visit among mothers who had a PNC visit
for their most recent delivery during the last 2 years -----------------------390
Table 7-68. Place of PNC visit among mothers who had a PNC visit
for their most recent delivery during the last 2 years ---------------------391
Table 8-1. Fertility status indicators, 2000-2016 -------------------------------406
Table 8-2. Coverage of key child health services (in percent) -------------406
page-XLII
Table 8-3. Association between intensity of HEP implementation
and household-level behavior -----------------------------------------------------415
Table 9-1. Percentage of HPs with HEP guidelines, by region

and livelihood -------------------------------------------------------------------------426
Table 9-2. Percentage of HPs reporting availability of annual plan and

community participation in the planning processes, by livelihood -----431
Table 9-3. Percentage of HPs reporting availability of

annual plan and community participation in the planning and
review process, by region -----------------------------------------------------432
Table 9-4. Percentage of HCs reporting their decentralized

roles in the HEP implementation -------------------------------------------435
Table 9-5. Percentage of WorHOs reporting their decentralized

roles in the HEP implementation ----------------------------------------------------436
Table 9-6. Percentage of HEWs reporting their perceived agreement

with the support they receive from kebele administration, HC,
and WorHO, by region and livelihood ---------------------------------------439
PART 3
Table 1-1: Sample size calculation for household survey ----------------------487
Table 1-2. Sample size for qualitative study ----------------------------491
Table 2-1. Comparison of model and non-model households by

awareness and practice of urban residents --------------------------512
Table 2-2. Effect of home visits on awareness and

practices of urban residents -------------------------------------------519
Table 3-1: Administrative and Technical Roles of Health Centers for

implementation of the UHEP --------------------------------------539
Table 3-2: Training, Education, and Career Development

Opportunities for UHEPr at HC Level ----------------------------------541
page- XLIII
Table 3-3: Logistics and Supply Chain Management of
UHEP at Health Center Level ------------------------------------------549
Table 4-1. Self-Reported Level of UHEPr’s Competency to

Conduct or Provide Health Services ---------------------------------------557
Table 4-2. Level of satisfaction of UHEPr by City Administration ----564
Table 4-3. Overall Satisfaction as UHEW by some

important background characteristics --------------------------------------566
Table 4-4. Magnitude of attrition by different background variables ----570
Table 4-5. Level of Burnout among Urban Health

Extension Professionals --------------------------------------------------------575
Table 4-6. Mental Health Status of UHEW ---------------------------577
Table 5-1. Frequency of Household Visit by UHEPr ----------------581
Table 5-2: Preparation, Planning and Monitoring of FHT Activities ----589
Table 5-3: Establishment of baseline community profile and

regular updating practices by the Family
Health Team by City Administration -----------------------------------------591
Table 5-4: Distribution of services provided through the

Family Health Team by City Administration ------------------------------592
Table 5-5: Distribution of delivery points used by the

Family Health Team to provide services ---------------------------------------594
Table 5-6: Planning, monitoring and evaluation practice of

the Family Health Team, UHEP assessment --------------------------------596
Table 5-7. Availability of basic equipment for

Family Health Team activities -----------------------------------------------598
Table 5-8: Availability of supply management system,

manuals and IEC–BCC materials for Family Health Team -------------599
page-XLIV
Table 5-9. Role of WDAs in the implementation of
Family Health Team activities ---------------------------------------600
Table 5-10. Involvement of different government sector offices in the

implementation of the FHT approaches -----------------------------------601
Table 5-11. Perceived Level of UHEP Package implementation

among Households -------------------------------------------------------------609
Table 5-12: UHEP Health Information System implementation

among HC in urban centers -----------------------------------------631
Table 5-13: UHEP Reporting and Data Quality

Assessment Practice in Ethiopia --------------------------------------------636
Table 5-14: Planning and Performance Monitoring for

UHEP at HC Level ------------------------------------------------------------645
PART 4
List of tables are presented within each specific study
page- XLV
ACRONYMS
ANC Antenatal care

AOR Adjusted Odds Ratio
ARI Acute respiratory infection
ARM Annual Review Meeting
ART Antiretroviral Therapy
BCG Bacillus Calmette-Guerin
CBHI Community-based health insurance
CBNC Community-based newborn care
CBRHA Community-based reproductive health agents
CD Communicable disease
CEA Cost effectiveness analysis
CEBS Community event-based surveillance
CHIS Community Health Information System
CHP Community-based health planning
CHW Community health worker
CI Confidence interval
CJSC Central Joint Steering Committee
CLTS Community-Led Total Sanitation
Communicable, maternal, neonatal, and nutritional
CMNND
diseases
CoC Certificate of Competency
CPD Continuous Professional Development
CPR Contraceptive prevalence rate
CSA Central Statistics Agency
page-XLVI
CYP Couple year of protection
DALY Disability-adjusted life year
DHIS District Health Information System
DHO District Health Office
DOTS Directly observed treatment short course therapy
DPC Disease prevention and control
DQ Data quality
EDHS Ethiopian Demographic and Health Survey
EFY Ethiopian Fiscal Year
EPHI Ethiopian Public Health Institution
EPI Expanded program of immunization
EPSA Ethiopian Pharmaceuticals Supply Agency
ETB Ethiopian birr
FEFO First Expiry First Out
FF Family Folder
FGD Focus Group Discussion
FGM Female genital mutilation
FHT Family Health Team
FP Family planning
GBD Global Burden of Disease
GDP Gross domestic product
GoE Government of Ethiopia
GDP Gross Domestic Product
GTP Growth and Transformation Plan
HC Health Center
HEP Health Extension Program
page- XLVII
HEW Health Extension Worker
HIS Health Information System
HIV Human immunodeficiency virus
HMIS Health Management Information System
HP Health post
HPARR Health post-adapted requisition and reporting format
HPV Human papilloma virus
HR Hazard Ratio
HRH Human Resources for Health
HSDP Health Sector Development Program
HSDP-II Health Sector Development Program
HSTP Health Sector Transformation Plan
iCCM Integrated Community Case Management
ICER Incremental cost-effectiveness ratio
IDSR Integrated Disease Surveillance and Response
Information Education Communication/Behavior Change
IEC/BCC
Communication
IMNCI Integrated management of newborn and childhood illness
IPLS Integrated Pharmaceuticals Logistics System
IQR Interquartile range
IRS Indoor residual spray
IRT Integrated Refresher Training
ISS Integrated supportive supervision
IST In-service training
ITN Insecticide-treated bed net
IUD Intra-uterine device
IYCF Infant and young children feeding
page-XLVIII
KII Key Informant Interview
LLIN Long-lasting insecticidal net
LMIS Logistrics Management Information System
LQAS Lot Quality Assurance Sampling
LYG Life years gained
M&E Monitoring and evaluation
MAD Minimum acceptable diet
MCH Maternal and child health
MDD Minimum dietary diversity
MDG Millennium Development Goal
MMF Minimum meal frequency
MMR Maternal Mortality Ratio
MNCH Maternal, newborn, and child health
MoE Ministry of Education
MoH Ministry of Health
MTCT Mother-to-child transmission
NCD Non-Communicable Disease
NGO Non-governmental organization
NHA National Health Accounts
OCP Oral contraceptive pills
ODK Open Data Kit
Organization for Economic Cooperation and
OECD/DAC
Development – Development Assistance Committee
OHEP Optimization of Health Extension Program
OPD Outpatient Department
OPV Oral Polio Vaccine
OR Odds Ratio
page- XLIX
ORS Oral Rehydration Salt
ORS Oral rehydration solution/salt
OTP Outpatient Therapeutic feeding Program
Program Accelerated and Sustained Development to End
PASDEP
Poverty
PCV Pneumococcal Conjugate Vaccine
PFSA Pharmaceuticals Fund and Supply Agency
PH Primary Hospital
PHC Primary Health Care
PHCPI Primary Health Care Performance Initiative
PHCU Primary Health Care Unit
PHE Public Health Emergency
PHEM Public Health Emergency Management
PMT Performance monitoring team
PNC Postnatal care
PPP Purchasing power parity
RHB Regional Health Bureau
RHF Recommended homemade fluid
RRF Report and Resupply Forms
RRT Rapid Response Team
SARA Service Availability and Readiness Assessment
SBA Skilled birth attendance
SDG Sustainable Development Goals
SMC Social Mobilization Committee
SNNPR Southern Nations, Nationalities and Peoples Region
SPA Service provision assessment
SRH Sexual and reproductive health
page-L
STI Sexually transmitted infection
TB Tuberculosis
TBA Traditional birth attendant
THE Total Health Expenditure
TIME Tuberculosis impact and modeling estimate
TT Tetanus Toxoid
TVET Technical and vocational education and training
TVETI Technical and vocational education and training institute
UHEP Urban Health Extension Program
UN United Nations
UNICEF United Nations Children’s Fund
US$ United States dollar
VF Verification factor
VHF Viral hemorrhagic fever
VIP Ventilated improved pit (latrine)
WaSH Water, hygiene, and sanitation
WDA Women’s Development Army
WHO World Health Organization
WHO/AFRO World Health Organization / Africa office
WorHO Woreda Health Office
ZHD Zonal Health Department
page- LI
The Health Extension Program
The Health Extension Program is a
flagship of the Ethiopian Health Sector
involving an innovative intervention
marked by the institutionalization of
community health services, government
leadership, and the alignment and
substantial support of development
partners.
page-2
Introduction
Ethiopia’s population is predominantly rural, with a rapid growth rate and

young age structure. These phenomena have remained unchanged for
over 2 decades. The country’s primary health problems remain preventable
communicable diseases; maternal, neonatal, and child health conditions; and
nutritional disorders, while non-communicable diseases (NCDs) are also on the
rise, predicting a double burden of disease.1 The underlying causes of these
health problems are poor socio-economic conditions, adverse health impacts
of climate change, food insecurity, lifestyle and nutritional habits, a low level
of awareness about health, and inadequate health service delivery across the
country. In response to these states of affairs, the Government of Ethiopia
(GoE) and its partners have made significant efforts that have brought
remarkable improvements in the health systems and health outcomes over the
last 2 decades. One such effort is the implementation of the Health Extension
Program (HEP).2-7
Ethiopia was an early adopter of the concept of primary healthcare (PHC) in

the Alma-Ata Declaration,8 putting in place the HEP as a strategy for reaching
its citizens with essential health services, among other services.4 The HEP is a
nationwide community-based health program that involves training 2 female
Health Extension Workers (HEW) per village, constructing a health post (HP),
and delivering 16 packages of carefully designed health services. This has been
considered an appropriate strategy for tackling the main health problems of
rural communities since its introduction in 2003. The program has evolved in
many ways since its inception, as have the national and global priorities and
health systems’ approaches. The agrarian HEP was adapted in various ways,
including in the designation of male HEWs, for pastoralist communities. In
addition, an UHEP was designed and implanted, with various adaptations
made to the package and composition of the cadres who deliver services. The
series of health strategic plans, as well as the Health Sector Development/
Transformation Plans (HSDP and HSTP), strongly emphasize ensuring universal
health coverage, as well as meeting the targets of the Millennium Development
Goals (MDG) and subsequently the Sustainable Development Goals (SDG),
with the HEP being a primary vehicle for doing so.2-4,6
page- 3
Despite these expectations and contributions, stakeholders have in recent
years been concerned by signs of deterioration in the performance
of the program that warrant prompt corrective actions. To this end,
the Ministry of Health (MoH), in collaboration with its partners and
funded by the Bill and Melinda Gates Foundation, has launched a
comprehensive national assessment of the HEP with the purpose of
understanding the status, determinants, and prospects of the program
and informing its programmatic and policy decisions. This assessment
was conducted by MERQ Consultancy Private Limited Company.
The purposes of this assessment were to generate the information

needed to meet the challenges and to form recommendations for
guiding the refinement and implementation of the program in the
coming decades. The general objective of the assessment was to
assess the status, determinants, and prospects of the HEP and identify
challenges and areas of intervention for programmatic and policy
decisions in the Ethiopian health sector. The study is a nationwide
assessment that looks into both the demand and supply sides through
a well-crafted approach that includes a review of the literature and
quantitative as well as qualitative methods. It touches on the critical
health system components based on the World Health Organization’s
(WHO) building blocks, while also looking deeper into selected areas
for deeper understanding.
The report is divided into four parts. Part 1 presents the context and
Part 1
objectives of the assessment in two sections. In the first section, it

describes the context of the Ethiopian health system, highlighting its tier
system, the health status of the population, and the policy environment,
and describing the HEP in both its content and modus operandi, as
stipulated in the strategic documents, manuals, and guidelines for its
implementation. In the second section, it lists the general and specific
objectives of the assessment.
Part 2 presents the methods and results of the assessment of the rural
Part 2
HEP. It includes 3 sections. Section I elaborates upon the methodology

of the study, including its scope, its 3 approaches (i.e., document review
and quantitative and qualitative approaches), and the health system
blocks investigated. It also details the study population, including
page-4
the sampling strategy that explains how the woredas, households,
health workers, and household members were sampled. The section
also describes study variables, data collection, and analysis methods.
Section II, the results section, describes the findings of the assessment
of the rural HEP, which includes agrarian and pastoralist settings. The
second section of Part 2 presents the findings of the assessment of the
rural HEP. While the focus of this section is on the rural HEP, some of
the findings presented also apply to urban settings. The results section
Chapter 1
is further divided into 9 chapters.
Chapter 1 presents a summary of the general characteristics of the

respondents to the quantitative and qualitative data collection activities.
Chapter 2 examines the relevance of the HEP service packages and

Chapter 2
the delivery modalities applied thus far. The epidemiological relevance

of the HEP is investigated from the perspectives of disease burden,
responsiveness to the needs of the beneficiary population, and the
community’s perception of the role of the HEP in promoting the right
behaviors to prevent and control disease. The section on the relevance
of service delivery modality reports on mechanisms that have been
put in place, including the model family approach and the Women’s
Development Army (WDA), among others. In addition, the modalities
are examined from the point of view of the level of service provision
achieved and harmonization with socio-cultural sensitivities within the
communities. The HEP’s own adaptability to changing circumstances is
also examined from different angles.
Chapter 3 discusses the status of the HEP’s inputs. This chapter

Chapter 3
covers several key inputs to the program, including human resources,

infrastructure, and pharmaceuticals. The human resources aspect
investigates Human Resources for Health (HRH) planning, recruitment,
training, deployment, skills, incentives, career paths, job satisfaction,
attrition, and the level of HEWs’ performance. The infrastructure aspect
of the input examines the availability and status of roads, transportation
systems, the physical infrastructure of the HPs, and basic utilities
and equipment within the facilities. Furthermore, the supply chain
management system has been comprehensively assessed, examining the
level of availability of essential pharmaceuticals required at the HP level,
page- 5
as well as the functionality of the Logistics Management Information
System (LMIS), the level of stockouts, inventory management, and the
patient counseling conducted by HEWs. The status of HEP financing
is also presented in this chapter with a special focus on description of
Total Expenditure & source financing at PHCU, share of HEP from the
PHCU expenditure and THE, and HEP expenditure by type.
Chapter 4 presents the status of health service delivery through the

Chapter 4
HEP. It describes the physical accessibility of HPs and the availability

of HEP services. The chapter also examines the level of awareness
of community members of available HEP services and their exposure
to the program through home visits, HP visits, and outreach services.
Finally, the chapter discusses the level of adoption of HEP-related
behaviors at the household level and factors associated with better
adoption of HEP-related behaviors.
Chapter 5 discusses community engagement and ownership in the

Chapter 5
HEP. The current status of strategies intended to engage community

members is presented. Coverage of the Model Family Training (MFT)
program, the availability and functionality of WDA and SMC structures,
and the characteristics of community health volunteers are examined in
this chapter.
Chapter 6 describes the Health Information System (HIS) for the HEP.
Chapter 6
It presents an assessment of the soundness of the design of the HEP’s

HIS, its simplicity, and its cost, as well as the status of the various
determinants of the availability and use of the HIS within the context
of the HEP. The latter provides a systematic review of and synthesizes
findings on the status of the structure, the inputs for the HIS, and
the data management process, including data quality assurance.
Monitoring and evaluation (M&E) as it relates to the HEP has also
been reviewed and is presented as the last part of this chapter.
Chapter 7 describes the coverage of HEP-related services at the

Chapter 7
population level. Standard indicators of health service coverage

are presented under the categories of hygiene and environmental
sanitation, disease prevention and control, and maternal, newborn, and
child health. For selected services, the chapter also presents the role of
page-6
the HEP as either a source of information or a service provider and the
implications of providing services through the HEP for quality of care.
Chapter 8 examines the outcomes of HEP service delivery. This chapter

synthesizes plausible evidence regarding the contribution of the HEP to
Chapter 8
health outcomes by establishing trends in morbidity, mortality, and health

behavioral indicators. It also presents a systematic review with meta-
analyses on the effectiveness of the HEP in improving health outcomes.
Lastly, the association between the strength of the implementation of
the HEP and household-level health outcomes is presented based on
data from the household survey conducted through this study.
Chapter 9 examines the governance, leadership, and management

of the HEP. This chapter describes the results of the study from the
Chapter 9
perspectives of the policy environment, the planning process, the

decentralized autonomy of the HEWs, manager commitment, the
coordination of HEP activities, and collaboration with other sectors.
Section III of Part 2 synthesizes findings presented in Section 2 to reach

conclusions and recommendations for future improvement actions. The
section presents conclusions and recommendations in 10 sub-sections, each
of which includes conclusions on specific aspects of the HEP, followed by
recommendations on what to maintain, modify, add, and drop.
Part 3 describes methods and results of the assessment on the UHEP

Part 3
followed by conclusions and recommendations. The findings are organized

with specific sections for different aspects of the program, including
relevance, human resources, and implementation.
Part 4 presents standalone reports of in-depth explorations of specific topics

related to the HEP, including the quality of HEWs’ training institutions,
Part 4
attrition among HEWs, the role of the HEP in public health emergency
management (PHEM), and cost-effectiveness analyses (CEA) of HEP
interventions.
page- 7
PART 1
Context and Objectives of the
Assessment
This part presents the context of the Ethiopian health system

in general and describes the features of the HEP followed by a
description of the objectives of the national assessment.
Section 1: Context of the Ethiopian Health System
Section 2: Objectives of the assessment
page-8
1 Context of the Ethiopian Health
SECTION I
System
1.1 Health status

Ethiopia’s major health problems largely remain preventable communicable
diseases (CDs) and nutritional disorders. Despite major progress in improving
the health status of the population over the last couple of decades, Ethiopia’s
population still faces high rates of morbidity and mortality. The maternal
mortality ratio (MMR) stands at 412 per 100 000 live births. This is a decline
from the estimates of 871 in the 2000 Ethiopian Demographic and Health
Survey (EDHS) and 676 in the 2011 EDHS. According to the Global Burden
of Disease (GBD) study for Ethiopia, communicable, maternal, neonatal, and
nutritional diseases (CMNNDs) account for 60% of the disability-adjusted life
years (DALYs) lost.9 Non-communicable diseases (NCDs) and injuries account
for 33% and 7% of the total DALYs lost, respectively.1
Several health indicators have shown improvement during the last 2 decades.
Modern contraceptive use by currently married Ethiopian women has steadily
increased over the last 15 years, jumping from 6% in 2000 to 35% in 2016 and
41% in 2019. The unmet need for family planning (FP) has declined over time,
from 37% in 2000 to 22% in 2016. The 2016 EDHS shows that 62% of women
received antenatal care (ANC) from a skilled provider at least once for their
last birth. In addition, 28% of births were delivered by a skilled provider, while
the majority of births (42%) are still attended by a traditional birth attendant
(TBA) or a nurse or midwife (20%), followed by a doctor (6%), an HEW (2%),
or a health officer (0.4%).9 Among women aged 15 to 49 giving birth, 17%
had a postnatal (PNC) check during the first 2 days after birth. Under-5 child
mortality stands at 58 per 1000 live births. The percentage of children aged
12 to 23 months who are fully vaccinated stood at 39% in 2016. Seven percent
of children under age 5 had symptoms of acute respiratory infection (ARI),
and 12% of children had diarrhea. The 2019 Mini-Demographic and Health
Survey also showed further improvements in several health indicators, including
increases in delivery at a health facility to 48%, in having 4 ANC visits to 43%,
and in the vaccination of children with all basic vaccines to 43%.9,10
page- 9
Context of the Ethiopian Health System
The prevalence of HIV is estimated to be 0.9%, while the incidence of TB is 277

per 100 000 population. As in several other low- and middle-income countries,
NCDs have in recent years become a major public health concern due to
nutritional problems, smoking, alcoholism, and sedentary lifestyles, among
other factors. In Ethiopia, 4% of men smoke any type of tobacco, although the
percentage of men aged 15 to 49 who do not smoke cigarettes has increased
slightly since 2011, from 93% to 95%. In addition, 35% of women and about half
of men (46%) reported drinking alcohol at some point in their lives. Six percent
of women and 9% of men consumed alcoholic drinks almost every day in the
30 days before the 2016 EDHS.9
1.2 Health policy

The Transitional Government of Ethiopia produced the existing Health Policy
in 1993, the first of its kind in the country; this was among a number of major
political and socio-economic transformation measures.11 The implementation of
the health policy was followed by the formulation of a series of four 5-year-
plan-based Health Sector Development Programs (HSDP I, II, III, and IV) in
line with the Plan for Accelerated and Sustained Development to End Poverty
(PASDEP), which is primarily aimed at achieving the health-related MDGs.2-5
One of the major policy initiatives that has been placed in motion as part
of the implementation of HSDP II and HSDP III is the institutionalization of
community/village health service through the HEP, which was introduced for
the first time in HSDP II as part of the Health Service Delivery and Quality of
Care components. HSDP II also clearly indicated that a package of essential
health services focusing on preventive health measures targeting households,
particularly women and mothers at the level of the kebele (cluster of villages),
was to be implemented through the HEP. This paved the way for pilot testing
and the subsequent formulation of standards and manuals to aid the scaling-up
of the program. Subsequently, HSDP III reiterated the importance of the HEP
as a main vehicle for achieving the PASDEP targets and the health MDGs. It
also provided further details on institutionalization as well as the types of HEP:
rural, urban, and pastoralist. Furthermore, a number of policy and investment
initiatives were introduced that contributed not only to the HEP but also to
the holistic strengthening of the health system. These include the Accelerated
page-10
Expansion of Primary Healthcare Unit (PHCU) Services; the implementation of

the Health Human Resource Development Plan, the Essential Health Services
Package, the Child Survival Strategy, and the National Reproductive Health
Strategy, among others. HSDP III, for the first time, costed the national health
plan in 3 different scenarios, all of which prioritized the full scaling-up of the
HEP. Furthermore, HSDP III called development partners to action and followed
through with the establishment of the MDG Performance Fund (now called the
SDG Fund). This fund, along with support through other channels, played a
pivotal role in scaling up the HEP with all its system requirements, marking the
program as a global success story among community health systems.
Following the successful completion of HSDP III, the formulation of HSDP IV

has largely become an expression of the government’s renewed commitment
to the achievement of the MDGs while bringing the issues of quality, equity,
and the career development of HEWs into the policy arena. Both HSDPs
III and IV have also taken into consideration the other “most influential
International commitments,” such as the African Health Strategy 2007-2015,
the Paris Declaration on Aid Harmonization (2005), the Accra Accord on Aid
Effectiveness (2008), and the Abuja Declaration on Healthcare Financing in
Africa.12-15
The completion of the four 5-year-based health planning programs prompted

the development of the next generation of health development planning based
on the general and national socio-economic development plan called “Growth
and Transformation Plan” (GTP), which has guided Ethiopia’s economic and
social development ever since. In line with the GTP, Ethiopia’s health sector
has developed its own Health Sector Transformation Plan (HSTP), which has
set ambitious goals of improving the equity, coverage, and use of essential
health services, including improving the quality of healthcare and enhancing
the implementation capacity of the health sector at all levels. The HSTP has
identified 4 mutually supporting transformation agendas to be accomplished
during the five years period (2015/16 through 2019/20): transformation in the
quality and equity of healthcare, transformation at the woreda (district) level,
progress toward the transformation of compassionate, respectful, and caring
health professionals, and, as a fourth transformation agenda, an information
revolution. The HEP, which largely serves segments of the population that are
mostly underserved by higher levels of the health service delivery tier system,
has been considered an important vehicle for narrowing urban-rural disparities
in health service coverage.6
page- 11
1.3 Organization of the health system

The Ethiopian public health system is built upon a 3-tier healthcare delivery
arrangement. The first level is a Primary Healthcare Unit (PHCU) composed of
health centers (HCs), each serving a population of 15 000 to 25 000, and their
satellite health posts (HPs), which serve a population of 3 000 to 5 000 each.
Primary-level healthcare is mostly provided by PHCUs and Primary Hospitals,
each with a catchment population of 60 000 to 100 000 people. The second
level in the tier is a General Hospital, with a population coverage of 1 to 1.5
million people each, and the third level is a Specialized Hospital, covering
a population of 3.5 to 5 million (Figure 1-1). In terms of administration and
management, the MoH deals mainly with policy matters, partnerships, and the
mobilization of international resources. Regional Health Bureaus (RHB) and
woreda health offices (WorHOs) focus on coordination and operational issues,
while Zonal Health Departments (ZHDs) provide technical support.2
page-12
Figure 1-1. Ethiopian Health Service delivery system (the three-tier system)
1.4 The Health Extension Program

While there is no doubt that the HEP has been initiated and owned by the
GoE, its development has passed through several learning processes. This
ranges from the recognition of many home-grown but scattered experiences to
learning from other countries, including the adoption of relevant international
initiatives. All these in combination have provided enormous input for the
inception, designing, piloting, and initial implementation of the program. Below,
we examine the structure and function of the program, the factors that led
to its introduction, and the major landmarks in its evolution. The description
indicated below mainly reflects the goals of the government as communicated
through policy and strategy documents, while the actual state of the program
is presented in the subsequent sections of this report.
page- 13
1.4.1 Factors leading to the introduction of the Health Extension

Program in Ethiopia
GLOBAL HEALTH SYSTEM FACTORS AND INTERNATIONAL

EXPERIENCES
At the global level, efforts to improve access to PHC have taken many forms
since the Alma-Ata Declaration of 1978, attracting enormous interest and
a global consensus that PHC improvements are a prerequisite to achieving
a minimum standard quality of life worldwide. The Declaration established
the “community health worker” (CHW) as a generic title and recognized this
position as a cornerstone of PHC.8 While this initiative had a mixed record
of successes and failures, its much-documented history in this regard has
focused on its failures, pinpointing its inability to deliver what was expected
from the program in many developing countries for several reasons, including
its inability to sustain the volunteer spirit of CHW. In this regard, the WHO
strongly advised that essential health services cannot be provided by people
working on a voluntary basis if they are to be sustainable.16 The Ethiopian case
of seeking a better mechanism through which to address the basic healthcare
needs of its largely agrarian society was no exception; rather, it was part of
the keenly followed global movement of PHC. The core principles underpinning
the HEP are community ownership and community empowerment to manage
the health problems specific to individual communities, thus enabling them to
produce their own health, which align well with the essence of the Alma-Ata
Declaration.4,8
In addition, learning from the operational history of the Chinese “barefoot

doctors” that worked well during the Cultural Revolution of the 1960s has been
considered seriously for emulation in many developing countries as part of
community-level health intervention toward achieving “Health For All” by the
year 2000. In this program, China deployed thousands of paid “barefoot doctors”
to improve the health of the rural poor. As a result of macroeconomic and
political forces, however, this integrated, CHW-driven PHC approach fell out
of favor in the 1980s and early 1990s. Barefoot doctors did not always work on
a voluntary basis but were also paid as community-based healthcare providers
whose payment came mostly from the surplus produce of the communities with
whom they lived and were assigned to serve.17,18
page-14
Learning from other countries seems to have intensified with the advent of the
South-to-South Collaboration. When the framers of the Ethiopian HEP reached
the critical stage of developing the national guidelines for the HEP, they took
the time to visit the Kerala CHW program with the objective of learning from
its many years of operation at the community level. The Pakistan program
in particular appeared very much closer and more relevant to the Ethiopian
learning process than the Chinese model and may have provided useful insight
in the initial design of the Ethiopian HEP. The Lady Health Workers Program
was initiated by the Federal Government of Pakistan in 1994 with the aim
of providing maternal and child health services at the community level with
workers who belonged to the local communities. These workers must complete
education to a grade 8 standard and be willing to undergo a 15-month training
program; after graduation, they are assigned to work independently, each
covering a catchment area of 200 families (1 000 population), covering more
than half of the population of the country’s rural areas and urban slums.19
POLITICAL FACTORS
The origins and development of the HEP tell a powerful tale, illuminating
some of the drivers of health policy development and factors contributing to
“
its success.
The drivers were historical (emerging out of conflict) as well as

ideological (a grassroots-based and pro-poor orientation) combined
with political imperatives (the need to deliver basic services to a large,
poor population as a new regime) and a healthy dose of pragmatism
(other options were not feasible with the resources available.20)
The quest for an alternative or supplementary structure for the delivery of basic
healthcare to the rural masses in Ethiopia has been seen as a critical political
gesture that was sought largely in response to the failure of previous efforts that
did not adequately address the healthcare needs of Ethiopia’s rural population.
In addition, the government’s 1993 Health Policy clearly states the need to
decentralize the health services and bring them closer to the community, setting
the stage for the exploration of an appropriate vehicle to this end.11
page- 15
This led to the government’s determined decision to launch 2 mutually supportive

programs in 2003: (a) the Accelerated Expansion of PHC Coverage and (b)
the HEP.21 As a flagship program of the government of Ethiopia (GoE),22-24 the
HEP was launched to improve access to basic health services, with a specific
focus on women and children in the 4 largest agrarian regions, and then
expanded to pastoral communities in 2006, and to urban areas in 2009.24 The
seriousness of the HEP’s thinking, inception, development, piloting, and eventual
commitment to a wider and deeper implementation is based in the resolve that
the GOE was genuinely convinced that a community-based approach was the
only realistic and feasible approach to meet the needs of its huge population
with limited resources.20
The Health Extension Program is a flagship of the HSDP involving an

innovative intervention marked by the institutionalization of community health
services, government leadership, and the alignment and substantial support
of development partners. The element of institutionalization was thought to
address the challenge of unsustainability faced by many community programs
by ensuring high political commitment, coordinating national policies, and
leveraging support from partners.22,25
EPIDEMIOLOGICAL AND HEALTH SYSTEM FACTORS
Epidemiological factors, with no doubt, played a substantial role in the GoE’s

decision to establish the HEP. For instance, in the early 2000s, over 300 000
women worldwide, including more than 162 000 in Sub-Saharan African
(SSA) countries, died due to complications of pregnancy and childbirth.26
Ethiopian women represented a large part of this death toll. In addition,
communicable diseases, such as HIV/AIDS, malaria, and tuberculosis, as well
as diseases related to poor hygiene and sanitation, were at the top of the list
of the country’s burden of disease. In addition to the recognition of such an
epidemiological pattern, there was a clear understanding of the health-systems
factors that contributed to such a state of affairs. The prevailing inadequacy of
the number of health workers and community-level health infrastructures was
exacerbated by maldistribution that left those most in need who were least
equipped to handle it: poor and marginalized groups, as well as those living in
rural areas.4,27,28
page-16
LOCAL EXPERIENCES
A positive lesson was also learned from the scattered but village-based small
health delivery units in northern Ethiopia during the armed struggle that
necessitated the provision of basic and essential services. The services delivered
through such health facilities at the community level focused mainly on health
promotion and the prevention and control of communicable diseases, including
the delivery of some basic curative care. These facilities worked well during
times of conflict but did not continue their expected determination and vigor
in northern Ethiopia during peacetime. The search for sustainable community-
based healthcare delivery must continue, then, with strong determination.
More recently, the experience was given expression with the field-level
experimentation and selective implementation of community-based healthcare
delivery in Tigray in the late 1990s. The early inception of the program was
developed under the “Healthy Family” initiative of the Tigray RHB and then the
southern ZHD in 2 villages of Southern Tigray in 1995-96. The program’s earlier
piloting and limited implementation focused on the provision of action-oriented
and consistent health messages in the areas of sanitation and hygiene and
maternal and child health, including immunization and FP, and the prevention
of some infectious diseases, such as tuberculosis.
More importantly, Ethiopia learned a useful lesson from its own practice: despite
its efforts, the community-based health services using voluntary community-
based health agents (CHAs), community-based reproductive health agents
(CBRHAs), and TBAs did not produce any significant changes in the health
conditions of the rural population. This might have been a factor in the
government’s determination to develop an effective, affective, and sustainable
community-based health delivery system. The GoE chose to invest its available
health resources in high-impact, low-cost interventions aimed at addressing the
most pressing health problems of the country leading to the decision to establish
the HEP as a means to realize universal PHC coverage. The government’s
decision to train and deploy HEWs as full-time salaried civil servants has
enabled the health sector to move successfully away from volunteerism.
page- 17
1.4.2 Health Extension Program packages and service delivery

modalities
Efforts to scale up the HEP were launched in 2003, right after the completion
of the pilot program. Several standards for recruiting and training HEWs, as
well as the operational processes of the HEP, were defined during this scaling-
up phase. Accordingly, HEWs were recruited based on nationally-agreed-upon
criteria that include completion of grade 10, residence in the village in which
they will practice, and the ability to speak the language of the community
they serve. All selected HEWs go through a year-long training, which includes
practical training in HCs. The training is provided in the technical/vocational
education centers under the auspices of the Regional State Governments . Upon
graduation, HEWs are assigned to the village from which they come to provide
HEP services. Two HEWs are deployed at each HP, serving a population of
3 000 to 5 000. The HEWs are supposed to select and train “model families.”
These families, after completing the training on the 16 packages of the HEP,
are expected not only to be able to implement these packages but also to
influence their relatives and neighbors to adopt the same practices.
These packages of the HEP are:

• Family Health
o Maternal and child health
o Family planning
o Immunization
o Nutrition
o Adolescent reproductive health
• Disease Prevention and Control
o HIV/AIDS and other sexually transmitted infections (STIs) and
tuberculosis (TB) prevention and control
o Malaria prevention and control
o First aid emergency measures
• Hygiene and Environmental Sanitation
o Safe excreta disposal
o Safe solid and liquid waste disposal
o Water supply and safety measures
o Food hygiene and safety measures
o Healthy home environment
o Control of insects and rodents
o Personal hygiene
• Health Education and Communication
page-18
HPs, the operational hubs of the HEP, are reasonably supplied with the basic
equipment, materials, basic and essential drugs, and medical supplies needed
to deliver the different packages of essential services. The 3 modalities of
service delivery of the HEP are static at the HP level, household visits, and
outreach services, which are well managed by over 40 000 HEWs who are fully
employed civil servants. The delivery of these services is free of charge.29
In recent years, the MoH has introduced “ the second generation of the HEP” and
the Optimization of the HEP (OHEP), which laid down severalinitiatives. The
rationale behind these goals was that changing demographic trends, shifting
epidemiology, and growing urbanization required a more comprehensive and
better quality of a wide range of curative, promotive, and preventive services.
The major components of this reform include upgrading HEWs to level IV,
institutionalizing Women’s Development Army (WDA)structures (networks of
households represented by women), expanding thecontent of existing services
and service packages with the inclusion of someclinical services (which have
grown from 16 to 18), constructing adequate HPs, and improving HEWs’ working
conditions through promotion and transfer. The Optimised HEP in cities and
urban areas suggested the introduction of the Family Health Team (FHT)
approach to effectively address the complex health problems of urban settings.
The FHT comprised Urban Health ExtensionProfessionals, clinicians, public
health professionals, environmental healthprofessionals, and social workers,
among others.
1.4.3 Structure of the Health Extension Program and roles of

various actors
According to the standard guidelines, HEWs are expected to play key roles
in managing the operations of the HPs, conduct home visits and outreach
services, provide referral services to HCs, identify and train model families,
evaluate the implementation of the HEP packages, and submit reports to their
cluster HCs.29 In addition, HEWs are expected to participate in the selection,
formation, establishment, training, and follow-up of WDA leaders.
The HEP envisioned the involvement of key actors, including HEWs, model
households (also called model families), WDAs, other community members,
HPs, HCs, WorHOs, RHBs, and the MoH. In addition, the participation of other
page- 19
sectors, particularly at the woreda and kebele levels, such as the Offices of
Education, Agriculture, Women, and Children’s Affairs, and social or traditional
associations (like Idir, Ekub, and Mahber) has been of paramount importance
in strengthening the implementation of the HEP.25,29 The structure of the HEP
at different levels is described in Figures 1-2 and 1-3.
Figure 1-2. Health Extension Program structure integrated into primary

healthcare units
page-20
Figure 1-3. Administrative structure of the Health Extension Program
page- 21
The HEP is expected to involve women in decision-making processes and

promote community ownership, empowerment, autonomy, and self-reliance as
core principles in its theory of change. Community engagement strategies for
the HEP include empowering women through model family trainings (MFTs)
and engaging community structures, including women. Since 2011, community
engagement strategies for the implementation of the HEP have relied heavily
on mobilizing WDAs. The aim of involving WDA structures was to improve the
implementation capacity of the HEP through the enhanced engagement of the
communities and to identify local challenges and corresponding solutions while
replicating best practices.21,22,29,30
MODEL FAMILIES
During the early stage of the HEP, a model family was defined as a family that
implements a minimum of 75% of the 16 packages after taking at least 75%
of the 96 hours of model family training. These model families are households
that: are trained in HEP-related topics, including maternal health, child health,
malaria prevention and control, and hygiene and environmental sanitation
packages; are able to implement these packages after the training; and are
able to influence their relatives and neighbors to adopt similar practices.
Before the introduction of the WDA, model families were expected to gather
regularly for experience-sharing. They now work as part of the WDA to engage
communities for health improvement. The subsequent revision of the health
extension implementation guidelines modified the model family training hours
to 60 and redefined “model household” as “a family that implemented all health
extension packages concerning its family with the support and close supervision
of a Health Extension Worker.” The expected changes required to be a model
household include making visible changes in behavior: for example, owning
and using a latrine, washing hands properly, eligible mothers’ and children’s
completing immunization schedule, and pregnant mothers’ accessing antenatal
care (ANC).22 A model family mainly implements all the HEP packages as
instructed by HEWs, is involved in other development work, serves as an early
adopter to help diffuse health messages leading to the community’s adoption
of desired practices and behavior, is part of the group network (which contains
6 household members), and has a household head (the mother) who shares her
stories of success in implementing the HEP.
page-22
WOMEN’S DEVELOPMENT ARMY/GROUP (HEALTH DEVELOPMENT

ARMY/GROUP)
The Women’s Development Army (WDA), also known as the Health

Development Army (HDA), is a movement of the community organized through
participatory learning and action meetings. The WDA consists of Women
Development Groups (WDGs), which comprise up to 30 households residing
in the same neighborhood. Each group is further divided into smaller groups
of 6 members, commonly referred as 1-to-5 networks. Leaders of the groups
and the 1-to-5 networks are selected by the respective team members. The
WDA has been providing a platform to engage the community in the planning,
implementation, monitoring, and evaluation of health and other programs in
the country since 2011/12. The HDA is designed to improve the implementation
capacity of the health sector by engaging communities to identify local
challenges and corresponding strategies. It is also designed for scaling up best
practices from one part of the country to another. The formation of the health
development teams and 1-to-5 networks is facilitated by the HCs, HEWs, and
kebele administrations. The key activities of WDA leaders include supporting
HEWs by educating and mobilizing community members and serving as role
models in their health behaviors. They are also expected to participate in the
identification of community needs, planning, and performance review at the
community level.23,25,31
While the main focus of the HEP is at the household or community level, its
implementation requires the coordinated efforts of all actors at all levels. The
unit of intervention is at the household level, followed by different forms and
levels of community engagement (e.g., WDAs/WDGs, HEWs, and community
participation in HP construction). As key vehicles for the implementation of the
HEP, 2 HEWs are deployed to kebeles for each HP serving a 3 000 to 5 000
population.25,32 The following are the key actors in implementing the HEP.
COMMUNITY
With the involvement of the woreda education office, woreda capacity-building

office, and kebele administration, local communities participate in the selection
of HEW candidates, the construction of HPs, and the identification of strengths
and weaknesses, and offer ideas for improvement in the review meetings.
page- 23
KEBELE COUNCIL
The kebele council has a critical role to play in coordinating the overall HEP
activities and in the planning, implementation, monitoring, and evaluation of the
HEP. It also solicits resources for the HEP, mobilizes community organizations
and community members for health action, strengthens community involvement
and participation in decision-making; promotes collaboration across sectors;
manages the HIS, and ensures the availability of HEP commodities (e.g.,
contraceptives, vaccines, anti-malaria drugs, and other medical supplies and
commodities).
PHCUS (HPS AND HCS)
HPs are where HEWs provide ANC, immunization, delivery, growth monitoring,
nutritional advice, FP counseling, and referral services to the general population
of the kebele. HCs provide technical support for the construction of HPs, as well
as referral care and technical and practical support to the HEP at the HPs.
WOREDA ADMINISTRATION AND WOREDA HEALTH OFFICE
WorHOs provide technical support for the construction of HPs and allocate
budgetary and other resources for the implementation of the HEP. WorHOs
also coordinate the implementation of HEP activity by different actors and are
responsible for monitoring and evaluating (M&E) the performance of HEP
activities. They provide technical, administrative, and financial support for the
HEP’s implementation, adapt communication materials, provide supportive
supervision to HEWs and HC and HP management in general, plan and
provide in-service training to HEWs and WorHO staff, obtain reports from HPs
and HCs, and provide information to RHBs and ZHDs.
RHB/ZONAL HEALTH DEPARTMENT
RHBs and ZHDs play a role in adapting implementation guidelines to local

conditions, provide pre-service training for HEWs, ensure the necessary financial
and political support, procure and provide drugs and supplies, provide technical
page-24
and administrative support to WorHOs, translate communication tools and

materials into local languages and distribute them to WorHOs, obtain reports
from WorHOs and provide information to the MoH, mobilize regional resources,
establish referral systems between HPs and HCs, and strengthen the Health
Management Information System (HMIS).
MINISTRY OF HEALTH (MOH)
The MoH is responsible for developing the overall program concept, standards,
and implementation guides, and determining the career structure for HEWs,
with the involvement of the Technical and Vocational Education Training
Schools (TVETs) of the MoE. The MoH is also involved in providing pre-service
training for HEWs, mobilizing national and international resources, providing
communication tools and materials, procuring and providing medical equipment
and supplies, and setting up the HMIS to ensure the necessary financial and
political support is offered.
1.5 Simplified logic model of the Health Extension

Program
The HEP is too large and complex for its full components to be represented in
a program logic model. A simplified logic model of the HEP at the kebele level
is presented in Figure 1-4.
page- 25
Figure 1-4. Simplified program logic model of the Health Extension Program
at the kebele level
page-26
Objectives of the Assessment
SECTION II
Ethiopia has been implementing the HEP as a strategy for reaching its
citizens with essential health services for the last 15 years. Despite significant
improvements in the health status of the population during the last decade,
there are increasing concerns among stakeholders at all levels of the health
sector that the performance of the program has deteriorated in recent years.
The specific remedial actions introduced after evidence from small-scale
studies are unlikely to reverse the program’s trajectory to a satisfactory level of
performance. The purpose of the HEP National Assessment was to generate
the information needed to inform the actions taken toward alleviating program
challenges and to formulate actionable recommendations that could guide the
refinement and implementation of the program during the coming decade.
2.1 General objective
The general objective of the assessment was to assess the status, determinants,
and prospects of the HEP and identify challenges and areas of intervention for
program and policy decisions in the Ethiopian health sector.
2.2 Specific objectives

The specific objectives of the assessment were to:
• assess the relevance of the HEP components to Ethiopians’ health needs;
• assess the implementation status of the HEP;
• assess the population coverage of essential services related to the HEP;
• assess the adequacy of resources needed to implement the HEP;
• explore the contribution of the HEP to recent gains in health status;
• identify the determinants of implementation of the HEP; and
• determine key areas of intervention for the future improvement of the
HEP.
page- 27
PART 2
Rural Health Extension Program
Assessment
This part presents the methods and findings of the national

assessment of the rural HEP (agrarian and pastoralist). It has
three sections:
Section I: Methods of the Rural Health Extension

Program Assessment
Section II: Results of the Rural Health Extension Program

Assessment
Section III: Conclusions and Recommendations
page-28
Methods of the Rural Health Extension
SECTION I
Program Assessment
1 General Approach
1.1 Study area and study period
The assessment was conducted in agrarian, pastoralist, and urban areas of
Ethiopia on a national scale. The study period was October 2018 to September
2019. Primary data were collected from March to May 2019. All 9 regional
states and the 2 city administrations were covered by the assessment. Data
were collected at all levels of the health system, including the MoH, RHBs,
WorHOs, HCs, HPs, and communities.
1.2 Framework and general approach
Figure 1-1. PHCPI framework adapted for assessment of the Health Extension
Program in Ethiopia
page- 29
Methods of the National Assessment of the HEP
The assessment was informed by a conceptual framework that was developed

based on the Primary Healthcare Performance Initiative (PHCPI) framework.33,34
Measures from the framework have been tested in resource-limited settings,
including different African countries.
The framework was prepared in the form of a program logic model showing
the cause-effect relationship among the 5 domains of measurement related to
PHC: (a) system, (b) inputs, (c) service delivery, (d) outputs, and (e) outcomes
(Figure 1-1).
Measuring each of these domains clarified the implementation status of the

HEP as a primary mechanism for the delivery of PHC in Ethiopia, the factors
that limited or facilitated its implementation, and its effectiveness in addressing
the population’s health needs. The framework was adapted to fit priority
information needs that should be targeted by the comprehensive assessment
of the HEP in Ethiopia.
Figure 1-2. Components of the National Assessment of the HEP

The assessment had 3 components: 2 components for evidence generation,
followed by 1 component for evidence synthesis and the formulation of
recommendations (Figure 1-2). The first component of the assessment involved
page-30
the quantitative and qualitative assessment of the current status of the HEP
through the collection and analysis of primary data. The second component
involved a thorough review of the HEP literature (including published and gray
literature) with the purpose of understanding the challenges, facilitators, and
effectiveness of the HEP at different points in time throughout its history. As
part of the third component, the findings from the first 2 sources of evidence
were synthesized, and practical recommendations were formulated to improve
the performance of the HEP. The first 2 evidence-generation components of
the assessment were handled by separate teams of professionals in order to
minimize confirmation bias.
2 Population and Sampling

2.1 Study population
Data for the HEP National Assessment were collected from different levels of
the health system, including from community members and health workers. Study
participants for the quantitative part of this assessment include representative
samples of:
• households selected from study woredas, including women, men, and
youth girls (aged 15-24 years);
• HPs selected from the catchment areas of the study woredas;
• HEWs working in study HPs;
• WDA leaders selected from study HPs; and
• HCs supervising study HPs.
Study participants for the qualitative part include deliberately selected program
experts and HEP coordinators/officers from the MoH, RHBs, WorHOs, HCs,
and partners supporting the implementation of the HEP at different levels.
2.2 Sample size determination
Ethiopia is administratively divided into 9 geographical regions (Tigray, Afar,

Amhara, Oromia, Somali, Benishangul-Gumuz, Gambela, Harari, and the
Southern Nations, Nationalities, and Peoples Region [the SNNPR]) and 2 city
administrations (Addis Ababa and Dire Dawa; Figure 1-3). Agrarian HEPs
are implemented across 7 regions. Pastoralist HEPs are implemented in Afar
page- 31
and Somali and in a few woredas in Oromia and the SNNPR. UHEPs are
implemented in Addis Ababa, Dire Dawa, and urban settings from all other
regions.
Figure 1-3. Study sites of the Assessment
Coverage of HEP-related services among households and service availability

and readiness among HPs were considered in order to calculate the minimum
number of households and HPs to be included in the assessment. The study
focused primarily on measuring the variables required for the estimation of
proportions for a single population. The following formula was used to estimate
the sample size for each of the pre-specified target variable:
ni=z2 Pi (1 – Pi) / D2
Where
ni=the required sample size for the outcome i;
Pi=an estimate of the true population proportion of variable i (sample
size was calculated using different estimates of P for different variables
and taking the variables that yielded the maximum sample size);
Z=1.96 for a 95% confidence level; and
D=a precision level of 5% (0.05).
page-32
2.2.1 Sample size calculation for household survey
The minimum number of households required for the household survey was
determined by using the above formula, which allowed a single proportion to be
estimated with the required degree of precision. Findings from the 2016 EDHS9
relevant to the HEP were used as estimates of population proportions for the
purpose of calculating sample size. Variables used for sample size calculation
were health facility delivery (26.2%), the contraceptive prevalence rate among
married women (35.9%), full immunization coverage among children aged
12 to 23 months (38.5%), the proportion of households with latrine facilities
(39.7%), and the proportion of women with comprehensive knowledge about
HIV (20.2%).
Sample sizes were adjusted for a design effect of 2 and an expected response
rate of 95%. The number of households to be included for each indicator was
then calculated by dividing each sample size by the expected proportion of
households with at least 1 eligible person. The proportion of households with at
least 1 eligible person was determined for each of the 5 indicators. The scenario
that yielded the maximum number of households was then taken as the final
sample size for the household survey. Full vaccination coverage required the
maximum sample size of 6 364 households. The final sample size was, therefore,
considered 6 364 households.
2.2.2 Sample size calculation for health post assessment
The number of HPs required for the HP assessment was calculated using
the sample size calculation formula for the estimation of single population
proportions. The 2016 service availability and readiness assessment35 was used
as a source of estimates for different proportions related to the HEP. Sample
size was calculated with a 95% confidence level and a 5% margin of error
(d=5%) for HP-level variables, including the percentage of HPs with latrine
facilities (60%), the availability of basic equipment among HPs (57%), the
percentage of HPs with a non-zero stock of oral rehydration solution (ORS;
40%), the percentage of HPs providing FP services (95%), and the percentage
of HPs with at least 1 staff member trained to diagnose and treat malaria
(47%). This last percentage (staff trained in malaria) yielded the maximum
sample size of 384 HPs.
page- 33
2.2.3 Sample size of woreda, HC, HEWs and households per kebele
The number of woredas and HCs was determined based on the sample size
of households and HPs. With due consideration to the assumptions behind
different statistical tests and logistics-related issues, the minimum number of
sample households per HP (or per kebele) was determined to be 30 and the
number of HPs per woreda was determined to be 6. This required sampling
of HPs to guide household survey and WDA leaders, which was, therefore,
conducted in the catchment kebeles of 3 HPs from each woreda. HEWs working
in the six study HPs and HCs that supervise these HPs were all eligible for the
study. The HPs were selected from 62 woredas distributed across the 9 regions
disproportionately, which led to the under-sampling of large regions and over-
sampling of smaller regions.
2.3 Sampling strategy
A 3-stage sampling strategy was employed to identify the study woredas, HPs,
HCs, and community members in both agrarian and pastoralist settings. The
woredas in each region were further classified by their predominant means
of livelihood to create 11 strata (7 agrarian and 4 pastoralist). Within each
stratum, a predetermined number of woredas were selected. Six kebeles were
then selected at random from each of these woredas. All HPs providing services
to communities in the selected kebeles were included in the HP assessment,
while the first 3 selected kebeles were included in the household survey.
In the kebeles randomly selected for the household survey, the sampling frame
of households was obtained from the HEWs. A predetermined number of
households (34-38) were randomly selected for the household survey. From
each selected household, the household head, the wife of the head (if the
household head was a man), and a youth girl (aged 15-24 years) were
interviewed. In addition to the households randomly selected to represent the
general population, an independent sample of 4 WDA leaders’ households
was included in the household survey for each kebele. In all 6 kebeles within
page-34
the woreda, a HP was assessed, and in each HP, all HEWs were interviewed to
understand their work type, performance, relationship with the WorHO and the
HC, the community, service delivery issues, job satisfaction, and mental health
status.
2.4 Inclusion and exclusion criteria
Households and residents who had lived in the study kebele for a minimum of 6
months were eligible for the study. Once the households were selected, specific
eligibility criteria were applied for specific data collection modules (Table 1-1).
Table 1-1. Eligibility criteria for study participants
Study Participants/Units Eligibility Criteria

Women Head or wife, age 15+
Men Head or husband, age 15+
Youth girls Age 15-24
HEWs Minimum of 6 months’ experience at HP
WDA leader 6 months’ experience as WDA leader
Key informants At least 2 years in the organization, familiar with the HEP
Abbreviations: HEW, Health Extension Worker; WDA, Women’s Development Army; HEP, Health
Extension Program.
3 Data collection
3.1 Development of data collection tools
Data collection tools were developed for different categories of data collection
activities. These include quantitative data collection tools and qualitative data
collection guides. Quantitative data collection tools included (a) a household
questionnaire with separate modules for different study units (household
module, women module, men module, and youth girls’ module), (b) an HP
assessment tool, (c) an HEW survey questionnaire, and (d) an HC assessment
page- 35
tool. Qualitative data collection guides include key informant interview (KII)
guides and focus group discussion (FGD) guides administered at different
levels of the health system, including at the community level.
All data collection tools were prepared through a process that involved 4 major
steps:
1. Sub-constructs related to the overarching research questions/constructs
of the PHCPI framework were identified through a literature review by
a team of professionals organized for each health system building block
component (governance and leadership, HRH, healthcare financing,
facilities and infrastructure, drugs and medical supplies, service delivery,
and health information).
2. For each sub-construct, the data need was determined at the federal,
regional, woreda, HC, HP, and community levels.
3. For each data need determined in step 2, standard questions were
identified from different sources. For those with no preexisting source
of standard questions, new questions were formulated by the respective
teams.
4. Questions were then arranged into data collection tools and guides
based on their respective data sources.
All survey tools were translated into local languages and then translated back
into English to ensure the accurate translation of each question. The translated
tools were pretested in communities outside of the sample woredas prior to
data collection. Observations from field-level pretesting were used to refine
the data collection tools and procedures. All quantitative data collection tools
were then prepared in the form of an Open Data Kit (ODK) data collection
template with integrated data quality assurance features.
3.2 Data collection, supervision, and

coordination team
Data were collected by a team of trained and experienced data collectors

and supervisors. A team of enumerators, supervisors, and coordinators was
established for data collection in each of the 9 regions. Separate teams were
established for the qualitative and quantitative data-collection activities. The
number of quantitative and qualitative data collection teams in each varied
page-36
according to sample size in each region. Each team comprised 4 to 6 data

collectors and a supervisor. In addition, a coordinator was assigned to each of
the 9 regions.
3.3 Training of data collectors and supervisors
Comprehensive training was provided to all data-collection and supervision

team members. Separate training sessions were prepared for (a) those
conducting the household survey and (b) those performing qualitative and
facility-level data collection. Each training session covered general guidelines
on data-collection methods, sampling and data collection procedures, and the
contents of each data-collection tool. Each training session lasted 7-10 days.
3.4 Data collection fieldwork

HOUSEHOLD IDENTIFICATION
In each region, the allocated number of woredas were randomly selected and
listed for data collectors before their deployment. Once data collectors and
supervisors arrived in each woreda, they sat with their respective WorHOs to
obtain a sampling frame of kebeles. After this list was prepared, supervisors,
along with relevant officials from the WorHOs, randomly selected 6 kebeles for
HP assessment and 3 HPs for the household survey. Similarly, after reaching a
selected kebele, supervisors obtained a sampling frame (i.e., a list of households)
from the HEWs and kebele administrators. Supervisors then used an electronic
random number generator to select 34 households from the sampling frame.
HOUSEHOLD SURVEY
A household survey was conducted through face-to-face interviews with women,

men, and youth girls from selected households. A household questionnaire was
used for the collection of data on household characteristics, exposure to the HEP,
and the coverage of essential HEP-related services. Household characteristics
were assessed for all sample households established at least 6 months before
the data-collection period. The remaining modules were completed based on
the availability of an eligible respondent in the household. Whenever available,
1 man and 1 youth girl were also interviewed from the sample households.
page- 37
SURVEY OF WDA LEADERS
Four WDA leaders (1 to 30 network leaders) were randomly selected from a

complete list of WDA leaders in each kebele selected for the household survey.
The household questionnaire used for the household survey in the general
population was administered to the household of each selected WDA leader.
HP ASSESSMENT
An HP assessment tool, including sections on the availability of inputs,

availability of services, and production of outputs, was used to assess selected
HPs. In addition, the HP assessment included questions on the functionality of
community structures, including WDAs, 1-to-5 networks, and social mobilization
committees in pastoralist settings. The assessment was completed by interviewing
the heads of HPs, making observations in the HPs, and reviewing HP records
(files, family folders [FFs], registers, and reports). The Health Post Assessment
Tool was used to guide the assessment.
SURVEY OF HEWS
A survey of HEWs was conducted through face-to-face interviews with

HEWs. All HEWs in the sampled HPs were included in the survey. The survey
focused on the identification of factors that determine the performance of the
HEP from the perspective of HEWs, as well as the characteristics of HEWs,
including their level of education, knowledge, and satisfaction with their jobs.
Knowledge assessment questions were also administered to assess objectively
how knowledgeable HEWs were. The Health Extension Worker questionnaire
was used for the HEW survey.
HC ASSESSMENT
A quantitative HC assessment tool was prepared to collect data on the role

of HCs in the implementation of the HEP, including HEP financing, supplies,
drugs, and equipment, training on HEP, human resources, the HMIS, and
other related activities. HC heads or their delegates were interviewed for this
assessment. In addition, observations and document reviews were performed to
assess the availability of HMIS tools and service statistics.
page-38
FGDS
WDA leaders participated in focus group discussions regarding 3 major

thematic areas: (a) the role of the WDA in the HEP, (b) characteristics of
HEWs, and (c) community perceptions of HEP-related services. The discussions
were conducted in separate HPs to keep their length manageable. FGD guides
were used to guide FGDs among the WDA leaders. In addition to the WDA
leaders, FGDs were also held with community leaders and with men and women
from the general population to explore external views. Audio from all FGDs
was recorded.
KEY INFORMANT INTERVIEWS
KIIs were conducted at all levels of the health system. The interviewees included
experts with experiences in HEP-related functions at WorHOs, RHBs, and
the MoH. KIIs were also conducted with HC staff and HEWs. Experienced
interviewers administered the KIIs at each level. KII guides were used to guide
the interviews. The audio from each KII was recorded to ensure its accurate and
complete transcription.
FGDS AT MOH AND RHBS
In addition to KIIs with officials at the MoH and RHB levels, FGDs were held
among health promotion, disease prevention, and maternal and child health
program experts who had been using the HEP as a platform to implement their
respective programs. Each FGD was facilitated by 1 facilitator and 1 note-taker.
The audio from all discussions was recorded.
4 Data Management
4.1 Quantitative data management and analyses
Electronic data were exported from ODK to Stata 15 for analysis. Descriptive
statistics were run with regional and livelihood disaggregation to determine
the implementation status of the HEP in each of the 9 regional states. Because
page- 39
of the disproportionate sample size allocation in different regions, aggregate-

level estimates involved appropriate weighting. The weights were calculated by
analyzing the number of possible study units represented by each study unit.
The population size of each region, the number of woredas in each region,
the number of kebeles in each woreda, and the number of households in each
of the selected kebeles were documented during the household survey. This
information was used to calculate the weights used in the analysis of household,
HP, and HEW data to generate national and sub-national estimates. The
primary sampling units (woredas from each of the 11 strata) were assigned
an importance weight equivalent to the relative population size of their strata
distributed to their respective sample size. The weights for kebeles, HPs, and
households were calculated as the inverse of their respective probabilities of
selection. Multivariable regression and logistic regression were used to identify
the factors associated with the intensity of implementation of the HEP and the
outcomes of the HEP at the individual and the household levels.
4.2 Qualitative data management and analyses
Each interviewer recorded answers to the demographic questions on a tablet

and audio-recorded the interview on a digital recorder. Recordings were then
uploaded to a central database, cataloged, and distributed to translators, who
transcribed them in the English language. Transcriptions and translations were
spot-checked for accuracy by a core team member and revised as necessary.
Every effort was made to maintain participants’ confidentiality during data
collection, analysis, and writing. No names were attached to any of the data.
In the Results section, quotations are identified only by source (e.g., KII, FGD),
location, and participant group (community, HEW, or program staff) where
relevant.
Data were entered into NVivo, version 12, and analyzed using thematic content
analysis. A codebook was developed by the analysis team, including researchers
involved in the project design and qualitative coders, and applied to the coding
of all transcripts. The steps involved in codebook development were as follows:
(a) initial codes derived from study goals and instrument questions, (b) codes
adapted and augmented by a reading of 2 transcripts and the conceptual
framework, (c) codes tested on 3 additional transcripts by multiple coders, and
page-40
(d) codebook edited via the addition and deletion of codes as appropriate.
All transcripts were open-coded using the final version of the codebook to
capture the key themes and relevant ideas as identified in the data. Each
transcript was coded by 2 independent coders, and disagreements were resolved
by the lead analyst, who reviewed all discrepancies and discussed them with
the second coder as necessary to reconcile the coding. Once the coding was
complete, code reports were produced for each code, cleaned, and prepared
for synthesis.
Each code report was synthesized by 1 team member as follows: (a) initially,
text excerpts no longer appearing relevant to the code were grayed out, (b) all
remaining excerpts were annotated with comments, (c) particularly illustrative
quotes were highlighted, (d) comments were summarized in theme domains,
and (e) subdomains with associated quotations were put into a table for each
code report.
5 Methods for systematic review of the

Health Extension Program literature
A review of the HEP literature was conducted in 2019 in parallel with the
collection and analysis of cross-sectional data. The review included national
studies like service provision assessment (SPA), service availability and readiness
assessment (SARA), emergency obstetric and newborn care assessment
(EmONC), the EDHS, other published papers, and gray literature. Systematic
review and meta-analysis were used as analytic approaches to synthesize the
evidence from the existing literature.
Different review databases were searched for the presence of a review.

Databases included: the Joanna Briggs Institute Database of Systematic
Reviews and Implementation Reports, the Cochrane Database of Systematic
Reviews, the Campbell Collaboration Library, the National Health Centre
Reviews and Dissemination Databases, and Health Technology Assessment,
and Evidence for Policy and Practice Information. Search terms used were
health extension program, Health Extension Worker, Health Extension Worker
motivation, health extension program actors, health extension program
implementation, health extension program implementation challenges, health
extension program implementation bottlenecks, community health, and
community health programs.
page- 41
A scientific literature search from the AJOL, PubMed, Google Scholar, EMBASE,
Ovid, and Scopus databases was conducted by combining “health extension,”
“community health worker,” “HP,” or “primary healthcare” and “Ethiopia” as search
phrases. In addition, the literature was searched using Research Gate and
cross-referenced using the reference lists of all identified articles for additional
studies that may have been missed in the electronic search. The MeSH terms
were identified using the above index and keywords from the latest MeSH
browser. The search for articles and reports from Google and Google Scholar
were conducted using BOOLEAN terms. Gray literature was searched from all
academic institutions of the country, with a focus on the HEP and PHC, as well
as government documents. All research reports and government reports on 1 or
more aspects of the Ethiopian HEP were included in the review.
Data were extracted using a data-extraction tool adapted for HEP-related

variables. All results were abstracted by 2 reviewers working independently to
avoid extraction errors. The information extracted was synthesized qualitatively,
summarizing the scope of studies conducted on the HEP, the gaps in the
research, and the main findings of the studies in terms of the implementation
status, effectiveness, challenges, and facilitators of the HEP. Meta-analysis
was also considered for the concepts assessed and reported by an adequate
number of studies.
6 Methods of synthesis of evidence and

formulation of recommendations
The quantitative assessment described previously, qualitative exploration, and
systematic review of the literature on the HEP identified several challenges to
the current level of performance of the HEP. These findings were synthesized
through a participatory process involving decision-makers at the regional and
federal levels. A series of consecutive workshops, supplemented by technical
input from the assessment team, were conducted to develop a feasible and cost-
effective set of recommendations based on the scientific evidence generated by
the assessment (Table 1-2).
page-42
Table 1-2. Health Extension Program assessment recommendation formulation

workshops and consultations
Number of
Workshop Participants Activities
sessions
Presentation of findings from
Experts with experience
HEP assessment quantitative, qualitative,
on HEP from MoH,
findings synthesis 1 and document-review teams,
RHBs, universities, and
workshop followed by discussions on
implementing partners
implications
MoH Directors
RHB heads and HEP
leaders Discussion on possible
HEP staff including WorHO solutions to major gaps
Recommendation
4 staff and HEWs identified by the assessment
formulation workshops
Funding agencies following presentation of key
Implementing partners findings of the assessment
supporting the HEP
Study team members
Experts with long-term
experiences on the HEP Alternative recommendations
Recommendation
Participants of the 21st on specific findings were
refinement consultations 3
Annual Review Meeting of discussed and/or debated by
and debates
the health sector participants
Study team members
Researchers used inputs from
Recommendation recommendation formulation
1 Assessment team members
refinement workshops in formulating
final set of recommendations
Abbreviations: HEP, Health Extension Program; MoH, Ministry of Health; RHBs, Regional Health
Bureaus; WorHO, woreda health office; HEW, Health Extension Worker.
7 Limitations of the National Assessment of

the Health Extension Program
This study combined multiple sources of data and information to assess
different aspects of the HEP. This feature of the assessment allowed the several
challenges of using any single method to be addressed. There are still, however,
some limitations that should be considered. Sources of these limitations include:
(a) respondents’ responses relying on recall, (b) universal coverage of the HEP,
leaving no room for a comparative study to determine its impact, and (c) the
possibility of social-desirability bias for the variables assessed using the self-
page- 43
reports of HEWs. These limitations might have under- or over-estimated the

resource adequacy, the availability of services at the HP, the level of exposure
of community members to the HEP, or the contribution of the HEP to health
outcomes.
7.1 Assessment of resource adequacy
Assessment of the adequacy of human resources relied on data reported by the

interviewed HEWs. Responses to questions on the number of staff in each HP
reflect the number of staff on payroll, not the actual number of staff working in
the HP at the time of the visit. This might have resulted in an over-estimation
of the available human resources at HPs. The interpretation of findings on
the number of HP staff should bear in mind absenteeism and reasons for
absenteeism, both of which are reported and assessed in this report.
The competence of HEWs was assessed by asking HEWs whether they had
taken a Certification of Competence (CoC) examination and whether they were
certified. Their responses were not verified. This might have under-estimated the
proportion of HEWs who have ever taken CoC exams and over-estimated the
proportion of HEWs who have been certified. The findings of this assessment
showed that large numbers of HEWs do not pass the CoC exams. Therefore,
this limitation will not affect our conclusion.
Objective methods for assessing the competence of HEWs were limited to

the assessment of their knowledge on a very limited number of topics by
administering exam-type questions. Due to logistical challenges, there was no
objective assessment of HEWs’ skills.
7.2 Assessment of the availability of services at HPs
HEWs’ feeling of responsibility for making services available at HPs might

have resulted in an over-reporting of service availability at HPs. In this study,
we tried to explore the objective evidence of service availability, including
the records of clients who received services during the 1 month preceding the
survey. There is still the possibility, however, of over-estimating the availability
page-44
of services, particularly for those that do not require service delivery data to be
recorded, like the provision of health education at home, at HPs, and during
outreach sessions.
7.3 Measurement of exposure to the HEP

Exposure to the HEP among household members was based on the responses
of women, men, and youth girls from each household regarding their experience
of ever having met an HEW for health information or service, both at any point
and over a 1-year recall period. Estimates of exposure at any point and exposure
during the last year may have led to under-estimates of the true exposure levels
because of a failure to recall. Contact with an HEW that a respondent cannot
remember could be considered insignificant in terms of its ability to influence
household behavior.
7.4 Assessment of the HEP’s contribution
The HEP in general and its specific components, packages, and services have
almost universal coverage across Ethiopia, leaving no room for a controlled
comparative study design to determine its effects. Assessment of the HEP’s
contribution in this national assessment had to rely on a theory-based approach
to synthesize the findings from other national and sub-national studies combined
with interpretation of associations from cross-sectional data.
The use of multiple sources of evidence, including cross-sectional data from a

national scale survey, qualitative exploration from communities and different
levels of the health system, and a review of the HEP literature allowed this study
to make justified conclusions and evidence-based recommendations. Procedures
for the collective interpretation of findings and recommendation-formulation
workshops allowed the research team to offer feasible recommendations that
can address the gaps observed while building on the program’s strengths.
page- 45
SECTION II
Results of the Rural Health Extension
Program Assessment
This section includes nine chapters:
Chapter 1: Sample and background characteristics of

study participants
Chapter 2: Relevance of HEP packages and service

delivery strategies
Chapter 3: Inputs of the Health Extension Program
Chapter 4: Health service delivery through the Health

Extension Program
Chapter 5: Community engagement and ownership in the

Chapter 6: Information system and Monitoring and

Evaluation in the Health Extension Program
Chapter 7: Coverage of Health Extension

Program – Related Services
Chapter 8: Health Extension Program service delivery

outcomes
Chapter 9: Governance, Leadership, and Management of

the Health Extension Program
page-46
1 Sample and Background
CHAPTER 1
Characteristics of Study Participants
The National Assessment of the HEP covered a representative sample of
households, HEWs, HPs, and cluster HCs from 62 (96.9% of the sample woredas)
randomly selected woredas in both agrarian and pastoralist settings. Data were
collected from 6 430 households from the general population, 618 WDA-leader
(1 to 30 network leaders) households, and 343 HPs, response rates of 100%,
81%, and 89%, respectively. At the household level, data were collected from
women, men, and youth girls. At the HP level, data were collected from HPs
and HEWs. In addition, data were collected from 179 HCs and 62 WorHOs
responsible for the supervision of selected HPs. The number of sample HPs in
each region ranged from 17 in Gambela to 74 in Oromia, and the number of
households ranged from 407 in Benishangul-Gumuz to 1 323 in Oromia (Table
1-1 and Table 1-2).
Table 1-1. Number of sample woredas, health posts, health centers, and Health
Extension Workers
Background Woredas Health posts Health centers HEWs

characteristics N (%) N (%) N (%) N (%)
Total 62 100.0 343 100.0 179 100.0 584 100.0
Livelihood
Agrarian 42 67.7 235 68.5 139 77.7 414 70.9
Pastoralist 20 32.3 108 31.5 40 22.3 170 29.1
Region
Tigray 6 9.7 32 9.3 27 15.1 63 10.8
Afar 4 6.5 18 5.2 7 3.9 19 3.3
Amhara 10 16.1 60 17.5 39 21.8 95 16.3
Oromia 13 21.0 74 21.6 46 25.7 123 21.1
Somali 8 12.9 43 12.5 10 5.6 75 12.8
Ben-Gum 4 6.5 24 7.0 6 3.4 37 6.3
SNNPR 10 16.1 59 17.2 35 19.6 96 16.4
Gambela 4 6.5 17 5.0 5 2.8 42 7.2
Harari 3 4.8 16 4.7 4 2.2 34 5.8
Abbreviations: HEW, Health Extension Worker; Ben-Gum, Benishangul-Gumuz; SNNPR, Southern

Nations, Nationalities, and Peoples.
page- 47
page-48
Table 1-2. Number of households, women, men, and youth girls among households from the general population and
from Women’s Development Army/SMC households
Households from the general population WDA/SMC households
Background Households Women Men Youth Households Women Men Youth
characteristics
N % N % N % N % N % N % N N %

Total 6504 100.0 6430 100.0 4416 100.0 900 100.0 618 100.0 613 100.0 389 100.0 120 100.0
Livelihood
Agrarian 4 454 68.5 4421 68.8 3157 71.5 658 73.1 400 64.7 400 65.3 250 64.3 88 73.3
Pastoralist 2 050 31.5 2009 31.2 1259 28.5 242 26.9 218 35.3 213 34.7 139 35.7 32 26.7
Region
Sample and Background Characteristics of Study Participants
Tigray 614 9.4 607 9.4 407 9.2 117 13.0 71 11.5 71 11.6 25 6.4 16 13.3
Afar 412 6.3 399 6.2 275 6.2 61 6.8 37 6.0 35 5.7 24 6.2 7 5.8
Amhara 1 066 16.4 1 060 16.5 603 13.7 169 18.8 71 11.5 71 11.6 33 8.5 16 13.3
Oromia 1 323 20.3 1 319 20.5 1139 25.8 162 18.0 152 24.6 152 24.8 121 31.1 30 25.0
Somali 821 12.6 798 12.4 376 8.5 97 10.8 93 15.0 90 14.7 63 16.2 13 10.8
Ben-Gum 407 6.3 406 6.3 340 7.7 50 5.6 46 7.4 46 7.5 36 9.3 16 13.3
SNNPR 1 023 15.7 1 009 15.7 759 17.2 150 16.7 112 18.1 112 18.3 60 15.4 17 14.2
Gambela 422 6.5 417 6.5 157 3.6 36 4.0 0 0.0 0 0.0 0 0.0 0 0.0
Harari 416 6.4 415 6.5 360 8.2 58 6.4 36 5.8 36 5.9 27 6.9 5 4.2
Abbreviations: SMC, Social Mobilization Committee; Ben-Gum, Benishangul-Gumuz; SNNPR, Southern Nations, Nationalities, and
Peoples.
1.1 Socio-demographic characteristics of household

survey respondents
The majority of female and male respondents were aged 25-44 years. There were
more older men than women. WDA leaders were in general older than women
from households from the general population. The majority of respondents in
both groups of households, both women and men, had no formal education,
while the majority of youth girls from both groups of households did have some
formal education. Currently married women constituted 80.3% of the women
from households and 76.0% of the women from WDA households. WDA leaders’
households were wealthier than households from the general population (Table
1-3).
Table 1-3. Socio-demographic characteristics of women, men, and youth girl

respondents, by type of household
Households from the general

WDA/SMC households
population
Number of respondents youth
women men women men youth girls
girls
6430 4416 900 613 389 120
15-19 4.4 0.5 80.0 1.1 0.3 80.0
20-24 11.9 4.8 20.0 6.2 2.1 20.0
25-29 18.8 12.8 15.5 6.4
30-34 14.7 14.9 13.9 10.0
Age category 35-39 14.4 14.4 20.4 15.2
of respondent 40-44 8.1 14.0 14.8 15.4
45-49 5.5 9.4 9.5 16.7
50-54 8.8 7.5 11.9 11.6
55-59 5.3 5.3 4.7 9.0
60+ 8.3 16.3 2.0 13.4
No formal
74.8 56.0 18.8 68.0 57.6 19.2
education
Educational
status (grades Grade 1-4 11.4 15.4 16.0 14.4 15.9 13.3
attended in Grade 5-8 9.4 17.7 41.0 13.1 14.7 43.3
formal school)
Grade 9 or
4.4 10.8 24.2 4.6 11.8 24.2
above
page- 49
Currently
80.3 98.3 8.3 76.0 99.0 5.0
married
Marital status
Divorced 4.6 0.5 3.9 7.7 0.0 5.0
of respondent Widowed 11.2 0.5 0.3 10.4 0.8 0.0
Separated 3.1 0.2 1.1 5.4 0.0 0.0
Never married 0.8 0.5 86.3 0.5 0.3 90.0
Tigray 9.4 9.2 13.0 11.6 6.4 13.3
Afar 6.2 6.2 6.8 5.7 6.2 5.8
Amhara 16.5 13.7 18.8 11.6 8.5 13.3
Oromia 20.5 25.8 18.0 24.8 31.1 25.0
Region Somali 12.4 8.5 10.8 14.7 16.2 10.8
Ben-Gum 6.3 7.7 5.6 7.5 9.3 13.3
SNNPR 15.7 17.2 16.7 18.3 15.4 14.2
Gambela 6.5 3.6 4.0 0.0 0.0 0.0
Harari 6.5 8.2 6.4 5.9 6.9 4.2
Lowest 20.9 16.0 15.9 12.2 11.3 11.7
Lower 20.1 19.1 17.4 17.5 15.2 13.3
Wealth
Middle 19.9 20.0 21.1 21.4 17.0 17.5
Quintile
Higher 19.8 23.3 21.2 20.6 21.9 26.7
Highest 19.3 21.6 24.3 28.4 34.7 30.8
Abbreviations: WDA, Women’s Development Army; SMC, Social Mobilization Committee; Ben-Gum,
Benishangul-Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples.
1.2 Characteristics of key informants and FGD

participants
Table 1-4 shows the categories of participants and which of the 17 tools were
used for each category. A total of 172 interviews and 109 FGDs were conducted
(MoH=5, partners=6, policy advisors=2, and, at the regional level=25,
WorHO=40, HC heads and supervisors=49, HEWs=38, kebele leaders=13,
male community member FGDs=13, female community member FGDs=29,
and WDA FGDs=52; Table 1-4).
page-50
Table 1-4. Characteristics of key informants and focus group discussion participants
Regions
Gambella
Amhara
Oromia
Somali
Harari
Tigray
SNNP
Total
Afar
BG
WDA/HDA 9 9 9 6 4 6 6 1 2 52
Women community
5 4 5 4 2 3 3 1 2 29
members
Men and community
2 4 3 4 1 2 3 2 1 22
leaders
Kebele
3 2 2 0 0 2 2 1 1 13
administrators
HEWs 7 6 6 6 2 4 4 1 2 38
HC head 3 4 6 3 2 3 3 1 0 25
HEP supervisors 2 6 2 2 3 3 2 2 2 24
WorHO head 1 1 2 1 1 1 1 0 2 10
WorHO HEP
2 1 4 1 2 0 0 3 0 13
coordinators
WorHO process
3 3 0 2 3 1 3 1 1 17
owners
RHB head 1 1 1 1 1 1 1 1 0 8
RHB HEP
1 1 1 1 1 1 1 1 1 9
coordinator
RHB program
1 1 1 1 1 1 1 1 0 8
officers
Total 40 43 42 32 23 28 30 16 14 268
Abbreviations: SMC, Social Mobilization Committee; BG, Benishangul-Gumuz; SNNPR, Southern

Nations, Nationalities, and Peoples; WDA, Women’s Development Army; HDA, Health Development
Army; HC, health center; HEP, Health Extension Program; WorHO, Woreda Health Office; RHB,
Regional Health Bureau.
page- 51
CHAPTER 2
Relevance of
HEP Packages
and Service Delivery
Strategies
page-52
2 Relevance of Health Extension
CHAPTER 2
Program Packages and Service
Delivery Strategies
SUMMARY OF KEY FINDINGS
• Despite substantial declines in MMR and infant mortality rate (IMR)

since the beginning of the HEP, CMNNDs still constitute 60% of the
total DALYs lost in Ethiopia.
• The emerging health burden of the country points to a list of NCDs
(neoplasms, cardio-vascular diseases, chronic respiratory diseases,
cirrhosis, digestive disorders, diabetes, mental illness, and substance use
disorders). Injuries and accidents have been rising in recent years and
already account for 7% of the total DALYs lost; this is emerging as a
major health problem for the country.
• The 16 HEP packages, categorized under the 4 umbrella interventions,
are still relevant to addressing the disease burden of the largely rural
communities of Ethiopia. The packages, however, have not been
adequately delivered for several reasons. One major reason is the limited
availability of services in the HPs.
• Government has made several adaptations of the HEP packages in
response to the communities’ health needs over the HEP’s lifetime. Still,
expansion of the health service packages is unable to meet the growing
needs and expectations of communities, particularly in the provision of
curative health services.
• In addition to the static HP service delivery modality, the HEP service
delivery through HEW home visits and outreach services are found to be
appropriate and are acknowledged by different groups of communities.
• The current arrangement of the HEP service providers (female HEWs
in agrarian HEP and male HEWs in pastoralist HEP) is found to be
appropriate, as evidenced by the responses of a large proportion of
community members. Mixing male and female HEWs in both agrarian
and pastoralist settings, however, has been suggested to address
problems arising from the assignment of either male or female HEWs.
page- 53
Relevance of HEP Packages and Service Delivery Strategies
• Communities perceive that HEWs are, in general, trustworthy and

friendly, considered models of good behavior, and seen as helpful in
empowering women and communities to solve their own health problems.
Trust was limited, however, when more clinical or curative services are
referenced.
INTRODUCTION
In general terms, relevance can be defined as the extent to which the objectives
of an intervention are consistent with beneficiaries’ requirements, country needs,
global priorities and partners’ and donors’ policies. It is 1 of the 5 distinctive
criteria (the other 4 being effectiveness, efficiency, impact, and sustainability)
for evaluating the development interventions that have been developed and
adopted from the OEDC/DAC evaluation guidelines.36,37 Different theories
of knowledge have different implications for what is considered relevant, and
these fundamental views have implications for all other fields, including health,
and particularly in the organization of a health delivery system. Relevance in
the context of the HEP may be interpreted as denoting a holistic program that
is meant to address pressing healthcare needs while meeting the expectations
of the communities to which it is directed.
Chianca (2008) suggests that “in evaluating the relevance of a program or

a project, it is useful to consider the following questions: To what extent are
the objectives of the program still valid? Are the activities and outputs of the
program consistent with the overall goal and the attainment of its objectives?
Are the activities and outputs of the program consistent with the intended
impacts and effects?”36
Relevance has a number of sub-dimensions. This chapter discusses the relevance

of the HEP, looking specifically into the epidemiological relevance of the
HEP packages, the relevance of the HEP service delivery modalities, and the
adaptability of the HEP over time by drawing findings from both qualitative
and quantitative sources.
page-54
2.1 Epidemiological relevance of HEP packages

2.1.1 Disease burden
Regular population-based surveys in Ethiopia have shown that Ethiopia has

made remarkable improvement in several health indicators, including those on
maternal and child health and communicable disease prevention and control.9,10
Despite substantial improvement in health indicators since the beginning of the
HEP, however, the 2017 Burden of Disease Analysis for Ethiopia showed that
CMNNDs still constitute 60% of the total DALYs lost in the country. Moreover,
there is an increasing burden of NCDs, which calls for increasing access to
prevention and control actions. Factors driving most deaths and disabilities are
related to behavioral and environmental factors, including water, sanitation,
and hygiene (WaSH; Figures 2-1, 2-2, and 2-3).1
Figure 2-1. Major causes of premature deaths in Ethiopia, 2007-20171
1 http: //www.healthdata.org/ethiopia
page- 55
Figure 2-2. Major causes of disability in Ethiopia, 2007-20172
Figure 2-3. Risk factors driving deaths and disability in Ethiopia, 2007-20173
2 http: //www.healthdata.org/ethiopia
3 http: //www.healthdata.org/ethiopia et
page-56
2.1.2 Community expectations

Community members involved in the qualitative explorations conducted as
part of the National Assessment of the HEP agreed with the relevance of the
services provided to their respective communities through the HEP. No HEP
package was considered irrelevant to their needs. The most common complaint
about HEP services was their lack of comprehensiveness. Communities have a
high demand for more comprehensive services at the HP level. This demand
has not been adequately addressed for different reasons, including (a) the
decision to keep clinical and curative services separate from health-promotion
and disease-prevention activities, which has led to the limited inclusion of clinical
care at the HP level, (b) the community’s limited trust of HEWs’ ability to
provide curative services, despite a generally high level of trust and acceptance
and reported availability of services at HPs (e.g., Integrated Community Case
Management [iCCM] and Community-Based Newborn Care [CBNC]), and
(c) the low awareness of communities regarding the services already available
at HPs.
Healthcare workers, community volunteers, and both female and male

community members in both agrarian and pastoralist communities listed
CMNNDs, including diarrhea, malaria, TB, typhoid fever, HIV, pneumonia,
and NCDs, including hypertension, as their major health problems. Some key
informants and FGD participants also mentioned cancer, liver disease, eye
diseases, scabies, kidney diseases, typhus, malnutrition, and heart diseases as
priority health problems. The household survey also assessed disease occurrence
at the household level to gain insight into the burden of disease in the 1-month
period prior to the survey. About 1 in 5 households reported that at least 1 of
their members had been sick in the past month, with the highest incidence in
Gambela (32%) and the lowest in Afar (11%). Disease occurrence in the last
month was higher in agrarian than pastoralist HEPs (Figure 2-4).
page- 57
Figure 2-4. Households with at least one sick member in the past 1 month
Abbreviations: HEP, Health Extension Program; SNNPR, Southern Nations, Nationalities, and Peoples.
The assessment also asked about the perceived cause of the most recent sickness
among household members. Chronic illnesses of unknown cause and malaria
were the most commonly reported illnesses in agrarian and pastoralist settings,
respectively. The common cold and diarrhea were among the most commonly
reported illnesses in both settings (Figure 2-5).
Figure 2-5. Reported causes of most recent illnesses among households with at
least 1 sick member during the month period preceding the survey
page-58
2.1.3 HEP packages in response to the health needs of the communities
Responding to the diseases accounting for 60% to 80% of the DALYs lost
in Ethiopia, the HEP service packages are largely focused on alleviating the
pressing health needs of rural communities. The packages were designed
based on the guiding principles of universal coverage of PHC for underserved
communities by bringing services closer to rural communities that used to be
marginalized under the former urban-based health service delivery arrangement.
The 4 essential health service components of the 16 health service packages
have been considered appropriate for tackling the healthcare problems of rural
communities since the HEP’s founding in 2003.4 The delivery of these 16 health
service packages has focused on preventive, promotive, and basic curative
services in addressing the major and critical health problems of families in a
community, especially in rural Ethiopia.22,29
According to the findings of this assessment (which are presented in the

section on coverage of HEP-related services), no HEP package has been
adequately implemented to the level that the community can sustain the
behaviors they have adopted. All HEP packages are relevant in addressing
the major causes of morbidity and mortality among rural communities, both
in the past and currently. Recently added packages on NCDs and mental
health have created opportunities to address the increasing burden of NCDs.
HEWs are providing clinical services, including ANC, other maternal health
services, and the treatment of sick children and newborns through iCCM and
CBNC. Expanding these services to the HP level has been largely considered
an appropriate action by health workers, health managers, voluntary CHWs,
and community members, with the most common justification being the need to
minimize barriers to service use, like distance, lack of transportation, and other
physical barriers. Strategic documents have also considered the HEP the major
community-based service delivery platform for the provision of PHC services.2,3,6
Community members’ perceptions about the HEP’s relevance in the prevention
and control of common diseases is also high (Table 2-1).
page- 59
Table 2-1. Perceived epidemiological relevance of the HEP
Proportion of respondents who agreed that:

HEP increased HEWs have brought
Total number of
awareness of good health habit in
Household members
communities on disease the
interviewed
prevention and control community
Youth Youth Youth
Women Men* Women Men Women Men
girls* girls girls
n n n % % % % % %
National 6,430 4 805 1 020 82.9 82.1 72.5 80.8 79.8 75.2
Region
Tigray 607 432 133 86.7 84.9 71.2 86.4 83.5 74.5
Afar 399 299 68 69.6 65.9 79.0 68.3 66.6 75.5
Amhara 1 060 636 185 91.1 87.8 83.0 90.1 87.5 86.9
Oromia 1 319 1 260 192 79.4 80.4 63.6 76.9 77.1 66.6
Somali 798 439 110 47.2 61.1 69.4 47.9 61.3 63.2
Benishangul-
406 376 66 93.3 90.3 61.0 93.0 87.4 96.9
Gumuz
SNNPR 1 009 819 167 83.0 82.2 77.3 80.1 80.1 77.9
Gambela 417 157 36 69.2 71.4 43.5 69.3 71.3 44.3
Harari 415 387 63 56.1 53.9 32.9 58.5 50.7 49.6
Livelihood
Agrarian 4 420 3 407 746 83.9 82.7 72.6 81.9 80.4 75.3
Pastoralist 2 010 1 398 274 58.7 64.8 70.0 56.5 63.5 70.4
*Number includes WDA households
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HEP, Health Extension Program;
HEW, Health Extension Worker.
Household members were also asked whether they wanted more services from
HPs in their respective kebeles. The recommendation for additional services was
almost universally made by the women, men, and youth girls who had ever visited
an HP. Although the communities cited “additional” services in their perceptions,
these services might actually already be part of the current packages. Among
study households whose members had ever visited an HP, 43.9% of women,
51.5% of men, and 49.3% of youth girls suggested that additional services be
provided there (Table 2-2). The treatment of sick adults, the treatment of sick
children, and health facility delivery services were the top 3 services women
recommended as additional HP-level services. Allowing women to have all their
ANC and PNC services at HPs was another recommendation made by women,
men, and youth girls alike (Figure 2-6).
page-60
Table 2-2. Community members recommending the addition of HP services to

current offerings, by region and livelihood
Among households whose primary respondent visited health post,

percent who recommended provision of additional services at HPs
Women Men* Youth girls*
National 43.9 51.5 49.3
Region
Tigray 74.5 83.2 62.9
Afar 44.3 39.1 29.0
Amhara 42.4 58.1 46.8
Oromia 41.3 48.5 46.1
Somali 18.0 23.5 13.8
Benishangul-Gumuz 72.3 83.3 67.2
SNNPR 44.6 43.4 51.3
Gambela 43.4 43.5 44.4
Harari 58.4 74.5 59.6
Livelihood
Agrarian 44.0 52.0 50.0
Pastoralist 39.8 35.0 26.7
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HP, health post.
page- 61
Figure 2-6. Percentage of household members who recommended additional

services at HPs in response to their health needs
Abbreviations: HP, health post; ANC, antenatal care.
2.2 Relevance of HEP service delivery modalities and

approaches
2.2.1 Acceptability of home visit as HEP service delivery modality
Household members were asked whether they agreed or disagreed with the
appropriateness of HEWs’ providing services at the household level. Large proportions
of women and men (82% each) and 77% of youth girls agreed that household-level
service provision by HEWs would be appropriate. In general, the acceptability of
a home visit as a modality of HEP service delivery was higher among agrarian
communities than among pastoralist communities. The acceptability of a home visit
was 82.4% in agrarian communities and 74.7% in pastoralist communities among
women, 82.0% in agrarian communities and 72.0% in pastoralist communities among
men, and 76.7% in agrarian communities and 68.5% in pastoralist settings among
youth girls. There was also marked regional variation in the acceptability of home
visits, with the highest level of acceptability observed in Benishangul-Gumuz (above
95%) and Amhara (above 86%) and the lowest in Harari (below 59%) and Gambela
(below 68%; Table 2-3).
page-62
Table 2-3. Percentage of household members who agreed on the appropriateness of

home visits as an HEP service delivery modality, by region and livelihood
Proportion of household members who

Unweighted number of respondents
consider home visit as HEP service
interviewed
delivery modality is appropriate
Women Men Youth girls Women Men* Youth girls*
National 82.2 81.7 76.6 6 430 4 805 1 020
Region
Tigray 77.4 74.3 65.3 607 432 133
Afar 67.7 62.1 67.6 399 299 68
Amhara 90.6 88.3 86.6 1 060 636 185
Oromia 79.9 80.6 68.8 1 319 1 260 192
Somali 47.9 68.7 59.2 798 439 110
Benishangul-
96.3 95.1 97.0 406 376 66
Gumuz
SNNPR 81.0 81.2 83.5 1 009 819 167
Gambela 67.7 66.9 52.8 417 157 36
Harari 58.8 55.2 54.0 415 387 63
Livelihood
Agrarian 82.4 82.0 76.7 4 420 3 407 746
Pastoralist 74.7 72.0 68.5 2 010 1 398 274
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HEP, Health Extension Program.
Findings from KIIs and FGDs with community members and program personnel
also showed that HEP service delivery at the household level through home
visits is critical for families in rural areas. Home visits allow HEWs to access
community members who are skeptical toward modern health services and
would never visit a health facility on their own. According to WDA leaders and
both female and male community members, home visits allow HEWs to observe
and facilitate improvements in the living conditions of rural families. The HEP-
related services provided through home visits as reported by FGD participants
include registering malnourished children and providing supplementary food,
screening for TB, checking the availability and cleanliness of latrines; vaccinating
children and following up on their status, providing contraceptives, assessing
the availability of waste-disposal pits and performing household inspections,
page- 63
assessing whether animal barns are separated from rooms where people
live, and even checking the cleanliness of the sleeping rooms, all of which are
considered relevant to the needs of rural families. Community members and
program staff asserted the usefulness of HEWs’ home visits in connecting health
professionals to the underserved community. Still, some service provision at the
HP level is challenged by the limited available time, as HEWs, particularly
those who are the only HEW at their HPs, spend a significant portion of their
time on home visits.
The reactions of community members to the recently declining trend in the

frequency of HEWs’ home visits also speaks to the acceptability and usefulness
of home visits in the provision of HEP services for rural families. One statement
“
by a female community member in Dire Tiyara, Harari is an example worth
mentioning:
The former HEWs visited all the households in the kebele. But the
current HEWs do not do that. They are not providing this service….
During the last 3 years, there were absolutely no house-to-house visits.
HEWs who were working in our kebele 3 years ago visited every
household in the kebele, they inspected our personal hygiene, they
advised us on several issues relating to healthy behaviors.…But now,
they come to the community only when they are forced by higher
officials to do so.
FGD, Female Community Members, Harari
Men also shared similar views on the relevance of home visits in the
implementation of HEP. FGD participants described home visits as helping
them save time and money to get treatment and improving their access to
health services. One male community member FGD participant from Bahir
“
Dar, Amhara indicated that
HEWs serve us like a family member; they deliver services house to

house, which minimizes our challenges related to time and costs.…
Almost all diseases are covered by HEWs. They prevent abortion
by providing family planning.…For children, HEWs teach mothers to
clean their breasts before breastfeeding…to make the water safe by
adding chlorine before drinking from taps.…
FGD, Male Community Members, Amhar
page-64
2.2.2 Socio-cultural relevance of service delivery modalities
Agrarian HEWs, all of whom are women, are selected from the communities in which
they live in order to address the linguistic, social, and cultural barriers and therefore
increase their community acceptance in service provision38,39 Female HEWs have
been recruited to manage the HEP because they are culturally more acceptable
than males for family health-related interactions in the agrarian setting. Male HEWs
are deployed in the pastoralist areas, given the relevant cultural and environmental
factors.22
In their service delivery, HEWs conduct outreach activities by going from house to
house. During these visits, they are expected to teach by example (e.g., by helping
mothers care for newborns, cook nutritious meals, construct latrines, and dispose of
waste in pits). This makes them socially and culturally acceptable to all groups in the
community and able to communicate directly with women about their health, their
children, and their families. The role of WDAs is also something to capitalize upon in
the socio-cultural aspect of the HEP service delivery. As Wang et al. explained, WDAs,
together with HEWs and trained model families, form key community-based actors
who provide socially and culturally plausible HEP services to the communities.25
The gender of HEWs is a critical socio-cultural element of the HEP. Household

members were asked 2 independent questions about the acceptability of a female
or male HEW for the provision of HEP services. A large majority of women, men,
and youth girls (88-89%) approved of female HEWs. The approval of male HEWs is
relatively lower (71.4-75.3%); it still shows, however, that there is a potential for male
HEWs (Table 2-4).
page- 65
Table 2-4. Percentage of household members who agreed with statement about
HEWs’ gender, by region and livelihood
Proportion of respondents who agreed that:

It is good that HEWs are females It is good that HEWs are males
Women Men Youth girls Women Men Youth girls
n % n % n % n % n % n %
National 6 164 89.4 4 647 89.3 949 88.2 6 101 71.4 4 610 75.3 936 71.8
Region
Tigray 578 91.4 410 89.4 116 84.5 543 77.6 386 76.0 112 66.3
Afar 371 77.3 286 76.2 67 85.9 378 65.5 280 64.3 65 75.0
Amhara 1 001 93.1 600 92.8 166 87.6 978 70.9 597 76.1 163 71.5
Oromia 1 285 89.7 1 233 89.8 179 86.5 1 277 70.7 1 225 73.1 176 73.5
Somali 730 49.8 410 68.2 95 63.6 734 50.7 410 67.7 96 62.3
Benishangul-
398 95.7 373 97.4 65 93.8 401 67.7 376 70.7 65 42.0
Gumuz
SNNPR 996 87.6 802 86.5 162 94.5 995 74.4 802 80.4 161 71.8
Gambela 392 73.6 148 78.2 36 72.6 383 66.1 150 71.2 35 63.7
Harari 413 75.2 385 71.1 63 62.2 412 54.4 384 52.5 63 30.3
Livelihood
Agrarian 4 268 90.3 3 305 89.7 694 88.6 4 200 71.9 3 277 75.4 684 72.1
Pastoralist 1 896 69.8 1 342 78.7 255 71.8 1 901 59.7 1 333 70.4 252 62.1
Abbreviations: HEW, Health Extension Worker; SNNPR, Southern Nations, Nationalities, and Peoples.
Gender preference was also assessed in the qualitative part of the study. As
justified in the HEP design, the qualitative study participants preferred female
HEWs to deliver HEP services at the HP and home visits. This is because
females are more culturally, socially, and biologically linked to maternal and
child health issues than are males. Both male and female beneficiaries prefer
female HEWs. Women prefer female HEWs in particular for delivery services.
The following statement from a WDA leader describes this preference:
page-66
“
In my understanding, males care more than females and are more
sympathetic than females, and even with their sympathy, we don’t
dare to be delivered by males, for we feel shame to show our private
parts.
WDA leader, Tigray
On the other hand, there were some situations in which male HEWs were
preferred. First, a long-standing cultural belief in Ethiopia is that males are
stronger and better qualified than females. Over the last 2 decades, however,
gender has become a persistent multisectoral issue among government and
private partners that has led to tremendous changes in people’s views. Still,
due to strong cultural issues and environmental factors, pastoralist communities
prefer male HEWs. Additional reasons raised in FGDs and KIIs for suggesting
male HEWs were related to physical challenges and security concerns.
Another line of thinking among the community (particularly among those with
more education) is that gender is not important; what matters in HEWs is their
education, competence, and commitment. The following statement from a man
“
in Malie Senata woreda in the SNNPR supports this statement:
There is no problem whether the HEW is male or female as long as

they have the necessary education and skill.
FGD, Male Community Members, SNNPR
The final relevant theme relating to gender preferences for HEWs is having
HEWs of both genders to fill gaps and buffer issues raised in terms of culture,
topography, and the gender of service users, and to minimize sexual violence
and abuse, particularly in remote areas.
page- 67
2.2.3 Trust and friendliness

The HEP assessment explored household members’ perceptions of HEWs’
friendliness and respect toward the communities and communities’ trust of
HEWs. Agrarian household members in general and women in particular
agreed that HEWs have friendliness and respect for the communities and that
the communities trust the HEWs (Figure 2-7, Table 2-5).
Figure 2-7. Trust and friendliness of HEWs
page-68
Table 2-5. Perceived trustworthiness and friendliness of HEWs
Total number The community
HEWs are The community HEWs serve us HEWs are ready
of household has high regard HEWs serve us
caring for the trust that HEWs with a friendly to learn from the
members and recognition with respect
community are growing approach community
interviewed for HEWs
girls
girls
girls
girls
girls
girls
girls
Men
Men
Men
Men
Men
Men
Men*
Youth
Youth
Youth
Youth
Youth
Youth
Youth*
Women
Women
Women
Women
Women
Women
Women
n n n % % % % % % % % % % % % % % % % % %
National 6 430 4 805 1 020 69.6 67.9 55.1 74.7 75.5 71.6 71.3 69.1 59.1 74.9 73.8 65.9 71.5 69.6 65.3 69.8 66.5 62.1
Region
Tigray 607 432 133 82.1 78.1 54.7 75.8 68.5 77.4 71.8 60.3 88.9 85.9 74.4 87.8 85.3 72.0 77.9 77.7 65.1
79.5
Afar 399 299 68 65.5 63.2 67.7 61.7 68.8 70.4 63.7 74.3 71.2 70.7 79.1 73.4 68.6 77.2 66.7 65.6 75.3
68.2
Amhara 1 060 636 185 83.6 80.5 70.9 86.1 83.7 79.6 83.3 80.3 75.5 87.1 84.2 74.6 85.1 83.7 76.5 83.5 77.6 73.3
Oromia 1 319 1 260 192 63.5 62.6 51.5 61.9 64.5 63.4 50.3 70.0 69.6 56.6 64.4 63.0 57.1 65.8 61.6 57.8
67.5 70.8
Somali 798 439 110 47.0 62.3 59.0 57.8 46.1 63.4 60.1 47.3 62.0 56.8 46.4 62.1 60.4 46.7 60.3 61.5
46.2 64.0
Benishangul-
406 376 66 85.1 86.2 53.9 92.3 92.3 89.2 88.1 87.8 48.2 90.8 91.8 77.9 89.4 87.1 77.4 87.6 85.7 84.8
Gumuz
SNNPR 1 009 819 167 65.3 68.2 42.9 78.2 80.3 82.1 72.4 73.1 55.6 69.7 73.1 71.1 68.6 70.8 65.5 61.8 66.4 55.9
Gambela 417 157 36 68.1 63.6 43.9 67.1 65.1 46.3 66.6 65.7 45.1 69.8 75.3 45.9 69.0 72.9 49.4 64.9 64.7 49.4
Harari 415 387 63 50.4 45.6 40.2 47.0 50.4 49.6 42.4 55.9 48.6 50.2 54.4 48.9 41.9 52.7 46.0 40.3
57.5 54.0
Livelihood
Agrarian 4 420 3 407 746 70.5 68.4 54.9 76.1 71.9 72.2 69.7 58.9 75.9 74.4 66.0 72.4 70.1 65.3 70.8 66.9 62.1
75.6
Pastoralist 2 010 1 398 274 49.4 52.4 61.5 54.1 63.1 50.1 54.7 64.2 51.8 55.8 63.0 51.1 55.1 62.7 49.1 53.7 63.4
58.6

Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HEW, Health Extension Worker; HEP, Health Extension Program.
page- 69

The qualitative results reinforce these quantitative findings. Communities’ trust

of HEWs has improved over time due to the multifaceted roles the HEWs are
playing, as well as their friendly approach. A statement from a male community
“
member in a Tigray woreda supports this:
The beneficiaries strongly believe in the Health Extension Workers.

The community believes that they are lifesavers.
FGD, Male Community Members, Tigray
Furthermore, WDA leaders also indicated that there is strong community trust
toward HEWs, with some reporting that women are naming their children
after the HEWs who assisted in their delivery. An impressive reflection from an
agrarian woreda in the SNNPR confirms this:
“
I gave birth to twins and came to the HP to get service because my
babies were ill and crying. At that time, they really supported me and
provided the necessary care for my babies. I gave my twins the names
of the HEWs, Tsedu and Aynalem, because of the trust I had in them.
FGD, WDA Leaders, SNNPR
HEWs themselves believed the communities placed trust in them. Communities

tell them their problems openly, as evidenced by an HEW from an HP in
“
Amhara:
They do trust us because we go homes to identify problems, and if

problems persist, they are honest when they explain to the Health
Extension professionals. If they didn’t trust us, they wouldn’t have told
us everything.
HEW, Amhara
In contrast to these positive reflections, some community members noted a

decline in communities’ trust of HEWs. Some female community members cited
page-70
HEWs’ lack of skills, drugs, equipment, and supplies at the HPs as reasons for
“
this declining trust. A female participant from Raya Kobo in Amhara confirmed
this perception:
The fact that HEWs are short of the competence required to conduct
some medical services means that our population doesn’t trust them.
FGD, Female Community Members, Amhara
2.2.4 HEP acceptance by the communities
For a general perspective, we asked household members whether the communities

accepted the HEP’s service delivery. Over 70% of agrarian women and men
nationally agreed that their communities trust that the HEWs model good
behavior. Similarly, over two thirds of female and male agrarian household
members nationally agreed that the HEP empowers communities to solve their
health problems themselves. Overall, a great majority of women and men, both
at the national level and in agrarian communities, confirmed that the HEP is
accepted by the communities (Figure 2-8, Table 2-6).
Figure 2-8. Perceived acceptability of the HEP

Abbreviations: HEP, Health Extension Program; HEW, Health Extension Worker.
page- 71
The qualitative results also consistently showed that women, men, and
community members in general have accepted the HEP through the HEWs’
education, communication, and services provided either in the HP or during
home visits. Women’s use of services like ANC, institutional delivery, PNC, and
FP methods (including by the wives of religious leaders), the use of insecticide-
treated nets (ITNs) in malaria-prone areas, improvements in personal and
environmental hygiene, the construction and use of latrines, a reduction in
female genital mutilation (FGM), treatment-seeking for sick children and
adults, and the reduction of early marriage and its associated consequences
were some of the key indications of communities’ acceptance of the HEP. In
addition, communities and program respondents agreed that men’s involvement
in supporting household members’ use of HEP services has improved over time.
FGD and KII participants mentioned that previously, women had hidden their
children during HEWs’ home visits, a sign of mistrust. In response, however,
to HEWs’ continuous education and demonstration of HEP services, the
communities accepted and recognized HEWs.
Some community participants in agrarian areas, however, disclosed that some

communities still have not accepted the HEP, as demonstrated by their open
defecation; non-burial of dead animals, non-acceptance of HEWs’ environmental
sanitation campaigns, and the continued practices of FGM and early marriage.
page-72
Table 2-6. Household members’ perceived acceptability of the HEP
Percentage of respondents who said:
Total number of I think that HEWs are a The HEP empowers The HEP has good
household members model for the community in communities to solve their acceptance by the
interviewed their behavior health problems themselves community
Youth* Youth Youth Youth

Women Men* Women Men Women Men Women Men
girls girls girls girls
National 6 430 4 805 1 020 74.9 72.6 59.2 67.2 66.1 52.6 86.3 86.8 78.7
Region
Tigray 607 432 133 82.8 81.7 62.6 82.6 78.9 51.1 87.0 86.0 70.1
Afar 399 299 68 68.2 63.9 75.1 71.0 67.8 76.1 70.5 68.9 73.1
Amhara 1 060 636 185 84.2 82.0 76.0 80.5 78.2 70.7 91.4 89.0 86.9
Oromia 1 319 1 260 192 72.4 69.4 55.0 63.5 63.9 50.3 84.2 84.9 72.0
Somali 798 439 110 45.8 60.5 57.0 45.9 61.3 57.7 48.3 65.4 63.5
Benishangul-
406 376 66 87.5 84.5 62.0 75.4 71.7 48.7 95.7 93.9 73.0
Gumuz
SNNPR 1 009 819 167 70.0 71.4 46.4 57.2 59.3 36.7 88.2 90.7 85.4
Gambela 417 157 36 65.0 64.8 44.2 66.7 69.2 49.8 74.8 79.5 59.9
Harari 415 387 63 56.0 51.3 39.6 51.1 46.2 41.6 62.4 56.5 50.5
Livelihood
Agrarian 4 420 3 407 746 75.9 73.1 59.2 68.0 66.6 52.4 87.3 87.3 78.9
Pastoralist 2 010 1 398 274 52.1 58.5 61.9 48.3 54.6 58.6 63.3 71.5 70.1
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HEP, Health Extension Program; HEW, Health Extension Worker.
page- 73

2.2.5 WDA and model family approaches
Some FGD participants (both women community members and WDAs)

confirmed that WDAs had contributed hugely to the mobilization of community
members in the 1-to-5 networks and the identification and referral of pregnant
women for services. They have also played a role in the remarkable reduction
in harmful traditional practices in their communities. WDAs share the burdens
of HEWs in health education and promotion and in the identification and
referral of sick women and children to HEWs. Two statements from a woman
community member (FGD) from Semen Mecha woreda, Amhara, and a WDA
“
from Kokossa kebele, Oromia, are good examples of this:
WDAs mobilize the community for immunization….WDAs already know

who pregnant women are, as they live together. They tell pregnant
women to go to a facility for ANC visits and remind them during
the last trimester to access a maternity waiting home in the HCs for
safe delivery. WDAs also tell them to get their babies vaccinated.
They teach them to share responsibilities with their husbands, because
a job burden could lead to preterm delivery, since it will lead to
intensive care for a preterm baby. So, this is how WDAs assist HEWs.
“
I have been working as a WDA leader for the last eight years. I
have been working for the community on norms that affect women
through harmful traditional practices, such as genital mutilation, early
marriage, polygamy, etc. As a result, I have brought many changes
and improvements to this kebele.
FGD, WDA Leaders, Oromia
On the other hand, some FGD participants mentioned that community

members are forced to accept HEP-related education and practices against
their will or before they are convinced. If they do not accept the teachings of
the HEP, local government appointees and WDAs punish them with fines and
intimidation. This approach to achieve behavior change through punishment,
not by diffusion, was reflected in statements made by FGD participants, a
WDA leader from Kokossa kebele, in Oromia, and a community member in
Atsbi Wenberta woreda, in Tigray:
page-74
“
I have served as a WDA leader for the last 6 years. Before we gained
awareness from HEWs, all mothers had been giving birth at home.
Now, we are following pregnant women within the 1-to-5 networks so
that they give birth in the facilities…. However, punishing a mother
who does not give birth in the facilities has become a problem for
us [WDAs]; we are also punished if we do not follow them and send
them to facilities.
“
FGD, WDA Leaders, Oromia
The HEW teaches us house to house. We [WDAs] also visit our

community house to house. We will be punished if we do not visit the
community. We pay 100 birr as a punishment if we do not inform the
HEW and let a pregnant mother deliver in a health facility. HEWs
are also liable for punishment if they do not help pregnant mothers to
deliver at the health facilities. So, if we find pregnant mothers during
our visits to homes, we report them to the HEW. The HEW visits the
pregnant mother we identified, and she refers or takes her to a health
facility for ANC visits and then for delivery.
FGD, Female Community Members, Tigray
2.3 Adaptability of HEP packages and service

delivery modalities
As indicated in previous sub-sections, the 16 HEP service packages were
carefully designed under the 4 major constructs during the early stages of the
agrarian HEP in 2003 and adapted for pastoralist communities in 2006. In
2009, the HEP packages were also modified for the urban setting, with a focus
on chronic health problems and environmental sanitation.2,4,22,25,29 In 2016, the
second-generation HEP also accommodated these packages, including NCDs
and other priority communicable and neglected tropical diseases in response
to the current disease burden and health profile of the country.40 The MoH
had undertaken an assessment of the HEP in which, as part of this process,
the HEP’s optimization efforts in family health services, particularly iCCM and
page- 75
CBNC, were among the myriad initiatives showing the government’s strong
commitment to and continued adaptation of the HEP. The assessment also
reported a gap in some of the HEWs’ practical skills and a growing demand
for curative services. In response, the MoH provided Integrated Refresher
Training (IRT) and upgraded HEWs from Level III to IV. In addition, the
provision of some curative and preventive services, such as the treatment of
pneumonia, iCCM, and the insertion of Implanon began at the HP level. As
indicated in previous chapters, in 2011, the WDA was initiated, bringing with it
the opportunity to strengthen community engagement and improve utilization
of HEP services in subsequent years.41,42
Several adjustments were also made to the modality of delivery of services,

a notable one of which was the proportion of HEWs’ time spent at the HPs
and in-home visits. As indicated, HEWs were at the beginning expected to
spend 75% of their working time conducting home visits and outreach activities
and the remaining 25% at their HP providing basic curative, promotive, and
preventive services. This time allocation for community services, however, has
been reduced to 50% as a result of several associated factors. One reason
could be the addition of selected curative services.25,41
Community participants from the qualitative assessments suggested a

few areas where HEP services need to expand. A notable one is the full
treatment of sick children and adults. Findings indicated that people expressed
disappointment in being unable to receive services from higher-level facilities
when they are referred, despite having paid their insurance premiums. Other
areas of change desired include better HP room design, the tailoring of health
education to clients, and an improvement in the quality of services, including
training in clinical care and the referral system. iCCM was mentioned as one
area that needs modification to reduce the community’s unnecessary cost in
terms of money and travel time.
page-76
page- 77
CHAPTER 3
Inputs of
The Health
Extension
Program
page-78
3 Inputs of the Health Extension
CHAPTER 3
Program
The availability of adequate inputs is a critical factor in implementing a health
program properly. Ensuring availability of the right input to the right degree
requires not only significant investment but also the efficiency and coordination
of various stakeholders at different levels of the health system. This section
assesses the status of various inputs to the HEP: human resources, health
infrastructure, drugs and other supplies, and funding.
3.1 Human resources in the HEP

3.1.1 Introduction
Ethiopia has made remarkable progress in expanding PHC services in recent
years through the expansion of health infrastructure and the development of
the health workforce. For example, the health workforce density has increased,
from 0.3 per 1 000 population in 2009 to 1.51 per 1 000 in 2016. Despite the
country’s huge strides in improving its health workforce, much more work
remains to attain the desired number and mix of health workers recommended
by the WHO.43,44
Ethiopia’s health policy prioritizes the expansion of PHC. The government

launched the HEP in 2003 and since then has accomplished the massive
construction of HPs in each kebele and trained HEWs to staff HPs. In this
regard, by 2018 a total of 36 642 HEWs were trained and deployed.45
The standard of health service delivery recommended in Ethiopia is that each

kebele should have at least 1 HP and that each HP should have at least 2
HEWs. An HP in an agrarian community is expected to serve up to 5000
people, while in pastoralist communities it is expected to serve up to 3000
people. The inclusion of additional packages in the HEP has also led in recent
years to the recommendation to have up to 4 HEWs in each HP and upgrade
all HEWs to level IV.44 This section presents the findings on the availability and
adequacy of human resources for the effective implementation of the HEP.
page- 79
Inputs of the Health Extension Program
3.1.2 Summary of key findings
A high proportion of HPs meet the minimum standard of 2 HEWs per HP. The
number of HEWs, however, is inadequate compared to the actual volume
of work. Almost all HEWs attended level III or level IV education and training
programs. Nevertheless, a considerable proportion of HEWs are not qualified
(i.e., CoC certified), and a substantial knowledge and skill gap was observed
among them. A low level of satisfaction and a relatively high level of burnout
symptoms were observed among HEWs. A high proportion of HEWs had
probable symptoms of depression. A relatively high level of intention to leave
the job was observed among HEWs. Nearly one third (32%) reported their
intention to leave their jobs. The actual rate of attrition was rather low, only
21% of HEWs had left their jobs from the start of the program to 2019. The
rate of attrition was 2.9 per 100 person-years.
3.1.3 Human resources planning
Human resources development and management has been given attention in

the national health sector plans.6 In line with this, different policy documents,
regulations, guidelines, and directives related to health human resources
development and management have been developed. The MoH has also
developed a national HRH strategic plan for the period 2016-2025. The plan
forecasts the HRH needs of the country through 2025 and estimates the cost
of realizing the plan.43 In Ethiopia, human resources planning and management
is decentralized: where regions, zones, and woredas have a mandate to hire
health workers.The human resources planning process at lower levels (region,
zone, and woreda) is not well developed and varies by region. An Electronic
Human Resource Information System was introduced in 2009 at the national
level, but it has not been established well enough to generate and disseminate
up-to-date evidence about the HRH situation of the country for decision-makers.
The system has not been functional at all levels.
The national HRH strategic plan projected that nearly 41 000 HEWs would
be needed in 2020 and 58 260 in 2025 to fulfill the minimum standard of
2 HEWs per HP.43 The plan did not take into account, however, the recent
recommendations to place up to 4 HEWs in each HP. Therefore, up to twice
page-80
the projected number of HEWs could be needed to staff HPs with 4 HEWs.46
The plan indicated that all HEWs would be upgraded to level IV, and some
to level V, during the implementation period. The hiring and placement of the
health workforce, including HEWs, is a decentralized process that mainly falls
under the purview of RHBs and WorHOs. The budget for human resources,
including salary and benefits, is allocated at the region, zone, or woreda level,
and WorHOs have a mandate to plan for and recruit health workers based on
the needs of the offices. The budget allocated, however, is usually inadequate
to hire the number of health workers required by the plan.43,46
A number of motivation and retention mechanisms were designed by the national

HRH strategic plan, including both financial and non-financial incentives.
The financial incentives include a professional risk allowance, transportation,
a housing and telephone allowance, and salary bonuses for less accessible
areas. Non-financial incentives include the provision of houses, scholarships
and other training opportunities, and research grants.43 Most of the proposed
incentive mechanisms, however, have not yet materialized, and it is not clear
from the plan how much of the proposed incentive packages consider HEWs.
Key informants and HEWs highlighted the absence of incentives and overtime
pay as major sources of dissatisfaction among HEWs. Moreover, the salary
and benefit packages for HEWs are not uniform across regions.
3.1.4 Recruitment and training of HEWs
The guidelines for implementing the HEP state that HEWs should be recruited
from the communities in which they will work. Selection and recruitment follows
specific criteria: they should be females, except in pastoralist areas, where males
can also be HEWs, be at least 18 years old, have at least a grade 10 education
(including grades 4 to 8 in pastoralist regions), and speak the local language.
In addition, HEWs are expected to be well versed in the communities’ culture
and norms.47
The recruitment and training of HEWs is designed on the basis of the human
resource needs of kebeles, as estimated by the WorHOs or ZHDs. A committee
comprising different stakeholders, including kebele leaders, select candidates
who fulfill the recruitment criteria. Selected candidates are then expected to
page- 81
be trained for 1 to 3 years and pass the CoC test before being deployed as
HEWs. The supervising HCs or WorHOs assign HEW supervisors and HEP
focal persons.
In practice, however, the recruitment process does not necessarily follow the
national directives. As a result, it is not uniform across the different levels
(regions, zones, and woredas). In some places, RHBs are primarily responsible for
recruiting trainees, while in most others, WorHOs or ZHDs are solely responsible
for the task. The relevant stakeholders, including WorHOs, education offices,
and community leaders, are generally involved in recruiting HEWs, although
this is not universal. There are also places where kebele leaders are responsible
for selecting candidates from their locality.
As stated in the HEP implementation manual, HEWs should be selected from

the kebele where they are expected to work to increase their cultural competence
and acceptance by the community. In some areas, however, particularly
pastoralist communities, HEWs are selected from nearby kebeles due to the
shortage of potential candidates. In some pastoralist communities, it is difficult
to find candidates who meet the education requirements (having attended
grades 10 to 12), and as a result, school dropouts with an education of grades
5 to 8 are recruited, which could potentially compromise the quality of the HEP
services. Moreover, upgrading to level IV or level V is not possible for them, as
completing grade 10 is a requirement for level-based college education. There
are also areas where HEWs are assigned to kebeles other than their local
communities.
In explaining the shortage of possible candidates and the resulting selection of

students with a grade 5 to 8 education for HEW training, one key informant
from a RHB stated:
“
Because of the shortage of educated people in the Afar region, both
men and women with a grade 5 to 8 education were selected for
training.
KII, Afar RHB
Key informants also underscored the need to revise the HEW selection and
recruitment criteria and apply more stringent requirements, as the current
page-82
requirements do not result in the selection of competent candidates for the

HEP. They argue that trainees have a limited capacity to acquire the necessary
knowledge and skills to be HEWs because, on the one hand, recruits are
either school dropouts or failed to pass grades 10 or 12 and have difficulty
understanding the medium (English) of instruction at colleges. On the other
hand, candidates join the program because they have no other training or
employment opportunities, which affects their motivation and performance.
Regarding the challenges in selecting and recruiting HEWs, a key informant
“
from one RHB said:
The HEP recruited those who left school because poor academic
performance or those who failed grade 10 or 12 and were unable to
attend college.
KII, SNNPR RHB
In explaining the consequences of loose selection criteria and the need to revise
the selection and recruitment criteria, a key informant from the Amhara RHB
“
said:
The selection criteria were not as strong as diploma programs’

selection criteria. I think that is why they have trouble passing the
CoC exams and improving their education. As I mentioned earlier,
we may need to modify the recruitment criteria to get competent
candidates.
KII, Amhara RHB
3.1.5 Pre-service training
Pre-service trainings for HEWs have been offered by TVET colleges operating under
the regional health and regional education bureaus. A total of 23 colleges are currently
involved in HEW training. In this study, the readiness of TVET colleges was assessed
using 5 criteria: (a) the relevance of the curriculum, (b) the adequacy of the inputs,
(c) the quality of course materials, (d) the appropriateness of the course delivery,
and (e) the methods of evaluation and competence of the programs’ graduates. A
number of capacity gaps that hinder the quality of education have been identified
page- 83
in this study. A complete report of the HEW training institutions’ assessment is

included in Part 4 of this report.
The assessment of the HEP curriculum showed that its contents are in line with
the country’s health policy and consistent with the actual job description of
the HEWs. The scope of the curriculum, however, is very broad, and the time
allotted to cover the content is relatively short. As a result, maintaining the
balance between theoretical and practical components (a 30% to 70% ratio,
as specified in the curriculum) was found to be problematic. The training was
found to focus more on theoretical subjects. The majority of HEW trainees
rated the relevance of the courses as very good or good in developing their
subject matter competence (85%) and skills (84%). Moreover, 81% of HEW
trainees rated the relevance of the courses to produce professionals who could
meet the needs of the community as either very good or good (Table 3-1).
Table 3-1. Trainees’ and instructors’ ratings of the relevance of the HEP curriculum
Aspect of Relevance Trainess (%) Instructors (%)
Percent of respondents who rated the relevance of the contents/

courses to develop subject matter competency of the HEP 84.8 70.8
trainees as very good/good
Percent of respondents on the relevance of the contents/courses
84.0 74.0
to develop the skills of HEP trainees as very good/good
Percent of respondents who rated the relevance of the courses
to produce professionals who could meet the needs of the 81.0 76.0
community as very good/good
Abbreviations: HEP, Health Extension Program.
Overall, HEW training institutions have a reasonable number and mix of in-
structors. The availability of facilities and equipment, however, is limited. A
total of 192 instructors work in the 23 institutions assessed, which range from 6
in Gambela to 93 in Mizan. About two thirds of instructors have a BSc degree,
and 27% have an MSc/MPH. There are a relatively large number of nurses,
midwives, and health officers among the instructors. There was no laboratory
technician or technologist in 5 of the 21 colleges assessed.
As shown in Table 3-2, although most of the required facilities (i.e., classrooms,
skill labs, practice sites, and library) are available in the colleges, most are
page-84
judged inadequate. For example, because the number of classrooms was so

limited, the class size could be as large as 70 or 80, compared to the standard
25 to 30 trainees in a class.
Table 3-2. State of HEP training facilities in Ethiopia, 2019
Internet
Toilet for
Library Skill Lab Computer Access Practice Water
College Classroom Males
Capacity Capacity Lab for Sites Point
(Females)
Trainees
Debre
14 250 + 25 +- ++ +- (+-) +-
Tabor
Minilik 1 100 + 30 +- ++ +- (+-) +-
Teda/
13 100 + 40 ++ ++ +- (+-) +-
Gondar
Metu 10 60 + 20 0 +- +- (+-) +-
Negele
15 30 + 40 +- +- +- (+-) +-
Borena
Dire
3 0 0 0 0 ++ +- (NS) ++
Dawa
Gambela 18 200 + 30 +- +- +- (+-) +-
Bahir
4 150 + 35 0 ++ +- (+-) +-
Dar
Hosana 10 400 0 0 0 ++ +- (+-) +-
Mekelle 38 200 + NS 0 +- +- (++) +-
Arba
15 160 0 20 0 ++ +- (+-) +-
Minch
Debre
12 40 + 45 ++ +- +- (+-) +-
Berhan
Semera 6 50 + NS 0 +- +- (+-) +-
Mizan 21 500 + 30 +- +- ++ (+-) ++
Dessie 14 80 + 25 0 +- +- (+-) +-
Abbreviations: HEP, Health Extension Program.

Note: 0=not available, +=available, +-=available but not adequate, ++=available and adequate,
NS=not sure
The assessment learned that the HEP teaching modules are up to standard.
Nevertheless, there is an imbalance between the amount of content and the
time allotted to it. In addition, no periodic revision is conducted to update the
page- 85
modules. The modular method of course delivery is preferred by most trainees

and instructors, but its effectiveness is hampered by a shortage of time, inadequate
training materials or modules, large class sizes, and the heavy workload of both
instructors and trainees.
Although the HEP promotes a practice-based teaching approach, in most

cases, teachers deliver the courses using a predominantly lecture method. There
is a limited focus on projects and collaborative learning. The instructors are
educationally and professionally qualified to deliver the courses. Nevertheless,
some instructors deliver courses outside their field of specialization. The
methods used to assess the trainees are mainly summative-type written tests.
Instructors do not often employ performance evaluation. The HEP trainees are
not competent in many aspects, a fact attributed to several factors, including
a problematic recruitment procedure, their poor English language proficiency,
and the TVET-RHB controversy over the responsibility for HEW training.
3.1.6 Deployment, roles, and responsibilities of HEWs
A considerable proportion (36.4%) of HEWs in the current study were not

given orientation before their deployment as HEWs. The practice of providing
orientation for HEWs is higher in pastoralist areas (68.3%) than agrarian areas
(33%). There is significant regional variation in orienting newly deployed HEWs,
ranging from 27.6% in Somali to 72.5% in Harari. The quality of orientation
given to HEWs is unclear, as there is no document describing how orientation
is to be provided to HEWs and by whom it is to be delivered. The qualitative
study, however, disclosed the presence of role ambiguity among HEWs.
“
A key informant HEW from an agrarian area confirmed the absence of orien-
tation and its effect on newly assigned HEWs:
Before we started the work, we didn’t have orientation. Education

and work are two different things. We were not aware of how we
should work when we were deployed in the community.
KII, Agrarian HEW
page-86
“
One HC head also identified the need for orientation during HEWs’ deployment as
follows:
The health center should give orientation for the HEWs before
deployment. They have to know what they will do there.
KII, Health Center Head, Pastoralist Area
The provision of a job description to newly assigned HEWs appears to be

overlooked. The current findings indicated a gap in the provision of a job
description. Only 21% of HEWs were given a job description before or within
6 weeks of their deployment. The percentage of HEWs who had ever been
provided with a job description was 24%. In addition to the rarity of providing
a job description to HEWs, there is also significant regional variability in this
matter. In relative terms, a higher proportion of HEWs (53%) were provided
with a job description in Gambela, while only 12% in Oromia, 17% in Benishangul-
Gumuz, 18% in Afar, and 23% in Harari were (Table 3-3). Key informants also
stated that HEWs do not have a clear job description and this causes confusion
in their roles and responsibilities. HEWs also complained about having multiple
supervisors, including kebele administrators, HC heads, and WorHO heads,
all of whom give orders without considering the HEWs’ workload. Different
government sectors in the kebele, such as Health, Agriculture, and Education,
are expected to work as a team to effect the desired changes. In some instances,
however, the team has no clearly defined roles and responsibilities, which creates
role ambiguity, and HEWs are obliged to perform tasks unrelated to health.
page- 87
Table 3-3. Provision of orientation and a job description to HEWs
Proportion of HEWs who were given:

Job description Unweighted # of
Background Orientation before
Job description before HEWs
deployment
deployment
National 24.3 21.4 36.4 584
Livelihood
Agrarian 24.4 21.3 33.0 414
Pastoralist 23.6 22.8 68.3 170
Region
Tigray 34.7 34.7 69.8 63
Afar 18.1 18.1 65.5 19
Amhara 34.4 30.6 69.3 95
Oromia 12.1 10.7 68.5 123
Somali 27.1 27.1 27.6 75
Benishangul-
16.8 12.4 59.6 37
Gumuz
SNNPR 35.8 30.1 58.0 96
Gambela 53.4 50.6 52.0 42
Harari 23.1 6.3 72.5 34
Abbreviations: HEW, Health Extension Worker.
Placing HEWs in their own localities to serve their own communities has obvi-
ous advantages. The practice, however has caused dissatisfaction and attrition
as reported by HEWs and other key informants. For instance, HEWs are forced
to work in one kebele for a long time despite their repeated transfer requests,
leading to grievances among HEWs. The national HRH strategic plan and
HEP implementation guidelines are not clear on how to reconcile the design of
the HEP, which requires HEWs to be placed in their local kebeles even after
upgrading their education level, with HEWs’ need to transfer to other areas for
various reasons. Moreover, the idea that it is a benefit for HEWs to serve their
own communities is not borne out universally.
page-88
3.1.7 Adequacy of the number of HEWs
The assessment showed that there were on average 2.5 HEWs per HP. The
majority (86%) of HPs were staffed with 2 or more HEWs, meeting the mini-
mum requirements for the number of HEWs per HP as indicated in the HEP
implementation manual, but 14% of HPs were staffed with only 1 HEW. In ad-
dition, some HEWs were enrolled in training for upgrading at the time of the
survey, creating a shortage of HEWs on active duty. Since there is no mecha-
nism to replace HEWs who are enrolled in upgrading training, very few (12%)
HPs were staffed by 4 or more HEWs, indicating the enormous challenge of
the task ahead to realize the recent recommendation to assign up to 4 HEWs
per HP (Table 3-4). Key informants also identified the shortage of HEWs as a
detriment to the overall performance of the HEP. Describing the shortage of
HEWs and its effect on service provision, one WorHo process owner said that:
“
The number of HEWs we have is inadequate to perform all the tasks
of the HEP. As a result, the coverage of HEP-related services is not
what it used to be.
KII, WorHo Process Owner
Even though a majority of HPs were staffed by 2 or more HEWs, the number may
not be adequate given their volume of work. In some instances, the vast geographic
area and large population size of a kebele causes an enormous workload for HPs.
Regarding this issue, key informants from WorHOs and HCs indicated that the usual
number of HEWs (2 per HP) is insufficient for some kebeles, which necessitates the
revision of the minimum standard to tailor it better to the specific needs of specific
kebeles. In explaining the need for more than 2 HEWs in some kebeles, a WorHO key
“
informant stated:
Some kebeles are too vast, requiring more than 3 hours’ walk to
provide services to the community, particularly in the pastoral kebeles.
This might affect the coverage of HEP-related activities.
KII, WorHo Process Owner
page- 89
One RHB head also explained the shortage of HEWs in some kebeles due to the
“
increasing population as follows:
The population number is increasing, and kebeles are continuously

divided into a greater number of kebeles. As a principle, there should
be 1 HP and two HEWs in each kebele, but we are not constructing
more HPs or recruiting more HEWs to cope with the increasing
number of kebeles.
KII, RHB Head
Table 3-4. Staffing level of HPs, by region and livelihood
Number Proportion of health posts having:

of
At At At least At least Unweighted
Background HEWs Two or
least 1 least 1 1 CoC 1 HEW # of HPs
per HP more
level IV nurse/ certified currently on
(mean) HEWs
HEW midwife HEW training
National 2.4 86.7 62.4 4.0 83.5 3.5 343
Livelihood
Agrarian 2.4 86.5 65.3 3.5 85.8 3.1 235
Pastoralist 2.6 88.5 33.5 8.9 60.5 7.6 108
Region
Tigray 2.5 90.0 80.9 6.4 100 6.4 32
Afar 3.0 100 20.7 91.0 29.7 71.5 18
Amhara 2.3 88.7 69.0 2.6 73.2 2.6 60
Oromia 2.3 84.5 64.9 1.8 93.1 1.8 74
Somali 2.6 85.7 33.0 4.2 41.6 4.2 43
Benishangul-
2.6 92.5 52.3 46.4 35.2 32.0 24
Gumuz
SNNPR 2.7 87.8 58.0 2.1 87.1 2.1 59
Gambela 5.8 94.9 66.5 74.5 87.2 51.4 17
Harari 3.1 100.0 58.7 13.5 69.1 8.7 16
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples; HP, health post; HEP, Health
Extension Program; HEW, Health Extension Worker; CoC, certificate of competence.
page-90
Absenteeism is a major problem facing the health systems of developing

countries. It causes staff shortages and compromises service delivery. Most
studies define work absenteeism among health workers as the proportion of
health workers who are unavailable in the health facility during the assessment
period. This definition, however, cannot be used for the current study because
HEWs are required to spend 50% of their time doing outreach services. In the
current assessment, work absenteeism is defined as the proportion of HEWs who
were not present in the HP or in the kebele during the time of the assessment.
From the total of 927 HEWs who were staff at the 343 HPs selected for the
study, 198 (21%) were absent during the time of assessment, exacerbating the
shortage of HEWs on active duty. At least 1 HEW was absent during the
time of this assessment from nearly half of the HPs (48%). The main reasons
for absence were maternity leave (23%), sick leave (13%), long-term training
(13%), IRT (6%), and annual leave (6%; Figure 3-1).
Key informants and FGD participants consistently mentioned that HEWs open
the HPs for only some days of the week and that clients and patients visiting
HPs return without receiving services. The main reasons identified by key
informants were a shortage of staff, a lack of motivation, and the probability
that an HEW’s residence is far from the HP.
One male community member who participated in the FGD described the
“
absence of HEWs as follows:
HEWs are not always available at HPs. Due to this, the community
gets health services from private clinics.
FGD, Male Community Members
page- 91
Figure 3-1. Reasons for HEWs’ absence during the time of visit
Abbreviations: HEW, Health Extension Worker; IRT, Integrated Refresher Training.
3.1.8 HEWs’ Qualifications and Competence
HEWs should have at least a level III qualification to be considered qualified.

Currently, upgrading all HEWs to level IV is underway to improve the extent
and quality of HEP services, fill the gaps identified in previous level III HEP
trainings, and improve the knowledge, skills, attitudes, and careers of HEWs.
Nurses and midwives are also assigned to some HPs, particularly in pastoralist
communities where referral HCs are located far from HPs. In this regard, the
current HEW assessment showed that 98% of HEWs were qualified (i.e., had a
level III or level IV qualification), 50% had a level III qualification, and 48% had
a level IV. Nearly 2% of the HEWs had a level I or level II qualification, even
though the current HEP implementation guidelines do not recognize HEWs with
qualifications below level III. All of the unqualified HEWs are located in HPs
in pastoralist communities (Table 3-5). The qualitative data from pastoralist
areas indicates that efforts to accelerate the upgrading of level I and level II
HEWs could not proceed as planned because level I and level II HEWs had
not finished high school, a requirement for enrolling in level III training. The
shortage of high school graduates from rural areas remains a challenge in
recruiting HEWs in pastoralist areas.
page-92
The HP assessment also showed that 61% of HPs have at least 1 level IV HEW,
and 5% of HPs have at least 1 nurse or midwife. HPs located in agrarian
communities have a higher proportion of level IV HEWs than do pastoralist
HPs. On the other hand, the proportion of nurses and midwives is higher among
HPs in pastoralist communities than those in agrarian communities (Table 3-5).
Mid-level professionals in Ethiopia are required to take a national CoC test

to protect institutions from hiring incompetent college graduates. HEWs, as
one category of mid-level professionals, are required to take the CoC test.
This requirement has not been consistently applied to graduates from training
institutions owned by the public sector. In addition, HEWs who were hired
before the directive was issued remained in the system even though they might
not have taken the exam or could not pass it. Getting a CoC has been used
as one criterion for HEWs’ career development and admission to upgrading
training.
A summary of the findings from the current study regarding qualifications and
CoC is presented in Table 3-5. The majority (82%) of HEWs have taken the
CoC certification test, but only 57% are certified, and 25% are not certified
despite having taken the test. Most HPs (84%) have at least 1 CoC-certified
HEW on staff. Among the HEWs who have passed the CoC test, 58% passed
and certified on their first attempt, 36% passed on their second attempt, and
6% passed after attempting the test more than twice. Similarly, among HEWs
who took the CoC test but did not pass, 62% attempted once, 26% attempted
twice, and 12% attempted more than twice (Table 3-5, Figure 3-2).
The failure of HEWs to get their CoC was repeatedly mentioned by different
key informants. Many stated that it indicated the HEWs’ incompetence. The
presence of a skill gap among HEWs due to the recruitment of individuals with
poor academic performance and a poor-quality college education that focused
more on theory than practice were mentioned as possible causes for the failure.
page- 93
Table 3-5. Qualification and CoC certification of HEWs, by livelihood (N=584)
Livelihood (%)
Characteristics Agrarian Pastoralist National (%)
HEW qualification
Level I or level II 0.0 22.7 10.3
Level III 49.4 49.8 49.5
Level IV 50.6 27.5 48.4
Took CoC certification test
Yes, and certified 58.1 49.5 57.3
Yes, but not certified 27.0 7.0 25.1
Did not take CoC test 15.0 43.5 17.6
Number of times took CoC test (n=450)
1 58.4 78.3 59.7
2 times 33.4 19.7 32.5
More than 2 times 8.2 2.0 7.8
Plan to upgrade current qualification level
Yes 91.3 83.4 90.6
No 8.7 16.6 9.4
Abbreviations: CoC, certificate of competence; HEW, Health Extension Worker.
Took COC ones Took COC twice Took COC three or more
Not certified
Certified
Figure 3-2. HEWs’ CoC status, by the number of times taking

the CoC test
Abbreviations: CoC, certificate of competence; HEW, Health Extension Worker.
page-94
3.1.9 HEWs’ knowledge and skills

As part of assessing their knowledge, HEWs were asked to name the first-
line drug of choice to treat uncomplicated malaria in children, adults, and
pregnant women, according to the Ethiopian malaria treatment guidelines.
This question was chosen based on the epidemiology of malaria in Ethiopia
and as a key duty of HEWs. Their knowledge of the type of vaccines currently
included as part of the Ethiopian Expanded Program on Immunization (EPI)
program schedule was also assessed. The current Ethiopian EPI includes 10
antigens scheduled to be provided over 5 visits. The schedule comprises Bacillus
Calmette-Guerin (BCG) and Oral Polio Vaccine (OPV) at birth; pentavalent
vaccines (Diphtheria-Pertussis-Tetanus, Hepatitis B/Hemophilus Influenza B),
Pneumococcal Conjugate Vaccine (PCV), and OPV at 6, 10, and 14 weeks;
rotavirus vaccine at 6 and 10 weeks; and inactivated polio vaccine (IPV) and
measles vaccine at 9 months.
The findings indicated a substantial knowledge gap among HEWs when

mentioning the EPI schedule. Among the HEWs asked about vaccines, 88%,
81%, 79%, and 62% correctly mentioned the vaccines given at birth, 6, 10, and 14
weeks, respectively. About half (51%) correctly mentioned all of the vaccination
schedules. Knowledge of the vaccination schedules was higher among HEWs
working in agrarian HPs (53%) than pastoralist HPs (31%). One possible reason
for this knowledge and skills gap could be the recent introduction of some of
the vaccines (i.e., PCV) to the EPI program; HEWs may not be trained on the
newly introduced vaccines. This reason alone, however, cannot fully explain the
observed knowledge gaps. It is not known whether HEWs who fail to correctly
mention the EPI schedule are involved in the provision of vaccination services.
It is possible that only some HEWs provide the vaccination service. Most (95%)
HEWs correctly answered that vaccines should be stored in a refrigerator at a
temperature between 2oC and 8oC (Table 3-6). The HEWs’ knowledge of the
EPI schedule increased with their level of education (Figure 3-3).
Regarding HEWs’ knowledge in treating malaria, the current assessment found

that only 40% of HEWs correctly identified the first-line drug of choice to treat
vivax malaria among adults, 39% correctly identified the first-line drug for vivax
malaria among children, and 16% identified the first-line drug for the treatment
of pregnant women. The remaining HEWs either identified the wrong drugs
or responded that they did not know. Similarly, the first-line drugs for the
page- 95
treatment of falciparum malaria for adults, children, and pregnant women

were correctly identified by 67%, 55%, and 35% of HEWs, respectively. HEWs
working in agrarian HPs had better knowledge in identifying the first-line drugs
used to treat uncomplicated malaria than did HEWs working in pastoralist
HPs (Table 3-6).
Table 3-6. HEWs’ knowledge of EPI schedule and malaria treatment

guidelines (N=584)
Category
Indicators Agrarian Pastoralist National (%)
(%) (%)
Knowledge on EPI schedule
Know vaccines given at birth 89.2 71.8 87.6
Know vaccines given at 6 weeks 82.2 67.7 80.8
Know all vaccines given at 10 weeks 80.2 67.3 79.0
Know all vaccines given at 14 weeks 63.9 44.1 62.0
Know vaccine given at 9 months 99 93 98.6
Know all vaccination schedules and vaccines 53.2 31.2 51.1
know vaccines should be kept in refrigerator at
95.0 92.3 94.7
2 to 8°C
Malaria treatment
Know the first line drug of choice to treat
44.8 24.1 40.1
uncomplicated vivax malaria in adults
40.1 29.1 39.1
uncomplicated vivax malaria in children
uncomplicated vivax malaria among pregnant 16.9 10.7 16.3
women
68.2 60.1 67.4
uncomplicated falciparum malaria in adults
53.9 60.4 54.5
uncomplicated falciparum malaria in children
uncomplicated falciparum malaria in pregnant 32.9 50.9 34.6
women
Abbreviations: HEW, Health Extension Worker; EPI, Expanded Program on Immunization.
page-96
Figure 3-3. HEWs’ knowledge of EPI schedule, by level of education
Abbreviations: HEW, Health Extension Worker; EPI, Expanded Program on Immunization.
Key informants from WorHOs and HCs also indicated the presence of a
knowledge and skills gap among HEWs. Similarly, some community leaders
perceived the health problems of their community to be beyond the knowledge
and skills of HEWs because curative services are not provided by HEWs. The
reasons for the skills gap among HEWs, as perceived by HEWs and other key
informants, include their lack of practical training (their training focuses only
on theoretical aspects) during their pre-service education and the recruitment
of individuals with poor academic performance.
Regarding the presence of a knowledge and skills gap among HEWs, a key
informant from a WorHO expressed his view thus:
“
Services provided by HEWs have some quality problems. HEWs have
skills gaps in providing service as per the required standard on the
MCH services, such as ANC, delivery, postnatal care, EPI, calculating
expected date of delivery, screening children for malnutrition and
treating malnourished child and conducting public conferences. This
is common particularly among recently graduated HEWs.
KII, WorHO
page- 97
A male FGD participant mentioned a lack of skill among HEWs to provide
“
curative services:
The diseases from which our communities suffers from are beyond
their capacity. For example, how can they treat a person with liver
disease? This patient have to go to hospitals. It is the same for other
diseases like hypertension, diabetes, and peptic ulcers.
In explaining the reason for the observed skill gap, a WorHO process owner
“
stated:
I think the main problem is during the selection and screening of

HEWs for training. They are not carefully recruited, and attention is
not paid to the selection process. Even the CoC exam given to them
was just for the sake of formality, not to assess their knowledge.
KII, WorHO Process Owner
The level of HEWs’ competence in accomplishing HEP-related tasks was also

evaluated by asking them to rate their own competence in different areas
and give their response as: “I am not able to do it,” “I can do it with some
guidance,” “I can do it by myself,” or “I can train others on how to do it.” The
areas of competence evaluated include detecting danger signs and managing
labor and delivery, providing FP services, and diagnosing common childhood
illnesses. The majority (85%) of HEWs reported that they could detect signs
of danger during pregnancy, 59% could attend a normal delivery, and 57%
could provide a timely diagnosis of labor complications. There was regional
variation in the self-rated competence of HEWs to manage labor and detect
complications.
FP provision is a core task of HEWs. Current findings show that HEWs have
a high level of competence in providing short-term FP methods, such as oral
contraceptive pills and injectable contraceptives. A majority of HEWs, however,
lack the competence to provide long-term FP methods. For instance, only 11%
of HEWs claimed that they could insert an intra-uterine contraceptive device
(IUD), and 8% could remove one. Similarly, 63% of HEWs were able to insert
page-98
Implanon, but only 23% could remove it. Only 1 of 5 level IV HEWs could insert
an IUD, compared to 3% of level III HEWs or 7% of level I or level II HEWs.
Similarly, 85% of level IV HEWs claimed that they are able to insert Implanon,
compared to 59% of level III HEWs and 72% of level I or level II HEWs (Table
3-7).
Table 3-7. HEWs’ self-reported competence in attending labor, detecting signs

of danger in delivery, and providing long-term FP, by background characteristics
HEWs self-reported competency (%)

Characteristics
Unweighted #
of HEWs
Detect
Attend Timely detect
danger Insert Insert Remove Remove
normal complications
signs in IUD Implant IUCD Implant
labor of labor
pregnancy
National 85.1 59.0 56.6 10.8 72.9 8.1 23.3 584

Qualification
Below level III 85.6 57.5 63.7 7.3 72.1 2.4 2.4 24
Level III 81.0 49.0 47.6 2.5 58.8 1.9 13.2 321
Level IV 88.6 67.8 62.6 18.7 85.2 14.7 36.3 239
Livelihood
Agrarian 85.7 59.3 56.3 10.6 76.9 8.0 24.4 414
Pastoralist 79.4 56.5 58.8 13.1 34.8 9.5 11.8 170
Region
Tigray 100.0 57.8 50.5 17.8 94.7 19.3 53.9 63
Afar 89.3 49.5 57.6 15.6 15.6 15.6 15.6 19
Amhara 87.1 53.1 59.8 10.8 77.1 9.3 20.0 95
Oromia 81.2 59.7 44.0 10.7 78.5 8.7 30.4 123
Somali 69.4 72.2 73.6 18.6 18.6 14.3 17.0 75
Benishangul-
94.9 49.9 55.8 7.0 35.9 4.8 23.9 37
Gumuz
SNNPR 92.3 60.5 74.2 7.9 73.7 2.1 9.6 96
Gambela 64.8 70.4 50.7 33.8 33.8 36.6 33.8 42
Harari 85.1 12.6 27.9 8.9 61.3 15.2 67.3 34
Abbreviations: HEW, Health Extension Worker; FP, family planning; IUD, intra-uterine device; SNNPR,
Southern Nations, Nationalities, and Peoples Region.
page- 99
3.1.10 In-service training of HEWs
In-service training is essential to improve workers’ knowledge, skills, performance,

and motivation, and it is considered a pillar of human resources development.
Therefore, most organizations incorporate in-service training into their human
resources development plans. It is particularly relevant for HEWs because
the HEP has evolved rapidly since its inception in 2003. Currently, the HEP
package incorporates high-impact curative services, such as iCCM, CBNC,
and the treatment of common childhood illnesses. Therefore, HEWs need
additional training on the newer intervention packages. To this end, the MoH
has developed IRT, a comprehensive in-service training package that all HEWs
should take at least every other year.
The current study assessed the IRT participation status of HEWs; the finding
is summarized in Table 3-8. About 42% of HEWs took IRT 2 or more times
(per the recommendation) in the 5 years prior to the survey, and 3 out of 4
HEWs participated at least once during the same period. The finding indicated
that IRT is not being implemented as intended. In addition, IRT is not fairly
distributed, as shown in the finding: 32% of HEWs did not get the chance to
participate in IRT, while 42% of HEWs got the chance to participate 2 or more
times. HEWs from Afar (33%) and Somali (49%) had the lowest proportion
of participating in IRT at least once, while HEWs from Amhara reported the
highest (84%) IRT participation, followed by HEWs from Oromia (68%) and
Benishangul-Gumuz (67%; Table 3-8).
page-100
Table 3-8. HEWs’ participation in IRT
Participation in IRT in the past 5 years (%)

IRT at IRT Unweighted no.
Background 2007 2008 2009 2010 2011
least 2 or of HEWs
EFY EFY EFY EFY EFY
once more
National 42.2 34.6 29.4 22.6 14.2 68.5 41.5 584
Livelihood
Agrarian 44.9 35.6 30.6 22.1 13.1 69.7 42.9 414
Pastoralist 16.9 24.3 17.4 27.4 25.5 56.5 27.6 170
Region
Tigray 24.3 21.3 29.5 10.8 5.3 50.4 22.9 63
Afar 12.0 0.0 4.1 18.3 10.3 32.7 12.0 19
Amhara 64.7 44.1 38.5 20.6 14.7 83.6 51.0 95
Oromia 45.6 37.2 33.3 21.3 8.3 68.0 45.4 123
Somali 13.0 18.3 16.5 28.0 23.9 48.5 26.5 75
Benishangul-
34.2 27.2 29.8 58.5 46.9 67.3 53.4 37
Gumuz
SNNPR 27.5 28.6 17.4 24.7 21.7 64.4 32.3 96
Gambela 15.5 24.0 30.9 19.6 9.9 63.4 19.7 42
Harari 17.1 38.3 34.2 36.1 32.0 57.3 40.5 34
Abbreviations: HEW, Health Extension Worker; IRT, Integrated Refresher Training; SNNPR, Southern
Nations, Nationalities, and Peoples Region.
The great majority of HEWs who had attended IRT recommended the training
to other HEWs. Only 15% of HEWs in Somali, 8% in Amhara, and 3% in Tigray
did not recommend IRT to other HEWs (Figure 3-4).
page- 101
Figure 3-4. HEWs’ recommending IRT to other HEWs, by region

Abbreviations: HEW, Health Extension Worker; IRT, Integrated Refresher Training; B/Gumz, Benishan-
gul-Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples Region.
IRT was introduced to limit the delivery of fragmented short-term trainings. The
assessment disclosed, however, that HEWs were participating in a number of
short-term trainings other than IRT. The result showed that newly introduced
high-impact curative interventions are the most commonly attended short-term
trainings. Accordingly, iCCM was the most widely attended (82%), followed by
CBNC (71%) and long-acting FP (68%). Community-Based Health Insurance
(CBHI) was a topic of training commonly attended by HEWs, although the
role of HEWs in CBHI implementation was not clearly articulated in the HEP
implementation guidelines. NCDs are the least attended short-term training;
only 29% of HEWs participated (Table 3-9).
page-102
Table 3-9. Short-term trainings attended by HEWs, by training topic and live-
lihood (N=584)
Livelihood
National
Topics of Short-term trainings
Agrarian Pastoralist (%)
(%) (%)
Short-acting family planning 56.4 40.0 54.8
Long-acting family planning 71.0 39.1 68.0
Antenatal care 33.8 30.3 33.5
Labor and delivery 34.9 27.0 34.1
Postnatal care 34.3 29.4 33.8
Community based newborn care 74.8 37.7 71.3
Integrated Community Case Management 85.4 60.0 83.0
IEC/BCC 44.5 24.9 42.6
Non-communicable diseases 29.2 28.5 29.1
Malaria 50.6 48.4 50.4
Tuberculosis 62.1 36.9 59.7
HIV/STIs 52.4 31.2 50.4
Community-based health insurance 57.3 26.7 54.4
Abbreviations: IEC/BCC, Information Education Communication/Behavior Change Communication
HEWs perceived that any kind of in-service training increases their knowledge and
skills. Specifically, IRT is perceived as the most useful training. Key informants also
highlighted that in-service training is essential, particularly for newly assigned HEWs,
as they are not confident in providing services unless they get additional in-service
training. Despite its relevance, there is an unmet need for some training topics. For
instance, HEWs requested and have interest in receiving training on NCDs, TB, nutri-
tion, the FF, and IUD insertion and removal. In contrast, some HEWs perceived that
some trainings did not contribute to their professional development.
Key informants also pointed out that language barriers and the limited time devoted
to the delivery of short-term trainings pose a challenge to understanding the training
content:
page- 103
“
Not all HEWs fully understand the instructional language of the
trainings. Language barrier is a great challenge to HEWs. Some of
the HEWs want the training modules to be translated into their local
languages. This is one of the main challenges in providing in-service
training.
KII, RHB Coordinator
3.1.11 HEP Career Development
The assessment found that HEWs’ career development was not given due
attention during the design of the HEP. Neither promotion nor educational
advancement were clearly described. Therefore, HEWs have long been denied
the opportunity to advance their careers. Cognizant of this problem, the MoH
has designed a career development path so that HEWs can advance. The
design states that HEWs who are currently at level III can upgrade to level IV
once they pass the CoC exam and engage in college education for an addi-
tional year. A path has also been designed to upgrade level IV HEWs to level
V (equivalent to a B.Sc. degree).44 Still, the career path of HEWs beyond the
degree level has not been clearly articulated. The other problem of HEWs’ cur-
rent career paths is that some HEWs, particularly those working in pastoralist
HPs, do not have a level III qualification. These HEWs dropped out of school
(receiving only a grade 5 to grade 8 education) and cannot upgrade to level III
or level IV because a diploma is a prerequisite for enrollment in college. From
the design it appears that HEWs are expected to work and live in rural areas
and have no opportunity to work in urban areas. The revised HEP implementa-
tion manual states that HEWs can work in WorHOs and HCs, but this has not
been put into practice in most places.
The path for rural HEWs to work as urban HEWs is unclear. Moreover, the ab-
sence of a path for HEWs working in rural area to become clinicians or other
public health professionals is a major problem in the career development of
HEWs.
page-104
In the qualitative study, key informants expressed diverse views about HEWs’
career development opportunities. Many believed that the fate of HEWs’ ca-
reer and educational development was uncertain and claimed that the govern-
ment was not sufficiently emphasizing career advancement opportunities for
HEWs, while others argued that the current career development opportunities
for HEWs was adequate, noting the massive upgrading training of level III to
level IV as an example.
Moreover, key informants have different views regarding the benefits of HEWs’
educational advancement for the program, with some indicating that HEWs’
leaving their workplaces for educational opportunities creates a staff shortage
at HPs and decreases the quality of their service provision. Others argue that
HEWs’ educational opportunities and professional growth improve their moti-
vation and decrease their attrition.
HEWs complain that even those who have worked for a long time still receive
few educational opportunities to advance their careers. This demotivates not
only senior HEWs but also the newly assigned HEWs contemplating a future
career as HEWs. The existing scholarship opportunities and salary improve-
ments are described as unsatisfactory.
One HC head described the lack of educational opportunity and promotion
“
thus:
There are HEWs who have worked for more than ten years. But there
is no promotion. Although HEWs suffer a lot, no benefit is provided
to them. This causes burnout. When I said promotion, it might be
education, giving recognition, and allowing them to transfer from
place to place. Even the newly deployed HEWs also see their seniors
and become demotivated. This has a negative impact on their work.
KII, HC Head
page- 105
3.1.12 Job satisfaction and burnout
A Likert-type scale measurement was used to assess HEWs’ job satisfaction.

All items were answered on a 4-point scale, where 1 means “very dissatisfied,”
2 “dissatisfied,” 3 “moderately satisfied,” and 4 “very satisfied.” The measure-
ment contains 33 items and consists of 8 dimensions of satisfaction: leadership,
promotion, autonomy, work environment, professional training opportunity, job
security and salary, recognition at work, and perception about other job opp-
ortunities. The reliability test of the instruments showed good internal consi-
stency (the Cronbach’s alpha coefficients for each of the dimensions ranged
from 0.72to 0.91). The overall level of satisfaction was measured using the
mean of thescale as a cutoff point to categorize respondents as either satis-
fied or dissatisfied. Accordingly, respondents who scored above the mean
were categorizedas satisfied, and those who scored at and below the mean
were considereddissatisfied. Similar procedures were followed to measure
the satisfaction of HEWs in each dimension.
Overall, more than half of the HEWs (51%) were dissatisfied with their jobs.
When the level of satisfaction in each domain was explored, the majority of
HEWs were satisfied with their autonomy (72%), their work environment, their
relationships with co-workers (66%), and the recognition they get at work (54%).
Only 13% of HEWs, however, were satisfied with their job security, salary, and
benefits, 25% were satisfied with their perceived alternative employment op-
portunities, and 29% were satisfied with their opportunities for promotion (Ta-
ble 3-10).
page-106
Table 3-10. Job-related satisfaction among HEWs
Proportion of HEWs satisfied by:
environment
Recognition
relationship
Job security
opportunity
opportunity
Leadership
Promotion
Autonomy
Perceived
other job
Training
&salary
at work
Unweighted
Work
Background
# HEWs
National 44.5 29.4 72.0 66.9 33.6 13.2 56.1 24.6 584
Livelihood
Agrarian 40.9 26.2 71.0 66.2 30.6 9.9 54.6 22.5 414
Pastoralist 79.9 60.6 81.5 73.9 60.7 45.7 69.9 44.9 170
Region
Tigray 48.3 22.7 70.4 65.4 37.8 12.3 42.4 19.5 63
Afar 68.0 27.2 84.0 74.7 19.3 9.1 56.6 29.0 19
Amhara 42.8 16.9 58.2 65.0 40.1 8.2 61.4 38.5 95
Oromia 37.2 31.3 74.7 64.3 22.8 12.5 40.9 9.8 123
Somali 90.4 76.5 92.1 91.2 78.7 65.2 85.3 63.9 75
Benishangul-
66.7 63.2 89.3 76.3 62.1 39.0 85.8 25.0 37
Gumuz
SNNPR 45.8 24.2 74.1 67.3 33.7 3.9 72.8 31.0 96
Gambela 55.0 26.8 36.7 52.0 52.1 29.8 34.7 35.2 42
Harari 47.2 42.1 89.4 93.7 65.0 0.0 54.4 25.7 34
HEWs’ overall satisfaction was estimated using the mean of the composite
score of the scale. The distribution of overall satisfaction of HEWs stratified by
selected background characteristics is summarized in Table 3-11. The HEWs’
level of satisfaction decreases with age and work experience; for example,
there is a 15% decrease in the level of satisfaction between HEWs aged 18
to 24 and those 30 and older. Similarly, HEWs with 1 to 5 years’ work experi-
ence had a satisfaction level 12% higher than HEWs with 11 to 16 years’ work
experience. In addition, HEWs living in the same kebele as the HP were more
satisfied with their jobs than HEWs living in a different kebele or nearby
town. The level of satisfaction varies significantly by region and livelihood.
HEWsworking in HPs located in pastoralist communities were more likely to
be sat-isfied than HEWs working in agrarian HPs (68% vs. 47%).
page- 107
Table 3-11. Distribution of level of satisfaction of HEWs, by background characteristics
HEWs level of satisfaction (%) Unweighted # of

Background characteristics
Dissatisfied Satisfied HEWs
National 51.4 48.6 584
Age category in years
18-24 38.3 61.7 189
25+29 55.0 45.0 272
30+ 55.6 44.5 123
Marital status
Never married 49.2 50.8 159
Married 50.9 49.2 399
Others 77.7 22.3 26
Level of qualification
Below level III 62.0 38.0 24
Level III 52.3 47.7 321
Level IV 48.4 51.6 239
Work experience
1-5 42.9 57.1 254
6-10 55.0 45.0 158
11-16 54.7 45.3 172
Region
Tigray 51.7 48.3 63
Afar 43.8 56.2 19
Amhara 53.1 46.9 95
Oromia 57.4 42.6 123
Somali 12.1 87.9 75
Benishangul-
21.8 78.3 37
Gumuz
SNNPR 50.2 49.8 96
Gambela 62.1 38.0 42
Harari 41.3 58.7 34
Livelihood
Agrarian 53.4 46.6 414
Pastoralist 32.0 67.9 170
Residence
Same kebele as HP 49.5 50.5 394
Other kebele/nearby town 55.0 45.0 190
Abbreviations: HEW, Health Extension Worker; HP, health post; SNNPR, Southern Nations,
Nationalities, and Peoples Region.
page-108
Key informants also noted the low level of satisfaction and motivation among
HEWs. A number of issues were identified as a cause for this dissatisfaction:
a lack of opportunity for career advancement, a low level of acceptance by
the community, limited transfer and promotion opportunities, and inadequate
“
salary and incentives. One HEW captured many of these issues well:
I feel disappointed that when you work as HEW there is no

advancement. It has been thirteen years since the program started,
and for the HEWs who were hired back then, there is no rating on
their performance. No educational opportunities of any sort. There
is no incentive, no change at all. Even if the government provides
education, there is no change in the monthly salary of HEWs.
When I think about this, I feel disappointed. The work is really hard.
Even society gives us names, such as “burned faces!” because of
the toughness of the job. We are sacrificing for our country. We go
from home to home to give services, walking for three hours. Though
we are paying such sacrifices, there is no incentive for it, and it is
disappointing.
KII, Agrarian HEW
Resistance from the community was also mentioned as a reason HEWs become
demotivated. In describing the resistance of the community, an HEW from an
“
agrarian area mentioned her personal experience thus:
During home visits, I once went to one house to provide health

education, but the woman yelled at me and ordered me to get out of
her house. I felt ashamed, and I don’t forget that moment.
KII, Agrarian HEW
HEWs also mentioned helping the community as a main source of their

satisfaction.Two HEWs in agrarian communities stated the following:
page- 109
“
First of all, we are helping our fathers, mothers, brothers, and sisters.
So helping them, changing their mind, is a lot for me. Also, observing
a reduction in mother and child death gives me mental satisfaction.
KII, Agrarian HEW
“
I am so happy working as an HEW because I have a mother, and
working for mothers, convincing them to take my advice, makes me
happy. So, supporting mothers to deliver is an honor for me. Plus,
working for my village, and when the community asks me for support,
believing in me again creates satisfaction.
KII, Agrarian HEW
Burnout, a combination of emotional exhaustion, depersonalization, and low

personal accomplishment caused by the chronic stress of medical practice, is
increasingly recognized as an important factor affecting the effectiveness of
health professionals.48 Some studies have reported that health worker burnout
is associated with increased medical errors, lower patient satisfaction, and
decreased performance. Burnout in this assessment was measured by an
instrument consisting of 15 items and asking HEWs to give a rating for each
item using 4 response options: 1, “Rarely,” 2, “Sometimes,” 3, “Often,” and 4, “Very
often.” Risk of burnout was analyzed by categorizing the total score of the scale
into 4 ranges: 15 to 18, “no risk of burnout,” 19 to 22, “little sign of burnout,” 33
to 49, “risk of burnout,” and 50 to 79, “severe risk of burnout.”
In the current assessment, a majority of the HEWs (75%) either showed little
sign of burnout or were at risk of burnout. A small but meaningful percentage
(4%) had a severe risk of burnout. A high proportion of HEWs from Tigray
(97%) and Amhara (96%) had either little sign of burnout, risk of burnout, or
severe risk of burnout, while only 12% of HEWs from Somali showed similar
burnout symptoms (either little sign of burnout, risk of burnout, or severe risk of
burnout). HEWs working in pastoralist communities were less likely to manifest
symptoms of burnout than those working in agrarian communities. HEWs aged
18 to 24 had a lower risk of burnout than did other age categories (Table 3-12).
page-110
Table 3-12. HEWs burnout status by background characteristics
HEWs burnout status (%)

No risk of Little sign of Risk of Severe risk of Unweighted # of
Characteristics
burnout burnout burnout burnout HEWs
National 15.5 44.7 35.1 4.8 584
Age category
in years
18-24 34.9 41.8 22.2 1.1 189
25-29 19.8 38.1 37.0 5.1 272
30+ 19.5 35.8 39.8 4.9 123
Marital status
Never married 34.6 35.2 27.7 2.5 159
Married 21.8 39.8 34.3 4.3 399
Others 7.7 46.2 42.3 3.9 26
Livelihood
Agrarian 15.2 41.0 39.8 4.1 414
Pastoralist 47.7 33.5 15.9 2.9 170
Education level
Below level III 28.7 46.0 22.6 2.9 24
Level III 10.0 46.7 37.2 6.1 321
Level IV 15.9 38.4 42.2 3.5 239
Region
Tigray 3.2 39.7 50.8 6.4 63
Afar 36.8 31.1 42.1 0.0 19
Amhara 4.2 41.1 48.4 6.3 95
Oromia 15.5 52.9 27.6 4.1 123
Somali 88.0 8.0 1.3 2.7 75
Benishangul-
46.0 43.2 10.8 0.0 37
Gumuz
SNNPR 10.4 53.1 32.3 4.2 96
Gambela 18.6 14.0 65.1 2.3 42
Harari 32.4 44.1 23.5 0.0 34
Abbreviations: HEW, Health Extension Worker; SNNPR, Southern Nations, Nationalities, and Peoples
Region.
page- 111
3.1.13 Intention to leave and attrition of HEWs
Intention to leave one’s job was evaluated by asking HEWs how many more
years they intended to work as HEWs. Accordingly, 21% of HEWs reported
that they intended to work as an HEW for 2 more years, and 11% reported
that they intended to work for 3 to 5 more years. In addition, when asked
whether they are currently looking for another job, 17% of HEWs reported that
they were currently looking for another job, which might be a sign of intention
to leave their current job. Only 1 in 4 HEWs reported that they intended to
work as HEWs for life Table 3-13. As shown in Figure 3-5, the most frequently
mentioned reasons for intending to leave one’s job were the low salary (31%),
lack of career development opportunities (25%), the desire to live in an urban
area (18%), and workload (15%).
Table 3-13. HEWs’ intention to leave their job, by region and livelihood
Intention to leave job among HEWs (%)
Unweighted #
Background
of HEWs
Intention to
Intention to
Intention to
another job
intention to
leave after
looking for
Undecided
to 5 years
leave in 2
leave in 3
Currently
leave job
5 years
years
No
National 21.3 11.0 5.1 20.1 42.6 17.1 584

Livelihood
Agrarian 22.8 11.4 2.6 19.1 44.2 17.6 414
Pastoralist 6.1 7.2 28.9 30.1 27.7 10.8 170
Region
Tigray 14.1 16.9 2.0 53.6 13.5 1.9 63
Afar 23.7 7.9 0.0 6.5 61.9 24.8 19
Amhara 42.2 9.7 1.1 7.2 39.7 42.8 95
Oromia 18.4 13.0 4.7 19.8 44.1 13.1 123
Somali 4.7 6.1 43.9 35.4 10.0 2.4 75
Benishangul-
5.8 8.8 10.7 46.0 28.8 8.0 37
Gumuz
SNNPR 14.7 8.9 0.0 22.0 54.4 7.2 96
Gambela 0.0 2.8 0.0 21.2 76.0 45.0 42
Harari 19.3 39.8 2.2 23.1 15.6 19.3 34
Abbreviations: HEW, Health Extension Worker; SNNPR, Southern Nations, Nationalities,

and Peoples Region.
page-112
Figure 3-5. Reasons for intending to leave one’s job among HEWs affirming
that they had a plan to change their jobs
HEWs who participated in the qualitative study also mentioned their intention
to quit their jobs because of the inadequate salary and benefits. In general,
HEWs’ intention to leave their jobs is influenced by the following factors: a
lack of educational opportunity, limited chance of promotion, limited transfer
opportunities, demands for false reports, judgmental appraisal, denial of
legitimate leave, unequal treatment, disrespect from officials, low salary, the
lack of benefits and incentive packages, the lack of skills, beliefs and attitudes,
personal conflicts, personal life, being assigned in a location other than one’s
birthplace, distance to the work site, lack of transportation, difficult topography,
a lack of trust and respect from the community, and the HEP’s and/or HEWs’
not being valued by the community.
Attrition among HEWs was defined in this study as ceasing to provide the
HEP service for any reason, including but not limited to resignation, dismissal,
disappearance, death, retirement, change in position, and transfer elsewhere.
HEWs’ personnel data were obtained from 77 (90.6%) of the 85 targeted
woredas. A total of 2828 personnel documents of HEWs from these 77 woredas
were reviewed to assess their rate of attrition. The majority (2 746, 93.8%),
of the HEWs were female, and 1623 (56.7%) were aged 20 to 24 years. The
magnitude of the attrition rate of HEWs from 2004 to 2019 was estimated at
21.1%, and significant regional variation was observed. The highest attrition was
page- 113
seen in Afar (36.1%), while the lowest attrition rate was seen in Oromia (15%).
Of those who left their jobs, the reason identified in their personnel file was
disappearing from work without any clue (41%), resignation (25.4%), change in
qualification (13.5%), transfer out (8.5%), dismissal (7.2%), and death (4.6%).
The median time to attrition was 6.1 years, with an inter-quartile range of 3.4
to 8.2 years, indicating that 50% of HEWs served 6 years before leaving their
work. The overall median time of service, however, was 7.9 years (IQR=3.4-
11.5). The overall incidence rate of attrition was 251.4 per 10 000 person-years
(95% CI, 231.0-273.6). The qualitative study depicts 2 major themes as causes
of attrition: personal causes, such as lack of educational opportunity, and
personal conflicts, and work-related causes, which includes denial of annual
leave, unfair wages, judgmental appraisal, poor support habits, and difficult
topography.
A complete report of the study on HEWs’ attrition rate, trends in attrition, and
trends for leaving the HEP is presented in Part 4.
3.1.14 Mental health status of HEWs
The HEWs’ mental health status was assessed using the patient health
questionnaire (PHQ-9). This is a 9-item instrument previously validated in
Ethiopia that has been found to have good sensitivity for screening depression.49
The 9 items measure the level of depression on scale, where 0 means “not at
all” and 3 means “nearly every day.” In this assessment, the instrument has an
acceptable level of internal consistency (Cronbach’s alpha=0.88).
Exploratory factor analysis showed that all 9 items load on 1 dimension and
have an eigenvalue ranging from 0.36 to 4.7. The instrument has a maximum
score of 27, and a score of 10 or above is considered a cut-off point for probable
symptoms of depression, in line with previous studies.50
The findings related to the mental health status of HEWs are summarized in
Table 3-14. The prevalence of depression among HEWs in the current study was
17%, with significant regional variability. For instance, the prevalence was 37%
in Gambela, 30% in Amhara, and 21% in Tigray, but 5% in the SNNPR, 8%
in Somali, and 9% in Benishangul-Gumuz. The probability of having elevated
depressive symptoms also differs by livelihood. HEWs in agrarian communities
page-114
are more likely to manifest probable symptoms of depression than pastoralist

HEWs. HEWs aged 25 to 29 years had a higher percentage of depression than
HEWs aged 18 to 24 years. Similarly, HEWs with 11 to 16 years’ experience are
more likely to have elevated symptoms of depression than HEWs with 1 to 5
years’ experience. The probability of manifesting symptoms of depression did
not differ by the HEWs’ level of qualification or place of residence (Table 3-14).
page- 115
Table 3-14. Prevalence of probable symptom of depression, by background

characteristics
Depression status of HEWs Unweighted # of

Background
No depression Probable symptom HEWs
National 83.5 16.6 584
Age category in years
18-24 89.0 11.0 189
25-29 80.8 19.2 272
30 and above 84.6 15.4 123
Marital status
Never married 84.2 15.8 159
Married 83.5 16.5 399
Divorced/widowed/
77.1 22.9 26
separated
Qualification
Below level III 83.4 16.6 24
Level III 83.7 16.4 321
Level IV 83.3 16.7 239
Work experience
1-5 years 90.3 9.7 254
6-10 years 82.0 18.0 158
11-16 years 79.9 20.1 172
Residence
Same kebele as HP 84.2 15.8 394
Other kebele 82.1 17.9 190
Livelihood
Agrarian 82.4 17.6 414
Region
Tigray 79.0 21.0 63
Afar 80.1 19.9 19
Amhara 69.7 30.3 95
Oromia 83.1 16.9 123
Somali 92.0 8.0 75
Benishangul-Gumuz 91.5 8.5 37
SNNPR 94.9 5.1 96
Gambela 63.4 36.7 42
Harari 89.6 10.4 34
page-116
In the qualitative study, HEWs shared that exposure to stressful conditions

at work, prolonged periods of workload and other reasons led to mental
“
disturbances.
The concern I have is, even if the program is good for the community,
it is creating a problem for the HEWs. Because we are not getting
transfers and there is a shortage of HEWs, we are getting fed up
with the community. Even doing the same thing in the same place
for a long time is not good for your mental health. The community
says, “You are still here, why don’t you get a transfer like other
professionals?” Those things are difficult for HEWs.
KII, Agrarian HEW
This HEW’s complaint is supported by other key informants. They revealed

their stressful working conditions that result in mental distress. They identified
the causes as competing family responsibilities as a woman and risky working
conditions.
“
In describing the situation, an HC head from an agrarian area said:
HEWs cannot improve themselves or transfer to other posts. There

is no professional development, and their salary is low and doesn’t
have a benefit package. Additionally, their jobs are risky; since they
are women, they face many problems. For example, in our cluster, 3
HEWs are sick or absent; 1 has a nerve problem, 1 has mental illness,
and the other has cancer. The sector did not provide any support for
them; therefore, the other HEWs are discouraged since the institution
does not have ways of assisting in cases like this.
KII, Agrarian HC Head
page- 117
3.1.15 Self-reported performance of HEWs
HEWs were asked to rate their performance in their day-to-day activities;

the findings are summarized in Table 3-15. The great majority of HEWs gave
themselves the highest score for their performance. To this point, 89% of
HEWs reported that they were highly productive in their day-to-day activities.
Moreover,93% of HEWs rated their leadership ability as high or very high, and
87% ratedtheir colleagues’ performance (i.e., productivity) as high or very high.
In relativeterms, a lower proportion of HEWs (75%) reported the timely
accomplishmentof their daily activities. A great majority of HEWs (90%)
highlighted their needfor additional staff in order to get through their current
workload.
Table 3-15. Community members’ involvement in the performance assessment

of HEWs
Livelihood (%)
Items National
Agrarian Pastoralist
I am highly productive in my day
88.8 88.2 88.8
to day activity
Feel I am as productive as I
89.1 85.7 88.8
could/should be
My colleagues are productive on
86.4 91.3 86.9
a typical day
I feel my colleagues are as
82.7 89.8 83.4
productive as they should be
I complete my “to do” list by the
74.3 85.9 75.3
end of each day
I am effective at leading staff for
94.3 89.9 93.9
high performance
I am energetic at the end of
85.4 82.8 85.2
each day
My team members are energetic 81.6 85.1 81.1
My team members are positive 91.5 89.3 91.3
page-118
3.1.16 Performance appraisal
In more than 76% of HPs, performance appraisal was conducted for all HEWs
working in the HP. In 6% of HPs, performance appraisal was performed only
for the head of the HP, and in 2% of HPs, only HEWs other than the head
were appraised (Table 3-16). More than half of performance appraisals were
conducted by the PHCU director/head, followed by the HEP supervisor from
the catchment HC.
Written feedback for the most recent performance assessment was available
in only 27% of HPs. Although the document could not be verified by the data
collectors, HEWs in 18% of the HPs claimed that there was written feedback.
In more than half (55%) of the HPs, there was neither written feedback nor a
performance appraisal (Table 3-17).
The performance appraisal processes and procedures were also explored by

the qualitative study. The information gathered from key informants shows that
the schedule for performance appraisal varies from place to place: various HPs
conducted performance appraisals weekly, biweekly, monthly, or every 4 months.
The criteria used to appraise the performance of HEWs also differ by woreda.
Some of the indicators used for performance appraisal include: creating a
model kebele/family, presence in the HP during working hours, achievement at
both HP and the kebele levels, effectiveness, and report quality. Quality of the
healthcare service and the community’s level of service use are also considered
in the appraisal of HEWs in some areas.
Key informants also highlighted that written feedback on how to improve

future performance is given based on the strengths and weaknesses identified
during the performance assessment and appraisal. In some cases, rewards
including promotion and encouragement for well-performing HEWs and giving
constructive feedback for poor-performing ones. Warnings and punitive actions,
including dismissal from posts, are also practiced.
page- 119
Table 3-16. Performance appraisal of HEWs, by region and livelihood
Appraisal Appraisal only Appraisal for

No appraisal Unweighted #
Background for all HEWs for the head HEWs who are
(%) of HEWs
(%) of the HP(%) not head(%)
National 76.2 6.4 1.5 16.0 584
Livelihood
Agrarian 80.4 6.3 1.3 12.0 414
Pastoralist 35.5 6.5 3.2 54.8 170
Region
Tigray 100 0.0 0.0 0.0 63
Afar 40.0 14.9 0.0 45.1 19
Amhara 94.5 2.8 0.0 2.7 95
Oromia 76.2 8.4 0.7 14.7 123
Somali 12.9 6.6 5.6 74.9 75
Benishangul-
85.6 0.0 0.0 14.4 37
Gumuz
SNNPR 72.4 6.9 3.4 17.4 96
Gambela 26.0 4.7 7.1 62.2 42
Harari 95.2 0.0 0.0 4.8 34
page-120
Table 3-17. Availability of written feedback for the most recent performance
assessment
Written
Written feedback Written
feedback
available and feedback not Unweighted #
Background available but
verified (%) available (%) of HEWs
not verified (%)
National 27.1 17.7 55.2 584
Livelihood
Agrarian 28.9 18.3 52.8 414
Pastoralist 9.4 11.5 79.1 170
Regions
Tigray 35.1 36.5 28.4 63
Afar 7.4 17.2 75.4 19
Amhara 47.7 19.3 33.0 95
Oromia 31.0 15.5 53.5 123
Somali 4.9 18.3 76.9 75
Benishangul-
40.4 17.8 41.8 37
Gumuz
SNNPR 6.5 17.6 75.8 96
Gambela 11.9 16.6 71.6 42
Harari 40.5 13.5 46.0 34
Region.
Although the HEP implementation guidelines recommend that community

members participate in the performance assessment of HEWs, only a few HPs
put this into practice. Gambela and Benishangul-Gumuz have higher levels of
community participation in HEWs’ performance assessment, while the SNNPR
and Tigray have the lowest levels. Community participation is relatively higher
in pastoralist communities than in agrarian communities (Table 3-18).
page- 121
Table 3-18. Community members’ involvement in the performance assessment

of HEWs
Involved in Involved in
Not involved Unweighted #
assessment of head assessment of all
Background (%) of HEWs
of HEP (%) of HEWs (%)
National 3.5 16.1 80.4 584
Livelihood
Agrarian 3.6 15.4 81.0 414
Pastoralist 1.3 29.3 69.4 170
Region
Tigray 0.0 15.3 84.7 63
Afar 0.0 30.1 69.9 19
Amhara 4.0 22.8 73.2 95
Oromia 5.5 12.4 82.1 123
Somali 4.0 42.0 53.9 75
Benishangul-
0.0 53.1 46.9 37
Gumuz
SNNPR 0.0 11.1 88.9 96
Gambela 12.5 62.7 24.9 42
Harari 0.0 33.3 66.7 34
Abbreviations: HEW, Health Extension Worker; HEP, Health Extension Program; SNNPR, Southern
3.2 Facilities and infrastructure

3.2.1 Introduction
Since its introduction in 2003, the HEP managed to construct and make
functional 17 162 HPs in rural areas across all the regional states.51 The standard
for HPs is that HPs shall have a minimum of the following: premises including
a waiting area, examination, counseling, and injection room, delivery and
postnatal room, store room, toilet, and an area for a placenta pit. Moreover,
the building should meet the following minimum requirements: the floor of the
HP shall be smooth, with no cracks, and easily washed, the walls and ceilings
shall be constructed with easily cleaned materials and adequate reflectance,
page-122
and all rooms shall have access to natural light and ventilation. In addition,
the standard recommends that HPs have the necessary equipment, including
stethoscope, sphygmomanometer, thermometer, kidney basin, delivery set,
delivery table, examination couch, storage shelves, medicine and supplies, cold
box, adult and child weighing scale, autoclave, and stretcher.52
Studies conducted at different time periods during the history of the HEP have
indicated, however, that most of the HPs fail to fulfill the requirements. This
section presents the availability and accessibility of HPs, the availability of
basic amenities, and basic equipment for the implementation of the HEP.
Health posts are universally available; almost all kebeles have at least 1 HP.
Less than half of the HPs, however, meet the standard for an HP-to-population
ratio of 1 HP for every 5 000 people. At the national level, the HP-to-population
ratio is 1:5 760, and the Health Extension Worker-to-population ratio is 1:2 599.
A majority of HPs have access to all-weather roads connecting them to the
referral or supervising health facility. The physical structure of the buildings of
a majority of HPs did not meet the required minimum standards. A majority
of HPs also lack basic utilities, such as electricity and improved water source.
There was a shortage of basic equipment in the HPs. Only 5% of HPs had all
basic tracer equipment (blood pressure apparatus, stethoscope, thermometer,
adult scale, child scale, and artificial light sources). The shortage was more
pronounced in pastoralist HPs.
3.2.3 Accessibility of HPs
HPs located in agrarian areas are expected to serve up to 5 000 people (1 000
households), and in pastoralist areas they are expected to serve up to 3 000
people to compensate for the low density and mobility of the population. The
current assessment indicates that, on average, 1 HP is serving 5 760 people. The
number of people served by HPs ranges from 2 120 in Benishangul-Gumuz to
6 498 in Oromia. In pastoralist communities, 1 HP serves on average a smaller
population (2 919 people), than an agrarian HP, which serves on average 6 057
page- 123
people. The HEW-to-population ratio is 1:2 599 nationally; HEWs in agrarian

regions serve an average population of 2 728, and those in pastoralist areas
serve an average of 1 361 people. Less than half (42%) of HPs meet the standard
for the population size they serve (i.e., fewer than 5 000 people; Table 3-19).
Table 3-19. HP and HEWs to population ratio, by region and livelihood
% HPs that % HPs that

Health
HEW to have 5000 have 3000
post to Unweighted # of
population or less or less
Background population HPs
ratio catchment catchment
ratio
population population
National 5760 2599 42.1 17.7 343
Livelihood
Agrarian 6057 2728 36.8 13.9 235
Pastoralist 2919 1361 93.3 53.9 108
Region
Tigray 5941 2875 36.9 12.0 32
Afar 3779 1256 82.5 57.4 18
Amhara 5963 2834 25.0 8.1 60
Oromia 6498 3110 33.4 6.7 74
Somali 2526 1268 100 58.5 43
B/Gumuz 2120 935 95.4 82.2 24
SNNPR 5361 1966 53.9 29.8 59
Gambela 3718 424 91.6 81.5 17
Harari 4576 1509 58.7 17.5 16
Abbreviations: HP, health post; HEW, Health Extension Worker; SNNPR, Southern Nations,
ROAD ACCESS
The absence of roads connecting HPs to their referral HCs impedes the smooth
functioning of the HPs. Without a road, it is difficult to establish a functional
referral link with the nearest HC, or make timely requests and receipt of drugs
and medical supplies. Per this study, 59% of HPs have an all-weather road
connecting them to the nearest referral HC. A substantial number of HPs
(35%), however, have access to roads that function only during dry seasons,
page-124
and 6% of HPs have no access to a paved road connecting them to the referral
HCs. There is an enormous disparity regarding access to all-weather roads
among the regions and between agrarian and pastoralist communities. Most
HPs (91%) located in Harari have access to an all-weather road, while more
than half of HPs in Tigray, Afar, the SNNPR, and Gambela have no access
to an all-weather road. Access to an all-weather road is slightly better among
HPs located in pastoralist communities than those in agrarian areas (62% vs.
58%; Table 3-20. Some participants in community FGDs stated that HPs are
located far from the community, as evidenced by their long distance to the HPs.
Poor road conditions and a shortage of road transportation forces HEWs to
walk long distances to provide outreach service to some communities. This in
turn affects the use of services at the HP. Moreover, because of the absence of
roads, laboring women cannot get ambulance services to reach the HP or HC.
By contrast, some community members stated that their HPs were accessible
to all community members.
Table 3-20. Road connectivity and accessibility of HPs, by region and livelihood
Percentage of HPs with road connecting to HC

Unweighted # of
Background All-weather road Dry weather road No paved road HPs
National 58.5 35.3 6.2 343
Livelihood
Agrarian 58.1 35.6 6.3 235
Pastoralist 62.2 32.6 5.2 108
Region
Tigray 41.9 41.1 17.1 32
Afar 41.2 58.9 0.0 18
Amhara 70.1 23.5 6.3 60
Oromia 60.1 37.1 2.9 74
Somali 65.0 25.9 9.2 43
Benishangul-
52.3 40.2 7.5 24
Gumuz
SNNPR 47.4 42.5 10.1 59
Gambela 47.2 52.8 0.0 17
Harari 91.3 8.7 0.0 16
Abbreviations: HP, health post; HC, health center; SNNPR, Southern Nations, Nationalities, and Peo-
ples Region.
page- 125
TRANSPORTATION FOR HEWS
In the current assessment, HEWs were asked about the means of transportation
they mostly used to provide outreach services. The findings indicated that
the great majority of HPs have no means of transportation for outreach
activities, such as home visits and vaccination services. Only 7% of HPs use
public transportation to provide outreach services, and 3% use bicycles or
motorbikes. In most HPs (89%), HEWs provide outreach services by walking.
Key informants from the MoH stated that a great number of bicycles were
distributed throughout the country, but the findings from the study indicate
that HEWs are not using them. The availability of transportation varies across
regions but not between agrarian and pastoralist areas (Table 3-21). In the
qualitative study, HEW noted that transportation services were limited and
“
inadequate in reaching the community:
The work is really tough; there is no transportation access, you go to

home visits alone, and it requires extensive physical energy besides
mind work. When you walk long distances each day, you face kidney
problems, and nowadays most HEWs are getting sick. Since the work
is really hard, we walk most of the time.
KII, Agrarian HEW
page-126
Table 3-21. Means of transportation used to conduct outreach activities
Means of transportation to do outreach service (%)

Unweighted # of
Background Bicycle/ Public
Walking HPs
Motorbike transportation
National 88.8 7.4 3.8 343
Livelihood
Agrarian 88.9 8.0 3.1 235
Pastoralist 88.2 1.9 9.9 108
Region
Tigray 97.0 0.0 3.0 32
Afar 67.1 28.3 4.6 18
Amhara 94.5 0.0 5.5 60
Oromia 87.8 8.8 3.5 74
Somali 84.2 0.0 15.8 43
Benishangul-
100 0.0 0.0 24
Gumuz
SNNPR 86.2 13.6 0.23 59
Gambela 62.3 23.5 14.1 17
Harari 86.5 0.0 13.5 16
Abbreviations: HP, health post; SNNPR, Southern Nations, Nationalities, and Peoples Region.
3.2.4 Physical structure of HPs
According to the standard, HPs should have premises with a minimum area of
300m2 with a dedicated and marked entrance and easily accessible by
persons with disabilities. The HP should preferably be provided with road access,
a water supply, a power grid connection, and communication facilities. The HP
should also have a minimum of the following: a waiting area, an examination/
counseling/injection room, a delivery and postnatal room, a storeroom, a toilet,
and an area for the placenta pit.
The findings from the current assessment showed that about 87% of HPs have
a stand-alone building, and 13% have a building shared with other government
institutions (e.g., kebele administration office or Agriculture Office). Only 43% of
HPs have their own fenced compound; 28% have a partially fenced compound,
page- 127
and the rest do not have an established compound, which could compromise
the quality of service delivery and the prevention of infection. The availability
of a standalone building ranged from 78% in Afar to 96% in Tigray. Afar
has the highest proportion of HPs with their own fenced compounds (86%),
followed by Gambela (76.5%) and Harari (72%). The Tigray (35%), Oromia
(36%) and Amhara (37%) regions have the lowest proportion of HPs with their
own fenced compounds. The absence of standalone buildings affect the privacy
of clients (i.e., in the provision of FP or ANC), as stated by key informants and
community members.
The HEP implementation guidelines recommend that HPs should have at least
3 rooms (1 outpatient department, 1 for delivery, and 1 for storage).The majority
(81%) of the HPs meet the minimum standard number of rooms (3 or more);
a sizable proportion (19%) of HPs, however, have fewer than 3 rooms, which
potentially compromises the quality of healthcare services. The key informants
frequently mentioned the inconvenience of HP buildings for receiving healthcare
services from HPs. On the one hand, the rooms are narrow to accommodate
clients, and on the other hand, the poor quality of the buildings (made of mud
and wood) are not inviting to clients. The assessment also indicated substantial
variation among the regions. All HPs in Afar met the minimum standard number
of rooms (3 or more), but only 65% of the HPs in Somali did. More HPs in
agrarian areas met the minimum standard number of rooms than did those
in pastoralist areas (82% vs. 72%). The assessment also indicated that 41% of
HPs (42% in agrarian and 40% in pastoralist areas) had a room dedicated to
delivery service, with substantial variation across the regions (Table 3-22).
FGD participants and key informants affirmed that service delivery in HPs was
compromised due to the shortage of rooms and the poor quality of the HPs’
physical structures. Women emphasized their lack of privacy while receiving
services like FP, ANC, and delivery. Moreover, HEWs stated that ensuring the
adequacy of rooms at HPs was vital to providing services for communicable
diseases, such as TB.
An HEW from an agrarian area described the problem she faced in providing
services as follows:
page-128
“
I have worries regarding rooms where we provide medicine for TB
patients because the room does not have windows. The ventilation
is poor. I am afraid of contracting the disease. The rooms where we
work do not have windows. So, we are working in fear in these rooms.
KII, Agrarian HEW
Table 3-22. Availability of basic facilities and infrastructure at health posts
Proportion of health posts:

Background Standalone Fenced Having at Having a Unweighted
building1 compound2 least 3 or delivery # HPs
more rooms room
National 87.1 43.2 81.0 41.3 343
Livelihood
Agrarian 87.3 43.1 81.9 41.5 235
Pastoralist 85.1 44.8 71.6 39.5 108
Region
Tigray 96.4 35.3 68.9 51.8 32
Afar 77.7 86.4 100 25.3 18
Amhara 87.8 37.1 69.0 51.5 60
Oromia 88.3 36.0 87.5 31.9 74
Somali 82.5 51.5 64.7 60.2 43
Benishangul- 93.2 65.1 82.2 56.9 24
Gumuz
SNNPR 83.9 56.8 84.7 43.2 59
Gambela 95.3 76.5 71.8 30.7 17
Harari 91.3 72.2 68.3 0.0 16
Abbreviation: HP, health post; SNNPR, Southern Nations, Nationalities, and Peoples Region.
1
A stand-alone facility is defined as one having a separate building dedicated to the HP.
2
A separate compound is defined as a clearly delineated/fenced compound.
page- 129
BUILDING STRUCTURE OF HEALTH POSTS
According to the standard for HPs, the floor of the HP should be washable,
smooth, and non-absorbent, and the walls and ceilings should be constructed
with easily cleaned materials and with adequate reflectance.52 The assessment
disclosed, however, that only 38% of HPs had walls made up of block or brick;
only 5% of the HPs were made of stone and cement, according to the required
standard. The rest of the HPs did not fulfill the building standard; half of HPs
had walls made of wood and mud. Most HPs (95%) have roofs made up of
corrugated iron sheets, but about 4% of HPs had a roof covered with grass or
plastic materials. Most HPs (86%) had a cement floor, and about 5% of HPs
had a floor covered with ceramic. Nearly 1 in 10 HPs had a floor covered with
dust with no cement cover (Table 3-23).
Table 3-23. Building materials used to construct health post, by livelihood

(N=343)
Building Materials Agrarian (%) Pastoralist (%) Total (%)

Wall
Block 21.8 64.1 27.4
Wood 54.4 19.7 49.8
Bricks 11.9 8.1 11.4
Stone 5.3 4.7 5.2
Others 1
6.6 3.4 6.2
Roof
Corrugated iron
97.6 81.0 95.4
sheets
Grass 0.7 16.1 2.8
Plastic 1.2 0.0 1.0
Others 2
0.6 2.8 0.9
Floor
Dust/Soil 9.7 2.8 8.8
Cement 84.4 92.8 85.6
Ceramic 5.1 1.6 4.6
Other 3
0.8 2.8 1.1
page-130
The proportion of HPs meeting the building standards is calculated using the
data collected in the current survey. An HP was considered to meet the mini-
mum standard when: the wall of the building is made of brick, block, or stone
with cement, the roof is covered with iron sheeting, and the floor is covered
with cement or ceramic. Accordingly, only 37% of HPs meet the building stan-
dard. Fulfillment of the building standard varies across regions and livelihoods.
Amhara has the lowest proportion of HPs fulfilling the standard (26%), while
all HPs in Afar and Tigray meet the building standard. Surprisingly, a higher
proportion of HPs in pastoralist areas meet the building standard than do HPs
in agrarian areas (50% and 36%, respectively; Table 3-24).
Key informants also mentioned the problem of poorly constructed HPs and the
challenges HEWs face in providing curative services in such a working environ-
ment. In describing the situation, a key informant from an RHB said that:
“
Some of the HPs [smiles] have no walls. They lock the door, but
have no walls….Many others don’t have a fence, have no toilet….They
teach about latrine construction, but the HP does not have a latrine
[smiles]. It has no furniture, when it rains, it leaks and destroys all
their files. They don’t have a budget. When we ask HEWs to send
reports, they don’t have pens and papers. No enabling environment….
The HP building has to be attractive not only to the professionals but
also to the customers.
KII, RHB
According to this participant, the community constructs the HPs with locally
available building materials:
“
Health posts are constructed by community contributions, which
could collapse within a year due to their poor quality. You see many
health posts collapsed in different kebeles. Some HPs constructed by
the government also have a quality problem.
KII, WorHO Process Owner
page- 131
Table 3-24. Health posts meeting the building standards
Percentage of health posts having:

A floor
Wall made of A roof Fulfilled the Unweighted #
Background covered with
brick/block or covered with building of HPs
cement /
stone iron sheet standard
ceramic
National 41.0 94.9 90.7 37.1 343
Livelihood
Agrarian 36.8 97.9 90.1 35.9 235
Pastoralist 82.7 65.9 97.0 49.5 108
Region
Tigray 100 100 100 100 32
Afar 100 100 100 100 18
Amhara 27.7 100 85.8 25.9 60
Oromia 36.8 99.8 90.3 36.8 74
Somali 95.1 42.0 100 37.2 43
Benishangul-
69.0 86.3 79.2 58.7 24
Gumuz
SNNPR 36.5 94.6 93.1 36.5 59
Gambela 81.2 40.1 85.9 35.4 17
Harari 100 100 100 100 16
3.2.5 Availability of incinerators, placenta pits, and open pits for

solid waste disposal
The HP building standards, national infection prevention guidelines, and HEP

implementation guidelines all recommend having an incinerator and placenta
pit in any health facility, including HPs. The current study finds that only 12% of
HPs have incinerators, 7% have placenta pits, and 74% have open pits for solid
waste disposal. The finding differs across regions. For example, an incinerator is
available in half (51%) of HPs located in Afar, while none of the HPs in Harari
have one (Table 3-25).
page-132
Table 3-25. Availability of incinerators, placenta pit, and open pit, by region
and livelihood
Proportion of health posts having:

Unweighted
Background Open pit for solid
Incinerator Placenta pit # of HPs
waste disposal
National 12.3 7.4 73.8 343
Livelihood
Agrarian 11.4 8.0 77.7 235
Pastoralist 21.9 2.0 35.8 108
Region
Tigray 46.4 39.9 80.2 32
Afar 50.8 10.3 34.0 18
Amhara 9.1 8.5 68.1 60
Oromia 10.3 0.1 73.7 74
Somali 11.4 1.5 30.7 43
Benishangul-
14.8 12.6 86.8 24
Gumuz
SNNPR 13.0 15.7 88.1 59
Gambela 35.5 18.8 49.7 17
Harari 0.0 0.0 85.7 16
3.2.6 Availability of basic utilities
The availability of electricity was assessed in 2 ways: (a) by assessing the

HPs’ connectivity with the national electric grid line and (b) by assessing the
availability of regular electricity. Regular electricity is considered available at
a facility if 1 of the following conditions is met: (a) the facility is connected to
a central power grid, and the power supply was not interrupted for more than
2 hours at a time during normal working hours in the 7 days before the survey,
(b) the facility had a functioning generator with fuel available on the day of
the survey, or (c) the facility has backup solar power.
In this regard, only 27% of HPs have an electric power source (9% from grid
power and 17% from solar), and only 18% of HPs have regular electricity. There
page- 133
is no significant difference in the availability of regular electricity between

agrarian HPs and pastoralist HPs (17% vs. 21%). There is also a regional
disparity regarding the connectivity of HPs to the national grid line and the
availability of regular electricity: Harari had the highest proportion of HPs with
regular electricity, and Oromia had the lowest.
Overall, 27% of HPs have an improved water source in the facility. The main
water source is the public tap. The proportion of HPs with an improved water
source varies by region and livelihood. The availability of an improved water
source was found to be highest (62%) in Gambela, followed by Somali (60%)
and Benishangul-Gumuz (57%), and lowest in Harari (18%) and Tigray (19%).
HPs located in agrarian communities were also found to be less likely to have
an improved water source than pastoralist HPs (25% vs. 48%).
Most HPs (87%) have a functioning latrine for clients. More HPs in agrarian
areas have a latrine than do HPs in pastoralist areas (91% vs. 78%). HPs
located in Afar and Gambela have higher latrine coverage (94%), followed by
Tigray, Amhara, and Oromia.
A majority of HPs (81%) have consultation rooms (3 or more rooms).

Communication equipment, however, was absent from most HPs, with the
exception of Tigray, where 94% of HPs have communication equipment. Only
1.2% of HPs have all basic amenities (Table 3-26).
page-134
Table 3-26. Availability of basic utilities in the HPs
Improved water
Communication
Unweighted #
Power source1
Consultation
Background
Equipment5
Sanitation
electricity2
amenities
facilities4
All basic
Regular
source3
of HPs
rooms
National 26.5 17.7 27.1 87.4 5.7 80.9 1.2 343
Livelihood
Agrarian 26.4 17.3 25.0 89.3 6.2 81.9 1.3 235
Pastoralist 27.2 21.1 47.7 69.2 0.9 71.0 0.0 108
Region
Tigray 43.8 37.5 18.6 93.1 93.5 68.9 9.5 32
Afar 46.2 22.3 38.6 97.2 4.6 100 0.0 18
Amhara 32.2 22.1 28.1 85.4 0.0 69.0 0.0 60
Oromia 22.5 13.0 21.4 86.8 1.7 87.3 0.0 74
Somali 26.8 20.3 59.7 66.4 0.0 64.7 0.0 43
Benishangul-
49.5 46.1 56.9 82.2 15.5 82.2 2.3 24
Gumuz
SNNPR 24.0 16.8 27.4 93.9 7.0 84.7 3.4 59
Gambela 38.6 28.3 61.5 94.9 5.1 69.4 0.0 17
Harari 64.3 64.3 18.3 100 49.2 68.3 0.0 16
1
Power source: Facility is connected to a central power grid, facility has a functioning generator or
invertor with fuel available on the day of the survey, or facility has backup solar power.
2
Regular electricity: There has not been an interruption in power supply lasting for more than 2 hours
at a time during normal working hours in the 7 days before the survey.
3
Improved water source: Water is piped into facility or onto facility grounds; or water is from a public
tap or standpipe, a tube well or borehole, a protected dug well, or protected spring, and the outlet
from this source is within 500 m of the facility.
4
Sanitation facilities: The facility has a functioning flush or pour-flush toilet, ventilated improved pit
latrine, pit latrine with slab, or composting toilet.
5
Communication equipment: The facility had a functioning land-line telephone, a functioning facili-
ty-owned cellular phone or wireless telephone, a private cellular phone supported by the facility, or a
functioning short-wave radio available in the facility.
page- 135
Mean availability was calculated by adding 5 items (regular electricity, improved

water source, sanitation facility, room for consultation, and communication
equipment). Findings showed that the mean availability of basic amenities was
2.2 out of 5 (Figure 3-6).
Figure 3-6. Mean availability of basic amenities, by region
In line with the quantitative findings, some key informants noted that HPs
primarily lack electricity, a water supply, and toilets. In some places, public taps
had been non functional for a number of years. Communities are worried about
“
TB transmission due to the crowded and small rooms at the HPs.
Health posts don’t have fence, no toilet… they teach about latrine
construction, but the HP does not have a latrine [smiled] no furniture,
when rains it leaks and destroy all their files, they don’t have budget,
when we ask them to send report they don’t have pen and papers.
KII, RHB
Though the HP construction design recommends including clinical services and

storage rooms, drugs were stored in the same rooms as other materials in the
studied HPs. Due to the lack of electric power, refrigerators are not functioning
well. As a result, transporting medicine from elsewhere by motorbike exposes
the medicines to direct sunlight. In some HPs, delivery services are also handled
using handheld torches.
page-136
“
We have a serious problem with electricity during delivery. The
handheld battery doesn’t last long enough. The solar battery doesn’t
charge fully during rainy seasons. There was a moment in which a
newborn fell down on the ground due to poor light.
FGD, WDA
3.2.7 Availability and functionality of basic equipment
The overall availability of medical equipment in HPs is very low. For example, only
39% of HPs (37% in agrarian communities and 51% in pastoralist communities) had
a refrigerator, although all HPs are expected to provide EPI services. The highest
availability was observed for vaccine carriers (87%), followed by child weighing scales
(83%) and thermometers (72%). The lowest level of availability was observed for
stretchers (18%) and dressing sets (23%). HPs in pastoralist communities have lower
availability of all basic equipment than HPs in agrarian communities (Table 3-27).
A majority of key informants indicated that the provision of supplies faces a number
of challenges, including delay, irregularity, and poor quality. While few respondents
reported that the HCs provided supplies to HPs annually, others repeatedly noted
absence at the HPs of HEP manuals, storage lockers, stationery, and registration books.
The inadequacy of supplies has an adverse effect on not only the service delivery but
also the acceptability and preferences of beneficiaries. As a result, communities prefer
to travel long distances to private clinics.
page- 137
Table 3-27. Availability of functional basic equipment at HPs (N=343)
Equipment Agrarian (%) Pastoralist (%) National (%)

Stethoscope 54.5 41.7 53.3
BP apparatus 36.5 30.0 35.9
Thermometer 74.3 57.0 72.7
Kidney dish 57.0 35.6 55.4
Delivery set 43.0 21.5 41.1
Delivery table 49.8 27.8 47.7
Examination table 46.7 26.7 44.8
Cold box 43.9 35.2 43.1
Adult scale 53.5 42.2 52.4
Child scale 83.6 80.2 83.3
Fetoscope 73.4 42.7 70.5
Sterilization 33.0 14.5 31.3
Stretcher 19.2 13.1 18.6
Refrigerator 37.3 50.9 38.5
Vaccine carrier 88.7 68.1 86.8
Drug shelf 66.5 58.6 65.7
Dressing set 24.5 10.2 23.2
Artificial light source 29.4 21.8 28.4
Abbreviations: HEP, Health Extension Program; BP, blood pressure.
The availability of basic equipment, including blood pressure apparatus,

stethoscopes, thermometers, adult scales, child scales, and artificial light sources,
that should always be available in the HPs were assessed in the current survey.
The findings show that only 5.4% of health facilities have all the basic equipment,
and the mean availability of basic equipment is 3.3 from the maximum of 6
(Table 3-28).
page-138
Table 3-28. Availability of basic equipment at health posts
Unweighted #
BP apparatus
%HPs with all

Thermometer
Light sources
Stethoscope
Mean avail-
Background
Adult scale
Child scale
of HPs
ability
equip
National 35.9 72.7 52.4 52.4 83.3 28.4 5.4 3.3 343
Livelihood
Agrarian 36.5 74.3 53.5 53.5 83.6 29.4 5.5 3.5 235
Pastoralist 30.0 57.0 42.2 43.9 80.2 21.8 4.3 2.9 108
Region
Tigray 52.2 91.5 60.8 84.0 94.5 32.3 10.8 4.2 32
Afar 89.9 78.9 83.2 75.6 88.1 46.2 23.7 4.6 18
Amhara 37.5 88.9 55.7 57.2 92.9 22.0 5.1 3.7 60
Oromia 35.6 61.9 50.8 45.8 75.4 33.2 2.5 2.9 74
Somali 30.5 54.5 46.6 33.5 72.6 23.7 9.3 2.6 43
Benishangul-
46.6 89.7 61.0 51.1 94.3 32.2 11.0 3.6 24
Gumuz
SNNPR 32.5 78.0 54.6 59.0 89.1 30.1 8.6 3.1 59
Gambela 20.4 53.9 61.6 69.2 56.3 43.5 0.0 3.1 17
Harari 59.5 63.5 67.5 81.8 90.5 13.5 8.7 3.6 16
Abbreviations: HP, health post; BP, blood pressure; SNNPR, Southern Nations, Nationalities, and
Peoples Region.
3.2.8 Availability of equipment for infection prevention and control
Overall, the availability of equipment for infection prevention and control

was very low. For instance, only 30% of HPs have sterilization equipment. The
availability of sterilization equipment varies by livelihood. Nearly one third of
HPs in agrarian areas have sterilization equipment; in pastoralist areas, only
10% do. Nearly 90% of all HPs had disposable syringes and 89% had safety
boxes at the time of the assessment. Only 46%, however, had sterile gloves and
56% had examination gloves. Meanwhile, waste receptacles were available in
56% of the HPs (Table 3-29).
page- 139
Table 3-29. Availability of IP equipment, by HEP modality (N=343)
Equipment Agrarian (%) Pastoralist (%) Total(%)

Sterilization 33.0 14.5 31.3
Waste receptacles 58.3 39.2 55.5
Face Mask 8.9 13.8 9.6
Safety box 90.3 82.5 89.2
Apron 33.9 17.6 31.7
Heavy-duty glove 11.2 7.3 11.7
Sterile glove 50.0 26.5 46.1
Examination glove 58.7 42.7 56.5
Boots 6.7 5.2 6.5
Disposable syringe 91.6 76.5 89.6
Abbreviations: HP, health post; IP, infection prevention; HEP, Health Extension Program.
3.3 Management of drugs and medical supplies at

health posts
3.3.1 Introduction
Drugs and medical supplies are crucial for the HEP to meet its overall goal of
achieving universal coverage. The availability of essential drugs and medical
supplies at health facilities is key to reducing maternal and child morbidi-
ty and mortality rates and supporting a healthy society. For this reason, the
Pharmaceuticals Fund and Supply Agency (PFSA), now called the Ethiopian
Pharmaceuticals Supply Agency (EPSA), was established. EPSA is responsible
for ensuring the uninterrupted supply of affordable, quality pharmaceuticals
(drugs and medical supplies) to all public health facilities. To this end, the
agency developed and implemented a system called Integrated Pharmaceuti-
cals Logistics System (IPLS) to improve the availability of drugs and medical
supplies at service delivery points.
According to the IPLS, health facilities (hospitals and HCs) send their requests
to their respective EPSA hubs bimonthly, and EPSA delivers the requested
page-140
pharmaceuticals to these facilities directly, or through their respective WorHOs

for less accessible health facilities. HPs then collect the pharmaceuticals and
report to their respective HCs. The HCs use the data in the HP report to esti-
mate consumption and re-supply the required quantities of pharmaceuticals to
the HP.
Logistics information is collected and reported monthly by HPs, and every other
month by HCs and hospitals, on LMIS forms. The LMIS is a system of records
and reports of essential logistics data (from all levels of the supply chain) that
can be used to make informed decisions and ultimately to properly manage the
pharmaceuticals supply chain. For direct-delivery facilities, HCs and hospitals
complete a combined report and requisition form and send it to their EPSA
hubs for requisition processing; the HC request includes the HPs’ pharmaceu-
tical requirements. For non-direct-delivery facilities, HCs complete a combined
report and requisition form, which is sent to EPSA branches through WorHOs.
To avoid stockouts of pharmaceuticals at service delivery points, an emergency

order point is set in the IPLS. The emergency order point for HCs and hospitals
stocks for 15 days and for HPs for 7 days. The LMIS also includes a mechanism
for upper-level facilities to provide feedback to lower-level facilities.
Ethiopia has developed a 5-year Pharmaceutical Supply Transformation Plan

(PSTP) based on the HSTP (2015/16 to 2019/20) and programmatic strategic
documents to meet the health commodity security needs for the transformed
HCs, with the objectives of increasing the availability of essential drugs from
65% to 100%, reducing the stockout of essential drugs from 11% to 0.5%, im-
proving bin-card accuracy from 43% to 85%, improving Report and Resupply
Forms (RRF) data accuracy from 46% to 85%, improving the percentage of
acceptable storage conditions from 55% to 85%, and reducing the wastage
rate from 8% to 2%. This assessment also examined HPs’ drug and medical
supply management. The assessment included the availability of IPLS formats,
the proper use of these formats, the availability of essential drugs, stockouts,
inventory management, the availability of proper storage, and patient-counsel-
ing practices in the HPs.
Tracer drugs were absent from the majority of HPs during the day of study
visit. Moreover, stockout of the drugs for long period of time was very common.
page- 141
Implementation of drug supply chain management was sub-optimal. Only 18%

of HPs had a bin card for stock management, and, of these, 70.4% of the HPs’
bin-card data were complete and accurate. A majority, 81%, of HPs were us-
ing HPMRR, and 25% of the HPs’ HPMRR data were complete and accurate.
More than half (56%) of the HPs were reporting HPMRR in a timely way, and
53% of the HPs resupplied in a timely way. Damaged and expired products
were available in 75% of the HPs, and only 57% of the HPs had implemented
FEFO. However, only 40% and 32% of the HPs had sufficient storage space
and a functional refrigerator for pharmaceuticals, respectively. Patient counsel-
ing about drug dose, frequency, and duration was inadequate at HPs.
3.3.3 Logistics Management Information System
The pharmaceutical supply chain management system in Ethiopia, various

recording and reporting formats have been designed to be used at different
levels of the supply chain. The availability and use of these recording and
reporting formats are critical supply chain indicators. At the health-facility level,
bin cards, the daily dispensing registration book, the Internal Facility Report and
Resupply Form (IFRR), the Health Post Monthly Report and Resupply Form
(HPMRR), and RRF were introduced to record pharmaceutical transactions
and report quantities for resupply. EPSA, together with its partners, has been
printing and distributing these formats to health facilities.
The IPLS recording formats are intended to capture critical logistics data at
each level of the health system. Bin cards and daily dispensing registration
books are the major formats used in recording logistics data. The data captured
on the logistics records are then aggregated to form logistics reports, which
are used to make crucial decisions about resupply quantity, forecasting, and
procurement.
The IPLS reporting formats are also designed to move data through the supply
chain system and help in decision-making. To facilitate correct and consistent
reporting and resupplying within the facility and among the different levels
in the pharmaceutical supply chain, IPLS introduced the HPMRR and RRF.
Hospitals and HCs use the RRF to report their consumption and request a
bimonthly resupply from EPSA, while HPs use the HPMRR every month to
report their consumption to the resupply HC, which can then calculate their
resupply quantity.
page-142
These being the expected standards, the sections below provide the status of
implementation as assessed by this study.
The availability of the standard drug list at HPs was assessed. The results of the
assessment indicated that the majority of HPs did not have their own standard
drug list. Only 17.6% of HPs had a copy of the standard drug list. Gambela and
Harari had the highest proportion of HPs with a copy of the standard drug list
(64.6% and 63.5%, respectively), and the SNNPR had the lowest (5.9%).
The findings of the assessment indicated that the average availability and
use of bin cards was 40.6%. Both were relatively high in HPs in agrarian areas
(42.9%), but in pastoralist areas only 18.3% of HPs were using bin cards. Among
the regions, the highest proportion of bin card availability and use was found
in HPs in Tigray (94.5%), followed by Harari (86.5%), while only 1.4% of the
HPs in Somali were using bin cards. Nationwide, the daily dispensing register
was used in 1 out of 4 HPs. The use of daily dispensing registers varied across
regions, with the highest use in the HPs of Harari (63.5%) and the lowest in
those of Somali (2.9%).
The HPMRR was available in most (81.3%) HPs. Across the regions, the HPMRR
was available in all HPs in Tigray, Afar, Harari, and Somali. Availability was
lowest in the HPs in the SNNPR (61%). In addition, a significant number of HPs
were using the HP-adapted requisition and reporting format (HPARR). The
proportion of HPs using this format was significantly larger in agrarian areas
than in pastoralist ones (51% vs. 26%).
The records and reports used in the IPLS system should be complete and
accurate enough to allow the appropriate bodies to make informed decisions.
Bin cards must be updated to capture all transactions involving the essential
data for logistics (i.e., the received quantity, losses and adjustments, quantity
issued, used, and consumed, and the balance and stock columns of a bin card).
In this assessment, to verify the bin cards’ completeness and currency, 3 bin
cards from each HP were picked at random; it was then verified whether these
columns had captured the most recent transactions. The percentage of updated
bin cards was calculated only for the HPs in which all 3 bin cards were updated.
Each bin card was considered updated and complete when these 4 columns
captured the most recent transactions.
page- 143
The assessment found that only 70.4% of the HPs’ bin cards were updated. A
significant difference was observed among the regions, however, ranging from
100% in Afar, Somali, and Gambela to 0% in Benishangul-Gumuz.
Similarly, the study also assessed the completeness of the HPMRR data.
An HPMRR should be completely filled out for the following essential data:
beginning balance, quantity received, losses and adjustments, and ending
balance.
The assessment revealed that only 25% of the HPs’ HPMRR data were
completely filled out. There was significant regional variation in the HPMRR
data’s completeness which ranged from 1.7% in Somali to 89.5% in Tigray. The
results also varied by livelihood: 26.3% of the HPs in agrarian areas and 8.6%
of HPs in pastoralist areas showed completely filled-out HPMRR data (Table
3-30).
page-144
Table 3-30. Drugs and supplies recording and reporting formats
Recording
Reporting
Recording formats & Reporting
Standard formats
Availability formats
drugs list Availability
completeness
Bin Dispensing Bin
HPMRR HPARR HPMRR
Card Register Card
Tigray 53.4 94.5 30.5 100.0 37.3 81.4 89.5
Afar 16.3 24.1 25.3 100.0 30.5 100.0 28.3
Amhara 26.0 65.8 57.9 94.3 68.1 76.8 42.5
Oromia 16.0 27.9 18.8 79.8 56.3 70.6 15.9
Somali 24.6 1.4 2.9 100.0 100.0 100.0 1.7
Region
Benishangul-
17.1 30.6 32.2 75.4 24.6 0.0 15.8
Gumuz
SNNPR 5.9 43.9 11.5 61.0 24.1 44.0 18.9
Gambela 64.6 26.0 30.7 81.9 54.4 100.0 6.6
Harari 63.5 86.5 63.5 100.0 31.2 41.4 16.2
Agrarian 17.5 42.9 26.4 81.7 51.0 69.4 26.3
Livelihood
Pastoralist 18.3 18.3 12.8 70.8 26.0 100.0 8.6
National 17.6 40.6 25.2 81.3 50.0 70.4 25.0
Abbreviations: HP, health post; HPMRR, Health Post Monthly Report and Resupply Form; HPARR,
health post-adapted requisition and reporting format; SNNPR, Southern Nations, Nationalities, and
Peoples Region.
3.3.4 Pharmaceutical availability
In this assessment, 15 tracer drugs were identified from the full list of products
managed by HEWs. The tracer drugs list included these essential drugs necessary
for treating common childhood illnesses, malaria, pneumonia, diarrhea, and
malnutrition, as well as some representative FP products and medical supplies.
In this survey, data were collected on stock availability on the day of the visit
and the incidence and duration of stockouts during the 6 months prior to the
survey. All analysis was done by product type for HPs that reported an incident.
Although the main task expected from HEWs at HPs is the provision of primary
prevention, they are also expected to provide some selected curative services
page- 145
for common illnesses. In this assessment, ORS, zinc acetate tablets (20 mg),
amoxicillin suspension, paracetamol tablets (500 mg), and paracetamol syrup
or suppositories were considered tracer drugs for HPs.
The assessment showed mixed results regarding the availability of tracer drugs.
The availability of amoxicillin suspension, paracetamol tablets, and paracetamol
syrup or suppositories was significantly low. ORS was available in most (86.5%)
HPs, and zinc acetate tablets (20 mg) were available in 83.7% of HPs. Similarly,
paracetamol tablets and paracetamol suppositories were available in 37.1%
and 15.1% of HPs, respectively. Among the regions, ORS and zinc acetate were
available in all HPs of Tigray but only 60.6% and 48.6%, respectively, of HPs
in Somali, the lowest percentage of all regions. The availability of both ORS
and zinc acetate was better in agrarian HPs than pastoralist ones. Amoxicillin
suspension was available only in 39% of HPs on average, with significant
variation across regions, ranging from 33% in Oromia to 72.8% in Somali.
Comparing the availability of amoxicillin suspension, paracetamol tablets, and
paracetamol suppositories by livelihood, the availability was better in the HPs
of pastoralist areas than agrarian ones. Among the tracer drugs, ORS was
more available, while paracetamol syrup and suppositories were available in a
minimal number of HPs (Table 3-31).
page-146
Table 3-31. Availability of tracer drugs in the HP at the time of visit
Oral rehydration salt
Unweighted #of HPs

Paracetamol 500mg
Paracetamol syrup/
Zinc acetate 20mg
dispersible tablet
suspension/DT
suppository
Amoxicillin
tablet
Tigray 100 96.7 56.8 83 36.4 32
Afar 72.9 63 45.7 35.6 30.8 18
Amhara 98.9 96 39.2 29.7 12.7 60
Oromia 86.9 84.3 33 39 13.4 74
Somali 60.6 48.6 72.8 54.5 59.1 43
Region
Benishangul-
90.9 92 49.3 36.1 19.4 24
Gumuz
SNNPR 78.9 78.3 38 30.4 7.5 59
Gambela 85.9 64.6 49.5 52.8 14.1 17
Harari 90.5 81 58.7 73 40.5 16
Agrarian 88.4 85.7 37.4 36.3 13 235
Livelihood
Pastoralist 67.8 64.3 52.6 45.3 36.1 108
National 86.5 83.7 38.9 37.1 15.1 343
HPs also provide FP, immunization, and preventive services (vitamin A supple-
mentation, deworming, and anemia prevention) and curative services, includ-
ing malaria treatment. Therefore, it was imperative to assess the availability of
the essential medicines needed for these services.
In this regard, the assessment results showed that pentavalent vaccine was
available in 41.9% of the HPs. Among the products needed for FP, Depo-Provera
was the most commonly available FP method (86.3%), followed by Implanon
(69.9%) and oral combined contraceptive pills (58.2%). The availability of
Depo-Provera was greater than 50% across all regions, except Somali, where
it was 17.7%. Similarly, the assessment revealed significant disparity across the
regions on the availability of Implanon, from 2.8% in Afar to 100% in the Harari
and Tigray regions. Implanon availability also varied by HP livelihood, showing
page- 147
availability in 1 out of 4 HPs in pastoralist areas and 3 out of 4 HPs in agrarian

areas. In addition, tetracycline eye ointment and artemether–lumefantrine
(Coartem) were available at fewer than 50% of HPs, at 31.6% and 33.7%,
respectively. Similarly, the availability of vitamin A capsules, Albendazole 400
mg, iron with folic acid, and disposable syringes of any size was 80.3%, 71.5%,
65.2%, and 89.2%, respectively (Table 3-32).
Table 3-32. Availability of tracer drugs in HPs at time of visit
(Medroxyprogesterone) Inj
Disposable Syringe of any

Tetracycline eye ointment
Lumefantrine (Coartem)
Vitamin A capsule any
Unweighted #of HPs

Albendazole 400mg
Pentavalent vaccine
Iron with folic acid
Oral Combined
Implanon NXT
Depo-Provera
contraceptive
Artemether-
strength
size
any
Tigray 41.7 100 96.4 96.4 76.1 96.4 97.7 65.8 66.1 96.4 32
Afar 36.8 2.8 92.6 94.5 63.7 63.2 43 53.6 78.9 100 18
Amhara 55.7 70.4 96 94.9 32.3 90.4 80 68.6 67.4 94.5 60
Oromia 41.3 76.4 93.8 80.5 30.8 69.9 51.3 63.5 24 90.8 74
Region
Somali 47.7 10.4 17.7 53.6 52.7 55.2 65.3 25.8 11 50.1 43
Benishangul-
49.1 72.4 78.8 27.6 22.4 39.5 60.3 44.7 96.6 85.6 24
Gumuz
SNNPR 28.7 70.2 79.2 75.1 21.7 61 73.9 47.9 17 89.7 59
Gambela 71.6 4.7 69.3 42.5 57.5 59.9 36.2 28.3 37.8 69.3 17
Harari 73.8 100 100 91.3 9.5 90.5 72.2 67.5 45.3 90.5 16
Agrarian 41.5 74.5 90 81.6 30.6 72.5 65.1 60.4 34.3 91.3 235
Livelihood
Pastoralist 45.8 24.5 50.2 66.9 41.6 61.4 66.5 37.2 28.1 68 108
National 41.9 69.9 86.3 80.3 31.6 71.5 65.2 58.2 33.7 89.2 343
page-148
3.3.5 Stockout of pharmaceuticals
Stockout of the tracer drugs was assessed to review the level of product availability
and stockout situation at the HPs. The stockout situation during the visit and/or
during the prior 6 months was assessed by reviewing bin cards and stock cards and
interviewing HEWs.
According to the results of the assessment, stockout of tracer drugs on the day of the
visit was significantly high, varying significantly by product type.
As indicated in Figure 3-7, 84.9%, 68.4%, and 66.3% of HPs were out of stock of
paracetamol syrup or suppositories, tetracycline eye ointment, and artemether–
lumefantrine, respectively, on the day of the visit. Only 10.8%, 13.5%, and 13.7% of HPs
were out of stock of disposable syringes, ORS, or Depo-Provera (medroxyprogesterone),
respectively, on the day of the visit. Disposable syringes, ORS, and Depo-Provera
(medroxyprogesterone) were stocked out in relatively fewer HPs.
Figure 3-7. Percentage of HPs with availability of tracer drugs on the day of visit
Abbreviations: HP, health post.
Examining the status of stockout over the last 6 months, it is more common among
HPs in pastoralist areas than among those in agrarian areas. Nationwide, 39.7%
of HPs experienced stockouts of paracetamol syrup or suppositories, followed by
paracetamol 500 mg tablets (38.1%) and iron with folic acid (35.8%). Fewer HPs were
page- 149
out of stock of disposable syringes (9.3%) or pentavalent vaccines (8.6%). Among the
regions, the Tigray, Afar, and Harari HPs experienced no stockouts of ORS in the
past 6 months, while in Somali 3 out of 4 HPs had encountered stockout of ORS in
the past 6 months. Similarly, the Tigray, Afar and Benishangul-Gumuz HPs did not
encounter stockout of pentavalent vaccine in the past 6 months, while in Somali 3 out
of 4 HPs experienced stockouts of pentavalent vaccine in the past 6 months. Implanon
was out of stock in all HPs in Gambela at some point during the past 6 months, and
no HPs in Harari had stockouts of Implanon during the past 6 months. Stockout of
Depo-Provera showed significant variation across regions, ranging from 0% in Afar to
94.3% in Somali. Among all regions, the HPs of Somali encountered the highest level
of stockouts in all the tracer drugs (Table 3-33).
page-150
Table 3-33. Percentage of HPs experiencing stockout of tracer drugs in the past 6 months
Region Livelihood
National
Medicine
Tigray
Afar
Amhara
Oromia
Somali
Benishangul-
Gumuz
SNNPR
Gambela
Harari
Agrarian
Pastoralist
Oral Rehydration Salt 0.0 0.0 21.8 11.5 75.8 10.8 42.4 24.8 0.0 21.4 45.9 23.0
Pentavalent vaccine 0.0 0.0 4.8 4.7 75.2 0.0 1.4 42.9 18.3 4.2 50.9 8.6
Implanon NXT 6.9 0.0 19.8 12.2 73.3 18.3 38.2 100.0 0.0 20.4 31.4 20.7
Depo Provera
16.6 0.0 21.7 9.9 94.3 26.4 48.0 27.2 0.0 22.1 36.1 22.8
(Medroxyprogesterone)
Zinc acetate 20mg dispersible
3.3 0.0 24.8 17.6 76.4 11.9 36.8 40.2 33.3 23.3 48.6 24.9
tablet
Vitamin A capsule, any strength 10.0 0.0 25.4 9.3 74.1 0.0 15.3 27.9 20.0 15.2 42.0 17.1
Tetracycline eye ointment 20.7 9.1 32.2 30.5 67.5 30.6 25.6 37.0 0.0 29.7 53.7 32.5
Amoxicillin suspension/DT 6.5 0.0 30.9 8.2 80.3 33.8 39.9 38.0 8.1 22.5 71.7 28.6
Paracetamol syrup/suppository 0.0 24.1 38.5 27.0 72.7 29.4 56.6 66.7 21.6 31.8 71.2 39.7
Albendazole 400mg 0.0 0.0 13.9 20.1 74.0 0.0 14.8 27.6 14.9 16.3 46.8 18.5
Paracetamol 500mg tablet 2.7 0.0 35.0 25.5 83.2 0.0 68.4 31.4 0.0 34.9 65.1 38.1
Iron with folic acid 7.8 0.0 36.9 37.9 76.6 11.4 30.9 32.7 12.1 34.0 55.2 35.8
Oral combined contraceptive 10.2 0.0 23.2 22.9 56.1 23.0 50.7 50.0 20.0 28.8 32.6 28.9
Artemether-lumefantrine
0.0 13.0 15.8 26.5 59.2 14.9 57.8 24.9 19.3 23.1 35.0 23.8
(Coartem)
Disposable syringe of any size 0.0 0.0 14.0 4.5 64.7 4.2 7.6 37.5 5.2 7.6 33.7 9.3
Unweighted # of HPs 32 18 60 74 43 24 59 17 16 235 108 343
page- 151

The duration of stockout of tracer drugs at HPs varied significantly across the regions
as well as among the tracer drugs. The longest duration of stockout over the past
6 months ranged from 23 days for disposable syringes to 65 days for iron with folic
acid. The average stockout duration for all tracer drugs was more than 45 days. This
implies that 25% of the time in the last 6 months HPs were out of stock of the tracer
drugs. For most tracer drugs, there were significant disparities in the longest duration
of stockout between agrarian and pastoralist HPs, with a longer duration of stockout
in agrarian HPs than those of pastoralist HPs (Table 3-34).
page-152
Table 3-34. Longest duration of stockout of drugs in the last 6 months
Inj
DT
pack
ment
tablet
ceptive
any size
strength
suppository
Depo-Provera
Implanon NXT
dispersible tablet
Iron with folic acid
Zinc acetate 20mg

Paracetamol syrup/
Pentavalent vaccine
Artemetherlumefan-
trine (Coartem) any
Albendazole 400mg
Paracetamol 500mg
Tetracycline eye oint-

Disposable syringe of
Unweighted # of HPs
Oral Rehydration Salt

Amoxicillin suspension/
Oral combined contra-
Vitamin A capsule, any
(Medroxyprogesterone)
Tigray 0 0 113 25 32 32 25 365 0 0 7 30 22 0 0 32
Afar 0 0 0 0 0 0 60 0 240 0 0 0 0 79 0 18
Amhara 26 1 26 33 52 42 90 31 51 41 56 44 40 5 22 60
Oromia 18 168 62 13 15 34 69 82 60 57 52 75 81 63 10 74
Somali 42 12 10 7 20 18 10 66 7 107 39 58 7 6 45 43
Benishan-
Region
gul-Gu- 3 0 19 42 10 0 75 38 180 0 0 53 62 34 120 24
muz
SNNPR 45 47 64 42 103 47 76 44 2 39 82 80 41 91 8 59
Gambela 10 51 7 76 13 10 16 16 7 6 5 18 4 30 15 17
Harari 0 4 0 0 50 60 0 60 125 15 0 7 17 60 30 16
Agrarian 33 121 56 33 59 41 72 49 42 50 66 66 57 54 16 235
Pastoralist 41 13 15 15 25 20 11 66 10 104 38 62 10 44 44 108
Livelihood
National 34 54 55 32 54 37 60 54 31 61 61 65 54 54 23 343
Abbreviations: HP, health post; HEP, Health Extension Program; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page- 153

3.3.6 Inventory management
Inventory management is key to the provision of the required service at an

acceptable standard at all levels of the supply chain. The science of inventory
management works to maintain an appropriate stock level of all products
to avoid shortages and oversupply through the timely ordering of the right
quantity of products.
The HPs’ inventory management was assessed, focusing on the timeliness

of reporting and requesting pharmaceuticals, the establishment and
implementation of emergency ordering procedures, the implementation of First
Expiry First Out (FEFO), near-expiry medicine re-distribution, the availability
of damaged or expired products in storage, storage conditions, and bin card
updating practices.
To ensure the availability of pharmaceuticals, HPs need to report and request

them on time. In this regard, the HPs’ HPMRR timeliness varied by region
and livelihood. The study indicated that only 56.1% of the HPs were providing
timely reporting of HPMRR, and that 53.2% of the HPs were receiving a timely
resupply. HPMRR reporting timeliness was highest in HPs in Tigray (94.7%),
followed by Amhara (90.5%), but only 23.2% in Somali. HPs in agrarian areas
had higher HPMRR reporting timeliness than did those in pastoralist areas
(59.3% vs. 24%). The resupply timeliness of HPs also varied significantly across
regions, with the highest in Amhara (86.9%) and the lowest in Somali (24.5%).
Redistribution of near-expiry pharmaceuticals is critical to avoiding expiry and

minimizing waste. Only 7.8% of the HPs, however, had executed redistribution
of near-expiry pharmaceuticals in the past 6 months.
HPs were also assessed on whether they were practicing FEFO. In this regard,
only 57.3% of the HPs implemented FEFO, with significant variation by region.
In the SNNPR, almost no HPs were implanting FEFO, while all HPs in Afar
and Somali were.
An emergency ordering procedure is decisive in ensuring the continuous

availability of products. Overall, only 1 out of 4 HPs knew and practiced the
emergency ordering procedure, with its highest use in Tigray and the lowest in
Afar and Somali.
page-154
The availability of damaged or expired products at HPs was also assessed.

The assessment revealed that damaged or expired products were available in
74.6% of HPs. Only 2.6% of HPs in Somali had damaged or expired products.
On the other hand, 93.6% of the HPs in the SNNPR had damaged or expired
products. The availability of damaged or expired products was higher in HPs
in agrarian areas than in those in pastoralist areas (79.4% vs. 26.5%). This
could be related to the availability of more experienced clinical staff, including
nurses and midwives, in pastoralist HPs than in agrarian HPs, where HEWs
with limited clinical skills constitute a majority of the workforce. The assessment
also revealed that 78.3% of the HPs separated damaged or expired products
from usable products. Only 36.1% of the HPs, however, updated their inventory
records and wrote off the damaged or expired products from their inventory
records (Table 3-35).
Table 3-35. Inventory management implementation among HPs
products separated
Procedures for EO
Damaged/expired
Damaged/expired
Damaged/expired
products removed
FEFO implement
Timely Resupply
Timely HPMRR
Re-distribution
from inventory
of near-expiry
established
Reporting
medicines
products
record
Tigray 80.4 81.4 21.6 53.4 82.2 56.9 94.7 82.4

Afar 100 0.0 30.3 49.7 100 42.6 42.8 40.7
Amhara 81 61.7 29.4 80.3 93.0 63.6 90.5 86.9
Oromia 67.3 16.7 1.1 71.9 62.8 24.9 41.5 47.5
Somali 100 0.0 0.0 2.6 58.8 0.0 23.2 24.5
Region
Benishangul-
50.7 26.5 7.5 65.5 89.5 43.9 71.2 70.1
Gumuz
SNNPR 1.0 7.4 0.0 93.6 86.8 28.1 53.0 35.5
Gambela 36.3 16.6 11.8 54.4 100 78.2 66.2 57.5
Harari 84.4 31.7 22.2 65.1 70.7 14.7 63.5 58.7
Agrarian 57.7 26.8 8.4 79.4 78.3 36.6 59.3 56.1
Livelihood
Pastoralist 43.4 8.0 2.0 26.5 78.2 18.8 24.0 23.7
National 57.3 25.1 7.8 74.6 78.3 36.1 56.1 53.2
Abbreviations: HP, health post; FEFO, First Expiry First Out; EO, emergency ordering; HPMRR, Health
Post Monthly Report and Resupply Form; SNNPR, Southern Nations, Nationalities, and Peoples
Region.
page- 155
HPs are expected to conduct physical inventory on a regular basis, at least

once per year and at most every month. The assessment revealed, however,
that only 1 out of 4 HPs were conducting physical inventory at least once per
year. More agrarian HPs conducted physical inventory at least once yearly
than did pastoralist HPs (30.1% vs. 15.6%). Among the regions, the highest
percentage of HPs conducting physical inventory yearly was found in the HPs
of Tigray (66.8%), while the lowest percentage was found in the HPs of Somali
(14.1%; Table 3-35).
3.3.7 Pharmaceutical storage
Pharmaceuticals require proper storage to ensure that their quality is maintained.

The HPs’ pharmaceutical storage conditions were assessed in this study.
Accordingly, 57.5% of the HPs kept their medicines either at the dispensary or
in storage. Only 39.7% of the HPs, however, had sufficient space in storage or
at the dispensary. Overall, 77% of storage areas were protected from direct
sunlight, while only 59% were clean and well ventilated. A minority of HPs had
storage areas with a functional refrigerator (31.5%). HPs in pastoralist areas
had a higher percentage of functional refrigerators than those in agrarian
areas (50.4% vs. 29.6%). The availability of a functional refrigerator ranged
significantly, from 18.9% in the SNNPR to 85.9% in Gambela. All HPs in Harari
had a functional thermometer, while Somali HPs had the lowest percentage of
available functional thermometers. Overall, 78.9% of the HPs had a functional
thermometer.
page-156
Table 3-36. Pharmaceutical storage conditions among study HPs
HP with functional
HP with functional
conducted at least
Physical inventory
dispensary/ store
Store/dispensary
Ventilated store
Store protected
sufficient space
once in a year
Medicines are
kept either as
thermometer
refrigerator
Clean store
from direct
sunlight
Tigray 61.5 91.4 80.3 89.0 80.0 63.2 80.2 66.8
Afar 89.7 94.9 91.8 86.7 71.8 52 77.4 52.2
Amhara 88.5 63.5 61 91.4 42.7 22.9 80.9 55
Oromia 47.2 53.0 53.5 58.3 14.5 36.5 81.9 18.8
Somali 26.1 60.2 75.0 44.0 53.3 54.1 49.9 14.1
Region
Benishangul-
69.4 72.7 90.1 95.1 80.3 45.5 75.6 39.7
Gumuz
SNNPR 53.0 55.3 57.8 84.9 61.7 18.9 85.1 20.4
Gambela 71.6 75.9 14.1 67.2 53.1 85.9 69.8 26.0
Harari 86.5 89.0 83.5 94.5 46.8 73.8 100 58.7
Agrarian 59.7 59.6 58.2 78.1 39.5 29.6 82.8 30.1
Livelihood
Pastoralist 36.6 55.3 75.6 58.6 43.7 50.4 56.4 15.6
National 57.5 59.4 59.2 77 39.7 31.5 78.9 28.7
3.3.8 Reported avaialbility of drugs
According to respondents, drugs are “non-existent” in some HPs, and the HPs
are “empty.” Shortages include basic medical equipment, as well as different
drugs, medical supplies, and medical equipment, like vaccines, insecticide-
treated bed nets (ITNs), weighing scales, blood pressure apparatus, and
ambulances. Alarmingly, a majority of the participants reported that drugs
to treat communicable and non-communicable diseases, ORS, anti-malarial
drugs, water-purifying tablets, and common pain relievers are not always
available at the HPs.
page- 157
“
The health facility is constructed, but there is shortage of professionals.
There is no doctor, and we don’t get medicine when we go, so we go
to Pawe, Bahir Dar, or Chagni. The building is good, but they don’t
have blood pressure measurement, let alone medicine. What is the
reason for these? There is only 1 ambulance, it has to go to different
areas, and there will be 3 to 4 deliveries a day; which of these should
it serve?
“
One of the biggest problems we have is emergency treatment. Our

HP does not have ORS for diarrhea treatment. The HP does not
have painkillers. For such treatments we travel to Sawla. We wish the
HP was well equipped.
FGD, WDA Leaders
Complaints about service delivery with limited resources were widespread.

Emergency services are given without medications. Getting screening services
at HPs is a nightmare. The lack of drugs also applies at the woreda level.
HCs are reported to be out of any drugs at all. Equipment and supplies are
outdated. Not only the medications’ inadequacy but their quality, quantity,
and effectiveness are problems. Communities prefer not to go to HPs because
of the lack of availability of drugs there. Community members with chronic
health problems reported going to the woreda to get medications and having
unsatisfactory outcomes. Traveling long distances to get diagnostic services
and drugs at private clinics and pharmacies is common.
“
The HP, as well the newly constructed HC, does not have medication.
The government did not supply medication. The community is
complaining about the lack of medical drugs. They are complaining
that, after constructing the health facility and hiring health
professionals, the facilities are not supplied with medication, and this
makes all the government activities and spending worthless. We are
forced to buy from private pharmacies.
page-158
The inadequacy of medical supplies is a serious problem for HEP services.

The HEP service packages, such as FP, are not well implemented, resulting
in, for instance, women being exposed to unwanted pregnancies. By the same
token, in situations where there is a shortage of options of contraceptives,
HEWs persuade beneficiaries to use available ones. The findings of the study
confirmed that the lack of medications at HPs and HCs makes the available
services incomplete.
As participants revealed, equipment provision has lacked regularity and

timeliness since the HEP’s founding. Medical supplies might be unavailable for
long periods, leading to serious problems; equipment provided or donated to
HPs vanishes for unknown reasons. There are situations in which 1 ambulance is
shared among 16 kebeles; therefore, ambulances are unreliable, so the community
prefers other local transportation services. In some places, ambulances are
not functional at all. People residing in places with difficult topography are
deprived of these services. HEWs are forced to find their own ways to reach
less-accessible areas and deliver services.
The lack of medication and ambulances has a negative impact on maternal-child

health (MCH) services too: children suffer from malaria; the unreliable arrival
of ambulances leads mothers to give birth at home. Lack of ambulances is one
factor in maternal mortality. Respondents also mentioned that ambulances are
unwilling to return mothers to their homes after delivery, inconveniencing them,
“
and, as a result, mothers prefer home birth.
The other problem we have is, when we called an ambulance to the

village at the time of labor, ambulances may come if there is a road
and help us with transportation to the facility. Nevertheless, when the
woman gave birth at the facility and was discharged, they were not
cooperative enough to serve the mother to get her back to her house,
and this contributes to the low coverage of the facility-based birth. You
know, mothers are suffering from the lack of transportation. They are
exposed to direct sunlight searching for transport services, and some
are bleeding while many people are waiting for transportation, and
they are ashamed of bleeding in front of many people. Mothers are
at a disadvantage in some conditions, you see, giving birth with lots
of suffering, but wandering here and there at different bus stations
for transportation is also culturally unacceptable. We are asking the
government to solve this critical problem, and other officials too.
page- 159
According to participants, the lack of medication affects health equity among

communities of different economic statuses. Ambulance services are not inclusive
of the whole population; rather, they serve only pregnant women. The services’
neglecting male community members has raised questions.
The lack of availability of drugs and medical equipment has brought reflection
to the HEP regarding the emerging needs of the community. People with NCDs
cannot get medications at the public facilities. The quantity and timeliness of
drug provision are challenges in addressing the needs of the population. The
allocated drugs are not proportionate to the population size. Inadequate drug
supply at HPs has led the community to question the purpose of building HPs
“
in their locality.
When it comes to medicine supply at the HP, there are only limited
types of medication available in the HP, yet, for example, antimalarial
doses, as malaria is widely prevalent here in this area, paracetamol
for the elderly, etc., are only available in the HP. Surprisingly, people
in this area often come to the HP and say, if the HP can’t provide
all the required medical treatment here, why it was established here?
They are claiming that the HP has to provide them with whatever
medical treatment they are looking for.
KII, HEP Supervisor
3.4 Financing of the Health Extension Program

Investment in the HEP has been increasing in absolute terms. Still, the share of
spending on the HEP in relation to total expenditure at the HC and HP levels
and total health expenditure in general has continuously declined since 2010.
Except for voluntary contributions of time and labor at the community level,
government and donors are almost the only financing sources of the HEP. The
government’s share in financing the HEP has increased over the years. The
HEP remains a highly donor-dependent program, with 65.3% of its spending
still coming from donors.
page-160
National spending on the HEP was estimated through further analyses of

data from the 2016/17 National Health Accounts. Health expenditures at
the HC and HP levels was apportioned to the HEP (at the HP level) and
HC-level expenditure using data from different secondary sources, including
EPSA drug cost, supply and medical equipment, human resources data, health
service coverage and use reports, and HSTP. The annual inflation rate from
the Central Statistics Agency (CSA) and purchasing power parity (PPP) from
the World Bank were also used to adjust the expenditure figures to standard
units. Primary data on HP-level inputs were also used to supplement secondary
sources in the estimation of the share of the HEP from the total PHCU-level
expenditure. The nominal spending data were adjusted based on the PPP and
exchange rates presented in (Table 3-37.)
Table 3-37. ETB to PPP and USD Conversion Factors for Ethiopia, 2010/11 -
2016/17
Conversion Factor 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17
Unit ETB per USD Purchasing

4.55 5.65 6.5 6.9 7.55 8.3 8.9
power parity conversion
Unit ETB per USD Average

15.7 17.3 18.2 19.1 20.1 21.2 22.8
Exchange Rate
Abbreviations: ETB, Ethiopian birr; USD, US dollar.
3.4.2 Total HC and HP (PHCU) expenditure
Total HC, HP, and PHCU expenditure, excluding the expenditure at primary
hospitals, indicated an increasing trend in absolute terms from 9.27 billion ETB
(USD 2.04 billion in terms of PPP) in 2010/11EFY to 23.73 billion ETB (USD
2.67 billion in terms of PPP) in 2016/17. The average recorded spending was
15.69 billion ETB (USD 2.22 billion in terms of PPP) based on the National
Health Accounts studies (NHA) from 2010-2017 (Figure 3-8).
page- 161
Figure 3-8. Total health center and health post spending, 2010/11 to 2016/17, in
billion ETB
Abbreviations: ETB, Ethiopian birr; USD, US dollar.
3.4.3 Total health center and health post expenditure by source
Contributions from outside Ethiopia constituted the major source of HC and

HP spending. Government contributions amounted to an average of 6.64
billion ETB (USD 0.91 billion in terms of PPP) per year, while the rest of the
world contributed an average of 9.05 billion ETB (USD 1.32 billion in terms of
PPP) per year. Expenditures from both sources have increased over the years.
Moreover, the funding gap between government and the rest of the world
has been narrowing, indicating a shift toward domestic financing, even though
the majority of HC and HP expenditure still comes from the rest of the world
(Figure 3-9).
page-162
Figure 3-9. Total health center and health post expenditure, by source, 2010/11
to 2016/17
3.4.4 Share of HEP spending from total HC- and HP-level

expenditures
Between 2010/11 and 2016/17, total spending on HC and HP increased by more

than double (254%) in nominal terms. This increase was more pronounced at
the HC level (269%) than the HP level, resulting in the trend of a declining
share of HP-level expenditure, from 25% in 2010/11 to 22% in 2016/17. On
average, HP-level expenditure constituted 23% of the total HC- and HP-level
expenditure (Figure 3-10).
page- 163
Figure 3-10. Magnitude of HEP spending (In billion ETB) and share of the HEP
from total PHCU (both health center and posts) level expenditure 2010/11 to
2016/17
Abbreviations: ETB, Ethiopian birr; HEP, Health Extension Program; PHCU, Primary Healthcare Unit.
3.4.5 Share of HEP spending from total health expenditure
Total HEP spending in nominal terms increased from 2.4 billion ETB (USD 0.52
billion in terms of PPP) in 2010/11 to 5.1 billion ETB (USD 0.58 billion in terms
of PPP) in 2016/17. On the other hand, the share of HEP spending in the Total
Health Expenditure (THE) declined from 8.89% in 2010/11 to 7.12% in 2016/17
(Figure 3-11).
page-164
Figure 3-11: Magnitude and share of HEP the HEP from the total health
spending, 2010/11-2016/17
Abbreviations: ETB, Ethiopian birr; USD, US dollar; PPP, purchasing power parity; HEP, Health
Extension Program, THE, Total Health Expenditure.
3.4.6 HEP spending by source of financing
HEP financing for the period 2010/11 to 2016/17 has come largely from the rest of
the world, with an average share of 65.3% from foreign sources. The remaining
34.7% was covered by the government. The share of government spending in
the total HEP spending increased from 20.8% in 2010/11 to 40.3% in 2016/17.
Despite the commendable increases in the share of government spending to
the HEP, the program is still predominantly financed by non-domestic sources.
Moreover, no increases were observed in the share of government spending
since 2013/14 (Figure 3-12).
Figure 3-12. Total HEP spending trend, by source, 2010/11-2016/17

Abbreviations: HEP, Health Extension Program. page- 165
3.4.7 HEP spending by economic classification
Spending on the HEP entails capital spending for the construction of HPs, which
is the program’s point of service delivery to the community. Of the total HEP
spending during the period 2010/11 to 2016/17, the average share of capital
spending accounted for 14%, while the remaining spending was a recurrent
expenditure.
Figure 3-13. Average total HEP spending share, by economic classifications,

2010/11 to 2016/17
3.4.8 HEP spending by input type
Total expenditure on the HEP was divided into 3 categories based on the type of
input. The large majority (62%) of HEP spending was related to drugs, supplies,
and medical equipment, while 24% was for human resources-related expenses,
including salaries, basic training, and refresher trainings. Recurrent expenditure
economic classifications consist of spending on 2 input type categories. Of the
average total 86% recurrent spending from 2010/11 to 2016/17, 62% went to
spending on drugs, supplies, and medical equipment, while 24% was dedicated
as human resources spending, entailing salaries, basic training, and periodic
refresher trainings. The remaining 14% was invested in infrastructure-building
and other related activities (Figure 3-14).
page-166
Figure 3-14. Average total HEP spending, by input type, 2010/11 to 2016/17
Abbreviations: HEP, Health Extension Program; HR, human resources.
3.4.9 HEP spending by type of health services
The HEP includes a number of packages, including essential health services

focusing on family health (reproductive and maternal health, child health,
and nutrition), communicable disease prevention and control, and hygiene
and environmental sanitation. Disaggregating HEP spending during the
period 2010/11 to 2016/17 by type of service indicates that child health services
accounted for the highest share of spending (46% of average spending),
followed, in decreasing order of their share of HEP spending, by nutrition,
hygiene and sanitation, reproductive and maternal health, and communicable
disease prevention and control. One possible reason for the significantly higher
spending share on child health might be drugs and supplies for vaccinations and
curative service for iCCM- and CBNC-targeted diseases, such as pneumonia,
diarrhea, sepsis, and early neonatal diagnosis and treatment of cases at the
HPs
Figure 3 15. Average total share of HEP spending,

by type of service, 2010/11 to 2016/17 page- 167
CHAPTER 4
Health Service
Delivery
Through The
HEP
page-168
4. Health Service Delivery through the
CHAPTER 4
INTRODUCTION
Ethiopia’s population is predominantly (85%) rural, with high population growth

and a young age structure. These phenomena have remained unchanged
for over 2 decades. The country’s primary health problems remain maternal,
newborn, and childhood conditions, communicable diseases, and nutritional
disorders. Evidence shows that most Ethiopians, especially those who live and
work in rural areas, face high morbidity and mortality rates that are largely
attributable to potentially preventable conditions linked to low health-seeking
behavior and inadequate health service delivery across the country. In response
to these health needs, the HEP was designed to deliver 16 packages of health
services intended to improve the living conditions of households, boost demand
for health services, and increase the use of high-impact, low-cost interventions.
The HEP intends to achieve these outcomes through a number of health service
delivery modalities, including home visits, community outreach, and HP-based
services. An HP staffed with at least 2 HEWs, closely working with a cluster
HC, is expected to provide HEP services for 1 000 households (5 000 people)
living in agrarian kebeles or 500 to 600 households (3 000 people) living in
pastoralist kebeles.
The HEP involves 3 primary implementation modalities: (a) home visits, (b) HP
visits, and (c) outreach services. HEWs are expected to visit households within
their catchment area at least once yearly; repeat visits are generally expected
in order to ensure the adequate household-level implementation of the HEP
packages and maintain current information about every household. Outreach
services are used to provide services like vaccination and growth monitoring
through scheduled sessions and to transmit health messages to a larger group
of respondents in different community gatherings. In addition to conducting
home visits and outreach sessions, HEWs are expected to provide selected
clinical services at their HPs. Model family training (MFT) constitutes the other
major mechanism for intensively training women on the 18 HEP packages, with
the assumption that trained women will change their households’ behaviors
and influence the adoption of the desired behaviors among their neighbors and
page- 169
Health Service Delivery through the HEP
relatives. The completion of trainings and the adoption of the desired behavior
at the household level leads to the designation of households as models. The
exposure of household members, particularly women, to the HEP packages
through home visits, HP visits, and community outreach sessions is ultimately
expected to lead to the implementation of the HEP at the household level,
reflecting the adoption of the desired behaviors by the household members.
The HEP Assessment included modules for assessing access to the HEP and
communities’ level of exposure to the HEP. The HP assessment module assessed
the availability of specific services related to the 18 HEP packages. In addition,
the household survey assessed whether community members were aware of the
availability of those services, whether they had had any exposure to HEWs/
HPs, the modality of their exposure, the time of their most recent exposure,
and the services received during their most recent contact with HEWs. This
section presents the findings of the assessment in terms of both household
access and household exposure to the HEP. This section presents the findings
on the availability of the HEP services at HPs, community members’ awareness
of the HEP services, and exposure to the HEP services.
HPs and HEWs are almost universally available to communities. Availability,

however, has not translated into real access to services. Of the 352 kebeles
included in the HEP National Assessment, 97.4% had at least 1 HP. More
than 90% of HPS reported that the majority of the HEP packages were
available. Interruption of services, the services’ lack of comprehensiveness, and
communities’ limited awareness of reportedly available services, however, were
major barriers to access. The household survey indicated that awareness by
community members of the availability of HEP services was only 58.8%. About
half or less of women, men, and youth girls had contact with HEWs. What
exposure there was related more often to HP-based services than to home-
and community-based health promotion activities. As a result, household-level
implementation of the HEP was very low. Men and youth were in general
marginalized from HEP services. During the 1-year period preceding the study,
the women, men, and youth girls reporting at least one exposure to the HEP
through any modality were 54.8%, 32.1%, and 21.9%, respectively. HPs were the
predominant source of exposure, compared to home visits and other settings,
in both agrarian and pastoralist communities, with an even higher proportion
in pastoralist settings.
page-170
4.1 Physical accessibility of HPs and availability of

services
As presented in previous section, HPs are almost universally available in the
study kebeles; at least 1 HP was available in 97.4% of the kebeles included in
the study. The availability of services in these HPs was also assessed for specific
services related to the HEP packages by asking HEWs (a) whether a service
was provided by the HP, (b) whether a service was available on the date of
data collection, and (c) how many clients had received a service from the HP
during the 1 month prior to data collection. The assessment included 4 areas
of the HEP, including 16 items on health promotion, 28 items on reproductive
health (10 on FP, 10 on ANC, and 8 on delivery and PNC), 13 items on child
healthcare, and 10 items on communicable disease prevention and control. A
total of 67 tracer services were assessed for availability. Service availability
scores were calculated as the proportion of HPs that provide each service.
The findings show that most of the basic services are widely available; there have
been limitations, however, in the comprehensiveness of the available services
and service interruptions. At least 1 FP method was reported to be available
in 94.5% of the HPs. Similarly, ANC and PNC were reported to be available
in 97.5% and 95.1% of HPs, respectively. Disease prevention- and control-
related services, including malaria diagnosis and treatment and TB screening
and/or treatment, were also reported to be available by large majorities of
the HPs. The provision of health education on topics related to hygiene and
environmental sanitation, including latrine use, personal hygiene, and the safe
disposal of solid and liquid waste, was reported as available in more than 90%
of the HPs. Despite the reported availability of services in large proportions of
the HPs, a smaller than reported number of HPs had services available on the
date of data collection, and much lower numbers of them had a record of at
least 1 client being provided with a service during the 1-month period preceding
the date of data collection (Table 4-1).
page- 171
Table 4-1. Availability of selected HEP services
Percent of HPs with:
At least
1 client
received the Service
Service area of HEP Reported
service during availability
service
the last at the time of
availability
completed data collection
month prior
the study
At least 1 FP method 94.50 93.70 94.5
Short-acting FP 92.50 91.30 92.5
Long-acting FP 79.60 63.90 79.6
Family health ANC 97.50 94.30 69.2
PNC 95.10 80.70 60.3
Immunization 93.50 91.70 38.9
Growth monitoring &
90.20 79.60 58.2
nutrition
Adolescent reproductive
87.90 51.80 54.1
health service
TB screening and/or
90.30 53.70 53.2
treatment follow-up
Disease Malaria diagnosis and
prevention and 81.90 42.50 58.4
treatment
control
HIV counseling and testing 72.90 46.10 36.6
First aid 71.80 21.00 34.8
Water supply and food
97.50 71.90 63.9
hygiene
Health
education on: Personal hygiene 97.40 80.50 66.2
Latrine use 97.30 77.90 63.3
Dry and liquid waste
95.30 74.50 61.5
management
Abbreviations: HEP, Health Extension Program; FP, family planning; PNC, postnatal care; TB,
tuberculosis.
The available services’ lack of comprehensiveness was manifested in the low

service availability mean score across different service packages. The service
availability score, calculated as the proportion of specific components of a
service or package that were reported as available, averaged across study
page-172
HPs, was low for all packages except health education and child health. The
average service availability score was 57.4% for FP, indicating a limited choice,
and 69.5% for ANC, indicating a lack of comprehensiveness of care (Figure
4-1).
Figure 4-1. Service availability mean score, by service category or package
Abbreviations: SA, service availability; ANC, antenatal care.
4.2 Community awareness of and familiarity with HEP services

Women, men, and youth girls were asked to list the services that HEWs
provided in their respective kebeles, without the data collectors’ prompting the
respondents for specific types of services. Once respondents listed the services
with which they were familiar, data collectors then asked them about their
awareness of additional services they expected from HEWs. Respondents who
listed a service before being asked about it were considered to be familiar with
that service. Respondents were considered to be aware of an HEP service if
either (a) they were familiar according to this definition or (b) they mentioned
that a service was provided by HEWs after being asked about it by a data
collector.
Awareness of HEP services was more than 50% for most HEP services; only
a small proportion of the respondents, however, were familiar with the HEP
services. Health education, child vaccination, FP, tetanus-toxoid (TT) vaccination,
and ANC were the most popular services across all groups of respondents.
page- 173
Women and men were relatively more familiar with HEP services than were
youth girls. Low levels of familiarity with and awareness about HEP services
was a problem even among women, who have been the primary targets of the
program throughout the HEP’s history. Women’s awareness was 75.1% for ANC,
80.0% for health education, 80.3% for FP, and 88.3% for child vaccination.
Familiarity with these services was even lower; less than half of women listed
these services as being provided by HEWs in their respective communities
(Table 4-2, Figure 4-2).
Figure 4-2. Women’s awareness of and familiarity with HEP services

Abbreviations: HEP, Health Extension Program; TT, tetanus-toxoid; ANC, antenatal care; NCD, non-
communicable diseases.
page-174
Table 4-2. Awareness of and familiarity with services provided by HEWs
Among respondents who were asked to list the services provided by HEWs. The proportion of respondents who
mentioned specific services spontaneously, recognized a service after probing, and were unaware of a service
Women Unweighted N=6430 Men Unweighted N=4416 Youth girls Unweighted N=900
Mentioned Recognized Not Mentioned Recognized Not Mentioned Recognized Not
HEP Services
spontaneously after probe aware spontaneously after probe aware spontaneously after probe aware
Health
49.2 30.8 20 55.4 25.9 18.7 43.6 31.5 24.9
education
Child
42.9 45.4 11.7 40.9 46.9 12.2 34.8 41.9 23.4
vaccination
Family planning 33.8 46.5 19.7 30 48.7 21.3 19 48.7 32.3
TT vaccination 24.4 52.5 23.2 22 53.4 24.6 22.5 44.9 32.6
Antenatal care 21.3 53.8 24.9 20.4 52.6 27.1 12.4 52.6 34.9
Deworming 9.1 50.3 40.6 13.9 45 41.2 13.6 33.8 52.7

Postnatal care 8.9 41.7 49.5 9.2 40.4 50.4 4 32.8 63.2
Treatment 8.2 37.4 54.4 10 34 56 5.5 29.9 64.7
Supplemental
8.2 32.4 59.4 8.6 32.1 59.3 4.2 26.1 69.7
food
Growth
7.8 44.4 47.8 6.6 42.8 50.6 3.9 31.2 64.9
monitoring
Food
6.5 34.1 59.4 6.1 29.6 64.3 3.4 25.3 71.3
demonstration
Referral 6.3 50.3 43.4 8.2 47.8 44 3.2 38.1 58.7
NCD screening 2.1 16.4 81.5 2.4 17.9 79.7 0.8 12 87.1
Abbreviations: HEW, Health Extension Worker; HEP, Health Extension Program; TT, tetanus-toxoid; NCD, non-communicable diseases.
page- 175

4.3 Community members’ level of exposure to the
HEP
4.3.1 Home visits by HEWs
The proportion of households reporting ever having had a visit by an HEW

was 55.1% (22.3% in pastoralist settings and 56.6% in agrarian settings). Only
32.0% of agrarian households and 14.7% of pastoralist households reported at
least 1 HEW visit to their homes during the 1-year period preceding the study.
Households were more likely to receive multiple visits by HEWs if they received
1 visit. Home visits were highest in Benishangul-Gumuz and lowest in Somali,
with 1-year rates of 76.4% and 4.8%, respectively. Variation in the coverage
of home visits was also observed across wealth quintiles. The proportion of
households with at least 1 HEW visit during the previous year decreased from
37.7% among households in the richest quintile to 23.8% in the poorest quintile
(Figure 4-3, Table 4-3).
Figure 4-3. Home visits by HEWs
Abbreviation: HEW, Health Extension Worker.
page-176
Table 4-3. Proportion of households receiving visits by HEWs, by background

characteristics of households
Proportion of
Households that Proportion of Households that
had at least 1 visit during the last 1 year had:
by an HEW Unweighted
Characteristics 4 or number of
2 or 3 households
No 1 visit more
Last 1 visits
Anytime visit by by visits
year by
HEW HEW by
HEW
HEW
Total 55.1 31.2 69.1 5.9 15.2 9.8 6504
Agrarian 56.6 32.0 68.4 6.0 15.6 10.0 4454
Livelihood
Pastoralist 22.3 14.7 85.4 2.4 6.9 5.3 2050
Tigray 67.2 39.5 61.4 10.3 17.4 10.9 614
Afar 31.8 25.2 75.2 2.1 12.6 10.1 412
Amhara 64.9 41.7 58.7 6.1 17.1 18.1 1066
Oromia 52.8 25.4 74.7 5.7 14.5 5.1 1323
Region Somali 5.5 4.8 95.2 0.2 1.8 2.8 821
Ben-Gum 90.6 76.4 23.8 7.8 37.8 30.6 407
SNNPR 49.4 30.4 70.4 5.6 15.0 9.1 1023
Gambela 51.6 45.3 55.7 5.6 26.7 12.0 422
Harari 42.9 25.2 74.8 5.3 14.1 5.9 416
Lowest 42.4 23.8 76.2 4.3 11.2 8.3 1348
Lower 50.5 29.2 70.9 6.5 14.5 8.1 1318
Wealth
Middle 56.8 32.7 68.5 6.8 14.6 10.2 1293
Index
Higher 59.2 31.5 68.9 6.5 16.4 8.1 1297
Highest 64.1 37.7 62.3 4.8 18.7 14.1 1248
Abbreviations: HEW, Health Extension Worker; Ben-Gum, Benishangul-Gumuz; SNNPR, Southern

Those respondents who reported at least 1 HEW visit during the 1-year period
preceding the survey were asked about the household members who were
contacted and the type of services provided during the home visit. The findings
showed that adult women were the most commonly contacted household
members, followed by adult men and children under the age of 5 years.
page- 177
Adolescent and youth boys were the least targeted household members. Health
education and child vaccinations were the most commonly provided services to
households with at least 1 HEW visit during the last year (Table 4-4).
Table 4-4. Characteristics of HEW visits to households among those having at

least 1 visit during the 1-year period preceding the survey, by livelihood
Livelihood
Agrarian Pastoralist Total
Unweighted number of Households with at least 1 HEW
1787 447 2234
visit during last 1 year
Adult women (25+ years) 69.5 56.2 69.3
Among households that Adult men (25+ years) 24.0 24.4 24.0
had at least 1 HEW visit Children (0-4 years) 20.8 38.5 21.2
during the last 1 year, Children (5-9 years) 12.6 19.4 12.7
proportion of households
where HEW at least once Youth girls (15-24 years) 8.2 13.2 8.3
met at least 1 household Early adolescent girls (10-14) 6.3 10.2 6.4
member: Youth boys (15-24 years) 6.2 9.9 6.3
Early adolescent boys (10-14) 4.5 7.1 4.5
Health education 83.6 80.1 83.5
Child vaccination 35.5 58.4 36.0
Deworming 21.4 28.8 21.5
Among households that Growth monitoring 17.9 27.7 18.1
had at least 1 HEW visit Food preparation 17.1 15.5 17.1
during the last 1 year,
Treat sick member 15.7 25.6 15.9
proportion of households
where HEW at least once TT vaccination 12.7 25.7 13.0
provided: Antenatal care 12.2 24.2 12.5
Postnatal care 12.3 18.6 12.4
Referral to health center 12.0 17.5 12.1
Supplemental food 7.3 18.7 7.6
Abbreviations: HEW, Health Extension Worker; TT, tetanus-toxoid.
The health education topics covered during home visits were largely related to
hygiene and environmental sanitation. Latrine construction and use, personal
hygiene, waste management, and water supply/food hygiene were reported
as topics of discussion during home visits by 92.9%, 84.5%, 82.1%, and 73.4%,
page-178
respectively, of households that reported receiving health education during an

HEW visit in the past 1 year. The least covered topics were HIV, TB, and
malaria (Table 4-5).
Table 4-5. Health education topics covered among households that received
health education during the most recent HEW visit during the 1-year period
preceding the study

Unweighted number of households who
received health education during most 1524 360 1884
recent home visit by HEW
Latrine construction and
93.0 87.9 92.9
use
Personnel hygiene 84.6 80.8 84.5
Dry and liquid waste
Percentage 82.2 80.4 82.1
management
of households
where the Water supply and food
73.5 67.5 73.4
following hygiene
topic was Family planning 59.0 71.9 59.2
discussed Maternal and newborn
during health 53.5 70.8 53.9
health
education
Immunization 52.4 69.0 52.8
provided to
household Community-Based
49.8 39.1 49.6
members Health Insurance
during the Nutrition 47.9 57.0 48.1
last year Care for sick children 42.8 57.7 43.1
Malaria 41.1 59.9 41.4
Tuberculosis 35.4 48.7 35.7
HIV 35.0 44.4 35.2
4.3.2 Health post visits by household members

Women, men, and youth girls were asked whether they had ever visited an
HP. Respondents who had had at least 1 HP visit at any point were also asked
about their most recent visit. Ever having visited an HP was reported by 58.3%
of women, 23.9% of men, and 14.9% of youth girls. At least 1 visit to an HP
during the 1-year period preceding the study was reported by 39.4% of women,
14.5% of men, and 10.0% of youth girls (Table 4-6).
page- 179
Table 4-6. HP visits during the last one year
Proportion of
respondents who Proportion of respondents who
visited a health during the last year had
post during: # of
respondents
4 or (unweighted)
No 2 or
Last 1 1 HP more
Livelihood Respondent Lifetime HP 3 HP
year visit HP
visit visits
visits
Women 59.0 39.7 60.3 6.3 16.4 16.9 4421
Agrarian Men 24.0 14.4 85.6 5.4 6.3 2.7 3157
Youth 14.9 9.9 90.1 5.3 3.8 0.7 658
Women 42.4 34.9 65.1 5.7 17.1 12.1 2009
Pastoralist Men 20.3 16.3 83.7 4.0 7.7 4.6 1259
Youth 14.6 12.1 87.9 3.3 5.8 3.0 242
Women 58.3 39.4 60.6 6.3 16.5 16.7 6430
Total Men 23.9 14.5 85.5 5.4 6.3 2.8 4416
Youth 14.9 10.0 90.0 5.3 3.9 0.8 900
Abbreviation: HP, health post.
Services provided at HPs during the last year were most frequently related to
child vaccination, health education, and ANC. HP delivery, referral to an HC,
and food preparation demonstration were reported by relatively few of the
respondents who had had at least 1 visit to an HP.
4.3.3 Services provided through outreach sessions

Respondents were asked whether they had ever received health education
or other HEP services in settings other than an HP or their homes. Lifetime
exposure to health education or other HEP services in a setting other than the
HP or home was 19.6%, 26.2%, and 19.5% among women, men, and youth girls,
respectively. Exposure to such a service during the 1-year period preceding
the study was 12.3%, 16.8%, and 12.2% among women, men, and youth girls,
respectively. Health education was the most common service reported by all
categories of respondents. Child vaccination, deworming, and TT vaccination
were also among the services received by respondents in outreach settings
(Table 4-7)
page-180
Table 4-7. Health education or other services received in settings other than home or HP, by respondent
category and livelihood
Proportion of
respondents
who received
health
Among respondents who received health
education or
education or other services in settings other than
other HEP
home and health post, percentage who received:
services in a
Unweighted
setting other Unweighted #
Livelihood Respondent # of
than HP and of respondents
respondents
Home during:
Life Last 1
time year
Health
education
Child
vaccination
Deworming
TT
vaccination
Growth
monitoring
Supplemental
food
Trachoma
prevention
and control
Women 20.1 12.6 4421 78.9 38.7 22.1 19.6 17.4 12.0 2.8 689
Agrarian Men 26.8 17.2 3157 90.0 0.0 30.3 0.0 0.0 0.0 1.4 615
Youth girls 19.9 12.4 658 52.3 0.0 29.0 0.0 0.0 0.0 1.8 108
Women 7.0 5.0 2009 68.2 50.6 29.6 33.8 23.6 11.1 0.0 143
Pastoralist Men 8.8 6.8 1259 91.8 0.0 41.4 0.0 0.0 0.0 0.0 102
Youth girls 7.1 5.9 242 95.8 0.0 50.7 0.0 0.0 0.0 0.0 17
Women 19.6 12.3 6430 78.7 38.9 22.2 19.9 17.5 12.0 2.8 832
Total Men 26.2 16.8 4416 90.0 0.0 30.5 0.0 0.0 0.0 1.4 717
Youth girls 19.5 12.2 900 53.0 0.0 29.3 0.0 0.0 0.0 1.7 125
Abbreviations: HEP, Health Extension Program; HP, health post; TT, tetanus-toxoid.
page- 181

4.3.4 Overall exposure of households to the HEP
Households’ overall exposure to the HEP was determined by assessing whether

a household member had exposure to the HEP through any of the 3 service
delivery modalities: (a) at home, (b) at an HP, or (c) at any other place (e.g.,
during outreach) during the 1-year period preceding the study. A household
member was considered to have been exposed to the HEP if he or she met any
of the following criteria:
1. At least 1 home visit during the 1-year period preceding the study, with
respondent her- or himself meeting the HEW during the visit;
2. At least 1 HP visit during the 1-year period preceding the study; or
3. At least 1 exposure to health education or other HEP services provided
in a setting other than HP or home.
A household was considered to have been exposed to the HEP if at least 1

respondent from the household (woman, man, or youth girl) had exposure to
the HEP during the 1-year period preceding the survey. Accordingly, 61.8% of
households were exposed to the HEP through at least 1 of their members. This
rate ranged from 62.7% in agrarian settings to 27.1% in pastoralist settings. In
only 26.8% of households were all respondents within a household exposed to
the HEP (Figure 4-4).
Figure 4-4. Exposure of household members to the HEP, by region
Abbreviations: HEP, Health Extension Program; HEW, Health Extension Worker; Ben-Gum, Benishangul-
Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page-182
The major source of exposure to the HEP in both agrarian and pastoralist
settings was an HP, followed by, for women, a home visit. The role of outreach
sessions was relatively larger for men and youth girls in agrarian areas than
pastoralist ones, while the HP remains the major source of exposure to the HEP
among men and youth girls in pastoralist settings (Figure 4-5, Table 4-8, Table
4-9, Table 4-10).
Figure 4-5. Source of exposure to the HEP among women, men, and youth girls
Abbreviations: HEW, Health Extension Worker; HP, health post.
page- 183
Table 4-8. Exposure of respondents to the HEP – Women
Proportion of women who had at least 1 direct

interaction with an HEW through: Unweighted
Health Any of number of
Home All 3 respondents
post Outreach the 3
visit modalities
visit modalities
National 27.5 39.4 12.3 54.8 3.9 6430
Agrarian 28.2 39.7 12.6 55.5 4.0 4421
Livelihood
Pastoralist 12.4 34.9 5.0 40.3 1.7 2009
Tigray 37.6 35.3 20.8 63.0 5.2 607
Afar 21.8 63.9 5.4 67.1 3.5 399
Amhara 38.1 41.6 17.3 61.5 5.4 1060
Oromia 21.0 34.0 6.0 47.2 1.1 1319
Region Somali 2.8 20.1 0.7 21.4 0.2 798
Ben-Gum 72.6 69.1 22.4 90.7 14.1 406
SNNPR 27.6 50.9 18.7 63.6 8.2 1009
Gambela 41.1 57.3 16.7 68.2 8.5 417
Harari 23.1 29.0 6.8 42.6 2.8 415
15-19 20.6 45.2 8.2 55.7 0.6 281
20-24 26.4 57.4 10.8 67.9 2.7 763
25-29 29.4 56.9 15.1 67.9 6.2 1209
30-34 28.0 52.9 10.7 64.7 4.6 943
Age 35-39 32.9 45.8 12.9 61.6 4.5 923
category of
respondent 40-44 29.7 28.2 13.2 47.8 4.3 519
45-49 32.7 23.2 17.8 47.0 6.3 353
50-54 21.8 17.6 13.0 36.5 2.9 563
55-59 23.2 12.4 8.8 32.8 1.8 340
60+ 20.7 10.3 7.7 30.1 0.1 536
No formal
24.8 33.9 10.8 49.8 2.6 4809
education
Educational Grade 1-4 30.6 48.9 15.0 62.7 6.3 735
status Grade 5-8 33.7 54.2 14.8 69.1 6.5 604
Grade 9 or
48.7 62.6 21.0 76.0 10.8 282
above
page-184
Currently
27.7 42.2 12.1 56.8 4.1 5162
married
Marital Divorced 30.4 35.9 19.7 51.6 6.6 297
status of Widowed 27.1 19.0 10.9 42.7 1.1 721
respondent Separated 17.6 19.5 8.8 31.7 1.8 201
Never
31.2 44.2 37.7 54.3 20.6 49
married
Lowest 20.1 39.9 10.4 52.1 2.0 1343
Lower 26.0 42.2 12.2 55.6 4.2 1291
Wealth
Middle 29.3 39.9 13.8 57.9 3.6 1278
Quintile
Higher 28.0 38.1 12.2 53.8 4.4 1274
Highest 33.1 37.6 12.5 54.5 5.0 1244
Abbreviations: HEW, Health Extension Worker; HEP, Health Extension Program; Ben-Gum, Benishangul-
Table 4-9. Exposure of respondents to the HEP – Men
Proportion of men who had at least 1 direct

Any of number of
Home Health All 3 respondents
Outreach the 3
visit post visit modalities
modalities
National 10.7 14.5 16.8 32.1 1.9 4416
Agrarian 10.8 14.4 17.2 32.4 1.9 3157
Livelihood
Pastoralist 5.9 16.3 6.8 24.2 0.8 1259
Tigray 19.0 22.6 37.4 56.1 2.9 407
Afar 5.3 50.8 8.6 53.9 1.8 275
Amhara 14.1 16.6 23.4 40.5 2.8 603
Oromia 8.3 10.5 9.7 22.9 1.0 1139
Region Somali 0.9 13.7 0.8 14.2 0.2 376
Ben-Gum 50.2 50.2 20.4 77.7 7.7 340
SNNPR 11.4 19.0 24.0 40.8 2.9 759
Gambela 17.3 40.4 12.6 47.9 3.0 157
Harari 9.8 13.5 4.8 20.3 1.2 360
page- 185
15-19 10.6 14.4 0.1 23.6 0.0 23

20-24 4.0 21.9 19.1 37.4 0.0 211
25-29 6.0 11.5 14.3 27.4 0.3 565
30-34 9.1 15.1 12.0 28.8 1.8 660
Age 35-39 8.7 12.2 14.4 25.8 1.5 635
category of
40-44 15.4 16.1 18.9 39.1 2.4 619
respondent
45-49 12.6 19.2 23.4 37.6 3.2 417
50-54 11.5 14.2 15.6 30.1 2.0 331
55-59 18.6 17.3 28.0 42.5 5.3 234
60+ 9.4 11.5 15.4 30.3 1.1 721
No formal
10.0 10.2 13.6 26.6 1.1 2475
education

status Grade 5-8 13.0 19.0 21.4 39.5 3.1 781
Grade 9
8.0 16.8 15.1 32.3 1.8 478
or above
Currently
10.6 14.3 16.7 31.9 1.8 4343
married
Never
0.0 13.2 17.6 30.8 0.0 24
married
Lowest 8.4 10.9 13.8 25.8 1.5 708
Lower 8.7 13.9 17.8 30.6 2.2 843
Wealth
Middle 10.5 13.9 21.0 35.3 1.4 881
Quintile
Higher 10.4 13.2 16.1 31.7 1.2 1028
Highest 14.0 19.1 14.7 34.7 3.1 956
page-186
Table 4-10. Exposure of respondents to the HEP – Youth girls
Proportion of youth girls who had at least 1 direct

Any of number of
Home Health Other All 3 respondents
the 3
visit post visit settings modalities
modalities
National 5.6 10.0 12.2 21.9 1.1 900
Agrarian 5.7 9.9 12.4 22.1 1.1 658
Livelihood
Pastoralist 3.7 12.1 5.9 16.7 0.3 242
Tigray 8.3 11.7 22.9 37.5 0.0 117
Afar 11.8 37.1 23.5 44.7 2.3 61
Amhara 5.1 14.9 16.0 27.3 1.4 169
Oromia 4.5 3.5 5.7 12.3 0.0 162
Region Somali 0.0 7.8 1.4 8.5 0.0 97
Ben-Gum 23.2 25.8 16.4 50.8 4.0 50
SNNPR 7.3 13.5 15.6 26.7 2.9 150
Gambela 38.4 44.7 20.0 58.0 10.1 36
Harari 17.8 17.9 18.7 35.6 1.8 58
Age 15-19 5.2 8.9 11.3 21.1 0.9 720
category of
respondent 20-24 6.9 13.9 15.3 24.9 1.7 180
No formal
0.5 12.7 6.1 14.0 0.0 169
education
status Grade 5-8 5.7 9.4 14.0 23.3 1.1 369
Grade 9 or
7.4 11.6 15.7 25.8 2.1 218
above
Currently
13.2 33.1 21.5 43.0 4.6 75
married
Never
5.1 7.6 10.8 19.7 0.7 777
married
Lowest 3.9 7.6 9.7 19.1 0.0 143
Lower 6.0 13.3 12.5 27.9 0.0 157
Wealth
Middle 2.9 10.8 13.7 19.7 1.2 190
Quintile
Higher 7.5 9.3 11.2 20.9 2.3 191
Highest 6.4 9.2 13.0 22.7 0.9 219
Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples Region
page- 187
.
4.3.5 Household level adoption of HEP-targeted behavior

Successful implementation of the HEP is expected to achieve behavioral
change observable at both the household and community levels. Household-
level implementation of the HEP was assessed using a set of criteria related
to the desired household-level behaviors (Table 4-11). Households were first
assessed for their eligibility for each criterion and then whether they met the
criteria for which they were eligible. Progress toward full implementation of the
HEP at the household level was then determined by dividing the number of
criteria met by the number of criteria for which a household was eligible.
Table 4-11. Criteria for assessment of the HEP implementation

at household level
Assessment checklist Less stringent criteria More stringent criteria
1. Antenatal care At least 1 visit At least 4 visits

2. Place of delivery Health facility
3. Postnatal care Within 1 week
4. Family planning Any method – ever use Long-acting – ever use
5. Child vaccination Complete by 1st birthday
6. Growth monitoring For all <2 years children
7. Latrine Any type With handwashing facility
Observed for:
8. Personal hygiene - Hand and face

- Clothes
- Shoes/sandals
9. Shower Shower room/place
10. Housing Observed for clearness
11. Solid waste disposal Pit
12. Solid waste disposal Pit
13. Livestock Separate from living room
14. Kitchen Separate from living room
15. Malaria control Participation of any
activities household member
16. Spray Spray and do not paint
17. ITN Use by all household members
Abbreviations: HEP, Health Extension Program; ITN, insecticide-treated net.
page-188
The average level of adoption of the HEP at the household level was 50.8%
when less stringent criteria were used and 40.9% when more stringent criteria
were used. Households of WDA leaders were in general better at implementing
the HEP; their progress, however, was much lower than was expected from a
group of women selected as models and community mobilizers (Figure 4-6).
Figure 4-6. Household-level implementation of the HEP among regular and

WDA households
Abbreviations: HEP, Health Extension Program; WDA, Women’s Development Army; SMC, Social
Mobilization Committee; HH, household.
Relatively better progress was observed in agrarian settings than pastoralist

settings. Household-level implementation of the HEP was highest in Benishangul-
Gumuz and lowest in Afar. Progress toward full implementation of the HEP at
the household level was found to increase with higher educational status and
wealth quintile (Table 4-12).
page- 189
Table 4-12. Household-level implementation of the HEP
page-190
Progress towards full implementation of HEP at household level on less stringent and more stringent criteria
Regular Households WDA/SMC
Progress on
Progress on less Progress on more # of Progress on more
less stringent # of households
stringent criteria stringent criteria Households stringent criteria
criteria (Unweighted)
(Unweighted)
Mean (95% CI) Mean (95% CI) Mean (95% CI) Mean (95% CI)
National 50.8 (50.6, 51.0) 40.9 (40.7, 41.1) 6430 60.6 (59.9, 61.3) 50.9 (50.3, 51.6) 613
Agrarian 51.3 (51.1, 51.6) 41.2 (41.0, 41.5) 4421 61.5 (60.8, 62.2) 51.7 (51.0, 52.3) 400
Livelihood

Pastoralist 38.4 (37.5, 39.3) 33.0 (32.1, 33.8) 2009 41.9 (39.2, 44.6) 34.7 (32.3, 37.1) 213
Tigray 53.7 (52.8, 54.6) 45.8 (45.0, 46.5) 607 63.3 (61.1, 65.5) 54.1 (52.3, 56.0) 71
Afar 39.0 (36.4, 41.7) 33.3 (31.0, 35.6) 399 33.1 (21.8, 44.3) 29.2 (19.8, 38.6) 35
Amhara 57.7 (57.3, 58.1) 46.6 (46.2, 47.1) 1060 70.9 (68.8, 72.9) 60.1 (58.1, 62.2) 71
Oromia 49.9 (49.6, 50.2) 41.1 (40.8, 41.4) 1319 58.9 (57.9, 59.8) 49.6 (48.8, 50.5) 152
Somali 39.2 (38.1, 40.4) 34.6 (33.4, 35.9) 798 39.4 (36.4, 42.4) 32.5 (29.7, 35.3) 90
Region Ben-Gum 60.7 (56.8, 64.5) 51.9 (48.2, 55.6) 406 68.8 (57.2, 80.4) 60.6 (48.5, 72.7) 46
SNNPR 44.1 (43.6, 44.6) 32.1 (31.7, 32.6) 1009 59.0 (57.9, 60.2) 48.5 (47.3, 49.7) 112
Gambela 42.0 (33.9, 50.1) 36.2 (28.6, 43.8) 417
57.9 (-236.5,
Harari 46.0 (36.8, 55.3) 38.8 (30.6, 47.1) 415 49.9 (-239.7, 339.6) 36
352.3)
15-19 48.2 (47.1, 49.3) 37.0 (36.0, 38.1) 281 33.5 (15.6, 51.5) 27.3 (11.9, 42.8) 7
20-24 50.2 (49.5, 50.8) 39.2 (38.6, 39.8) 763 68.4 (65.4, 71.4) 61.0 (57.8, 64.3) 38
25-29 52.2 (51.7, 52.8) 41.9 (41.4, 42.4) 1209 62.6 (60.7, 64.4) 53.5 (51.5, 55.4) 95
30-34 50.3 (49.7, 50.9) 39.8 (39.2, 40.4) 943 63.4 (61.8, 65.0) 52.1 (50.7, 53.5) 85
35-39 52.3 (51.7, 52.8) 43.3 (42.7, 43.8) 923 60.9 (59.3, 62.5) 50.1 (48.6, 51.6) 125
Age 40-44 51.3 (50.5, 52.1) 41.2 (40.4, 41.9) 519 57.3 (55.8, 58.8) 47.4 (46.0, 48.8) 91
45-49 52.4 (51.5, 53.3) 43.9 (43.0, 44.7) 353 62.8 (60.4, 65.1) 52.8 (50.7, 54.8) 58
50-54 51.4 (50.7, 52.1) 42.0 (41.3, 42.6) 563 52.6 (50.7, 54.5) 46.4 (44.5, 48.3) 73
55-59 47.3 (46.5, 48.1) 38.7 (37.9, 39.4) 340 64.0 (61.1, 66.9) 54.1 (51.2, 56.9) 29
60+ 47.0 (46.1, 47.9) 37.0 (36.2, 37.8) 536 65.5 (60.4, 70.5) 54.6 (50.1, 59.1) 12
No formal
49.2 (48.9, 49.5) 39.7 (39.4, 39.9) 4809 58.3 (57.3, 59.3) 49.6 (48.7, 50.5) 417
education
Grade 1-4 51.5 (51.0, 52.0) 41.7 (41.2, 42.2) 735 59.4 (57.9, 60.9) 49.1 (47.8, 50.4) 88
Education Grade 5-8 57.4 (56.8, 58.0) 45.8 (45.2, 46.4) 604 66.4 (64.8, 67.9) 54.6 (53.0, 56.3) 80
Grade 9
58.4 (57.5, 59.3) 45.8 (44.9, 46.6) 282 65.0 (63.2, 66.8) 54.6 (52.9, 56.4) 28
or above
Currently
51.1 (50.9, 51.3) 41.4 (41.2, 41.7) 5162 60.3 (59.5, 61.0) 50.7 (50.0, 51.4) 466
married
Divorced 50.5 (49.4, 51.6) 38.8 (37.7, 40.0) 297 68.4 (66.0, 70.9) 61.0 (58.6, 63.4) 47
Marital
status Widowed 47.9 (47.1, 48.7) 37.0 (36.3, 37.7) 721 50.4 (47.3, 53.5) 42.1 (39.4, 44.8) 64
Separated 50.0 (48.1, 51.8) 38.6 (36.9, 40.3) 201 68.0 (65.2, 70.8) 52.8 (49.6, 56.0) 33
Never
50.8 (47.7, 53.9) 38.7 (36.0, 41.3) 49 71.6 (56.3, 86.9) 53.6 (43.8, 63.3) 3
married
Lowest 41.5 (41.0, 42.1) 31.9 (31.4, 32.5) 1343 45.6 (42.5, 48.6) 35.3 (32.5, 38.2) 75
Lower 46.2 (45.7, 46.7) 35.7 (35.2, 36.1) 1291 57.7 (54.9, 60.5) 48.5 (45.8, 51.3) 107
Wealth
Quintile Middle 50.8 (50.4, 51.3) 40.7 (40.3, 41.1) 1278 58.1 (56.6, 59.5) 48.2 (46.9, 49.5) 131
Higher 52.8 (52.3, 53.2) 43.5 (43.1, 44.0) 1274 60.7 (59.5, 61.9) 51.5 (50.3, 52.6) 126
Highest 60.6 (60.2, 61.1) 50.6 (50.2, 51.0) 1244 65.8 (64.8, 66.8) 55.8 (55.0, 56.7) 174
Abbreviations: HEW, Health Extension Worker; HEP, Health Extension Program; WDA, Women’s Development Army; SMC, Social
Mobilization Committee Ben-Gum, Benishangul-Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples Region; CI, confidence
interval.
page- 191

4.3.6 Factors associated with implementation of the HEP
Summary of key findings

• Human resources-related factors are likely to be the primary drivers of
the intensity of HEP implementation.
• The professional mix and level of education, rather than the number of
HEWs in an HP, are associated with better implementation of the HEP
through home and HP visits. HPs with midwives or nurses or level IV
HEWs had better implementation of the HEP in terms of both home
visits and HP visits.
• Population per HP is weakly and negatively associated with the coverage
of home visits; the number of HEWs within an HP (i.e., on the payroll),
however, has no significant association with the coverage of home visits.
• The limited human capacity at the HP level, both in terms of
employee numbers and skill sets, is recognized as a major challenge to
implementation. Attempts have been made to address these challenges
by assigning HC staff to work in and support HPs but have faced the
logistical challenge of the need for repeated travel from HCs to HPs.
Factors associated with the implementation intensity of the HEP were identified
by running separate regression models predicting the proportions of households
reached by the HEP during the prior year through home visits, HP visits, and
outreach sessions. In the first step, bivariate analyses were used to assess the
association between individual covariates and each of the 3 measures of HEP
implementation intensity. Covariates that showed a significant association with
implementation intensity at a 10% level of significance were included in the final
models. Findings from the final model showed that the availability of at least 1
midwife or nurse and of at least 1 level IV HEW were strongly associated with
the coverage of home visits and HP visits. The availability of a midwife or nurse
at an HP was associated with an 11.1% and 22.4% increase in the proportion of
catchment households reached through home visits and HP visits, respectively.
The availability of at least 1 level IV HEW in an HP was also associated with
a 9.4% and 9.5% increase in the proportion of catchment households reached
through home visits and HP visits, respectively (Table 4-13).
page-192
Table 4-13. Factors associated with level of implementation of the HEP

through home visits, HP visits, and outreach sessions
B coefficients for proportion of households reached

with HEP during the last 1 year through:
Covariates
Home visit HP Visit Outreach
Population (in thousands) per HP -1.95* 0.01 0.63
Proportion of villages/gotes within 5

0.07 0.04 0.05
km from HP
Number of infrastructure/facility
2.49* 1.60 0.81
standards met (maximum of 8)
Population (in thousands) per HEW 1.32 0.44 1.05
Availability of at least 1 midwife or

11.06* 22.37* 3.42
nurse
Availability of at least 1 level IV HEW 9.44* 9.50* 2.45
Number of required equipment and

** 0.44 **
supplies available (out of 29 items)
Number of drugs and supplies without
** 0.44 **
stockout in 6 months (out of 20 items)
Abbreviations: HEW, Health Extension Worker; HEP, Health Extension Program; HP, health post.
Potential confounders accounted for: livelihood, formal education, wealth index, median age of
women
* P value <0.05
** P value >0.1 during first step
Evidence from KIIs and FGDs revealed that the human-resources capacity at
the HP level poses a major challenge to the provision of health services through
PHCUs. Assigning HC staff, including nurses and midwives, to regularly visit
and support HPs has become a common practice aimed at improving the HEP’s
implementation. This approach has been criticized, however, by HC staff and
program experts for its inefficiency and logistical challenges. The deployment
of mid-level health workers like clinical nurses was recommended by HC staff
with the argument that it would strengthen health-promotion and disease-
page- 193
prevention activities by HEWs, as it gives them more time to do their jobs. It

was also argued that it would improve the quality of the services provided at
HPs and reduce the workload at HCs. One HC head noted that:
“
In relation to HPs, in another kebele there was a clinical nurse
assigned, and there is a difference in the service quality. Therefore,
it’s better if professionals are also assigned to work with them at
HPs in remote areas. Previously, clinical nurses were assigned. Yes.
They manage and treat emergency cases, like labor and severe
diseases in under-5s. I’m not saying the extension projects are not
doing this, but there is a difference in service quality. But that
is not implemented now; it would be good if we assigned other
professionals to HPs. Our health center has only 7 professionals in
total, and there are 4 kebeles in our catchment. When we send 1
professional to each kebele [to support HPs], we are left with 3 in
the health center. Therefore, the woreda should assign additional
health professionals.…We are treating more than 100 cases per day,
but if we strengthen the Health Extension Program, we can treat
them earlier at the HP. If we can perform the sanitation and other
tasks properly at the kebele level, we can reduce the cases. So, we
should strengthen the activities in the villages with support from
the woreda.
HC Head
A KII with an HEW supervisor in Amhara revealed that improvements in the

qualifications of health professionals at HPs not only meets the standard of the
program but also improves HEWs’ acceptance by the community:
page-194
“
The community considers Health Extension Workers as being only
for preventive activities and tries to compare them with traditional
birth attendants. The community is not confident enough about the
clinical services HEWs give at HPs. This is because the community’s
health-seeking behavior and awareness are increasing due to media
exposure and other sources of information. The community, after
all, considers HEWs as not being health professionals and simply
working on prevention activities like the construction of latrines
and environmental sanitation. The community undermines the
educational status of HEWs. The community needs and searches
for better health services from health centers and hospitals, and
these are the current conditions in our area.
KII, HC Director, Tigray
“
Generally, I can say that the community has confidence in Health
Extension Workers regarding community health activities, such
as vaccinations, care for under-5 children, and the treatment of
malaria. However, the community is less confident in them when it
comes to diagnosis and treatment services at the HP. Because the
HP provides only a few clinical services, most of the society’s clinical
demands are addressed at and referred to HCs and hospitals. I
hope this will be solved by the second generation of the Health
Extension Program.
In a KII with an HC director in Tigray, the key informant suggested that the
professionalization of HP services was critical, even though it may take time
and needs a feasibility study:
page- 195
“
Well, I am aware that the Health Extension Program is using low-
level Health Extension Workers in place of higher-level, better
trained health professionals, and I think that is fine. You know, if
the service is going to be given at the HP, then people need to
be properly trained and qualified to treat minor illnesses in basic
primary healthcare and ensure access to rural areas. I think that
over time the increasing professionalization of the people who
are doing the community-based work will be a natural evolution
based on career advancement opportunities, and all of those are
very important, I think, in the approaches of the Heath Extension
Program. But of course you have got to balance resources and
availability in everything. It needs to be worked out because they
are not easy to achieve.
KII, HC Director, Tigray
Respondents reported desiring the professionalization of HP services, as it

could increase the community’s acceptance of the program and its staff. The
professionalization of HP services is possible not only through the deployment
of new health cadres but also by upgrading current HEWs, who tend to
become fed up with the HEP. The acceptance of the program has declined
since the era when both clients and HEWs were from the community. As a
solution, the HEP should work with the community and convince them of the
advantages of the program, and HEWs should receive support from all their
respective communities and the government. The HEWs have begun to treat
TB at HPs and therefore, should be academically upgraded. They should
alsobe professionally skilled and knowledgeable. The number of HEWs
should beincreased to meet the needs of the communities and decrease their
individualworkloads. HPs should also provide full-time services and be
equipped withbasic equipment and furniture.
page-196
page- 197
CHAPTER 5
Community
Engagement
and Ownership
in HEP Service Delivery
page-198
5. Community Engagement and
CHAPTER 5
Ownership in the Health Extension
Program Service Delivery
INTRODUCTION
The HEP has evolved over time and become deeply rooted in communities
as a program that encourages families to be responsible for their own health.
Many recognize the HEP as an innovative community-based health service
delivery program that targets households.22,25 It is designed to improve the
health status of families with their full cooperation, using local technologies and
the community’s skills and wisdom. The HEP’s underlying philosophy focuses on
building community ownership, responsibility, and the maintenance of health
by transferring health knowledge and skills to parents, especially mothers, in
the communities. The HEP reflects the government’s strategy of community
empowerment, self-reliance, responsibility, and ownership for one’s own health.4
In the operationalization of the HEP, community engagement and ownership

is to be realized through several entry points. The first is the community’s
participation in and contribution to the resourcing of the HEP. In this regard,
communities have actively engaged and made contributions (mostly in kind)
for the construction of HPs.25 The other form of engagement is community
participation in the selection of candidates for HEW. Another form of planned
community engagement and ownership is found in the M&E of the HEP
services and HEWs themselves. Through their representatives, communities are
expected to participate in the governance of the health services provided in
their respective kebeles.
By design, the HEP requires community ownership and engagement to achieve

its purposes. Community engagement is not a package itself, but it is vital to
implementing all the HEP packages for the betterment of society. The HSTP
clearly stipulates 2 strategies of community engagement and ownership: training
model families and mobilizing WDAs, which, because they are networks of
households represented by women, are assumed to be a good strategy for
fostering networking in the community. Five households are networked under 1
leader to form a 1-to-5 network. Five such networks are then further networked
page- 199
Community Engagement and Ownership on HEP Service Delivery
to form what is known as a WDA, which is always led by a woman. At full

coverage, a kebele is expected to have 30-40 WDAs, including a broader
network of 25-30 households. WDA leaders report directly to HEWs on
matters related to health. Both the 1-to-5 network leaders and WDA leaders
are voluntary community members who receive orientation and/or training to
serve as role models in their behavior and to mobilize the members of their
networks to follow a healthier lifestyle. Community engagement in pastoralist
settings relies largely on SMCs, which are expected to support HEWs in the
implementation of the HEP. Unlike WDAs, which focus mostly on health-related
activities, SMCs have a broader role in the governance of multi-sectoral kebele-
level activities, one of which is health.
The HEP National Assessment evaluated the level of coverage of the model
family training sessions and the existence and functionality of community
structures, including WDAs and SMCs.
The provision of MFT has been a major strategy of the HEP since its inception.
The strategy involves intensively training selected households with the purpose
of creating role models for other community members. This is one strategy for
community engagement to improve the community’s health. After completing
the training and implementing the relevant HEP services, women are expected
to play a role in educating their neighbors and relatives.
In recent years, there has been a shift toward a scaling-up strategy that requires
HEWs to educate every household instead of focusing on a few model families.
The HSTP reflects this intention through its goal of graduating 80% of kebeles
as model families. This requires graduation (i.e., the completion of training by
all households in 80% of the kebeles in Ethiopia).
5.1 Awareness of and enrollment in model family

training
The national assessment of the HEP collected information on the awareness

of women regarding model family trainings, enrollment, and the completion
status of households. These data were also collected among an independent
sample of WDA leaders. Findings reveal that MFT reached only a very small
proportion of households. Only 14.9% of women in agrarian settings and 8.0%
page-200
of women in pastoralist settings reported being aware of MFT. Enrollment

and graduation rates were very small, with only 2.9% of agrarian and 2.1% of
pastoralist households reporting having ever been enrolled in MFT. Relatively
higher coverage of training was observed among WDA leaders. Awareness
of MFT was universal among WDA leaders and SMC members. Enrollment
and training completion rates, however, were very low compared to what was
expected from community leaders. Awareness of and enrollment in MFT was
relatively higher in Benishangul-Gumuz and Tigray and lowest in Somali (Figure
5-1, Table 5-1).
Figure 5-1. Model family training among regular and Women’s Development
Army households
Abbreviation: WDA, Women’s Development Army.
page- 201
Table 5-1. Proportion of households’ awareness of and enrollment status in

model family training, as reported by their female representatives
Proportion of households by women’s awareness and enrollment

status in relation to model family training
Regular Households WDA/SMC Households
Background
households
households
Completed
Completed
Enrolled
Enrolled
characteristics
Aware
Aware
# of
# of
National 14.6 2.9 1.9 6430 99.6 42.8 26.1 613
Agrarian 14.9 2.9 1.9 4421 99.6 43.8 26.5 400
Livelihood
Pastoralist 8.0 2.1 1.5 2009 99.3 20.1 18.4 213
Tigray 27.6 5.3 3.4 607 99.8 36.7 18.5 71
Afar 11.6 0.9 0.4 399 99.2 6.5 6.5 35
Amhara 12.7 2.8 1.9 1060 99.6 39.9 38.0 71
Oromia 12.7 2.3 1.8 1319 99.6 40.9 25.5 152
Region Somali 1.1 0.9 0.9 798 99.0 0.0 0.0 90
Ben-Gum 36.0 8.1 4.1 406 99.7 42.3 14.8 46
SNNPR 19.3 3.7 1.8 1009 99.8 54.1 24.2 112
Gambela 15.8 8.0 2.6 417
Harari 10.4 5.3 3.3 415 99.7 46.7 33.0 36
15-19 4.6 0.0 0.0 281 99.0 0.0 0.0 7
20-24 15.5 1.0 0.4 763 99.9 80.3 61.7 38
25-29 15.6 2.3 1.7 1209 99.6 39.5 19.4 95
30-34 14.1 2.6 1.3 943 99.6 39.0 30.8 85
Age 35-39 15.6 3.4 2.3 923 99.6 46.7 26.3 125
category of
women 40-44 15.8 4.8 3.1 519 99.5 36.2 24.6 91
45-49 14.6 4.7 3.8 353 99.7 43.8 15.1 58
50-54 15.4 3.7 2.5 563 99.7 30.7 21.0 73
55-59 15.2 4.0 3.2 340 99.8 56.1 22.0 29
60+ 10.0 1.4 0.9 536 99.8 66.4 66.4 12
No formal
12.9 2.7 2.0 4809 99.6 36.0 23.1 417
education
Educational Grade 1-4 14.6 2.6 1.4 735 99.5 43.6 27.1 88
status of
women Grade 5-8 17.8 3.5 1.5 604 99.7 45.9 28.0 80
Grade 9
35.6 5.3 3.7 282 99.9 71.7 37.0 28
or above
page-202
Currently
14.9 2.7 1.7 5162 99.7 43.7 25.6 466
married
Marital Divorced 16.4 3.9 3.4 297 99.7 48.4 40.2 47

status of Widowed 13.8 4.6 3.5 721 99.5 30.6 18.3 64
women Separated 5.7 1.3 1.3 201 99.4 34.6 22.0 33
Never
2.8 0.0 0.0 49 99.7 73.8 73.8 3
married
Lowest 9.4 1.1 0.4 1343 99.5 24.5 12.9 75
Lower 13.0 2.2 1.4 1291 99.6 32.3 24.3 107
Wealth
Middle 14.1 2.8 2.1 1278 99.6 46.6 33.8 131
Quintile
Higher 15.5 2.6 1.6 1274 99.6 42.0 20.3 126
Highest 20.0 5.4 3.8 1244 99.7 47.7 28.9 174
Abbreviations: WDA, Women’s Development Army; SMC, Social Mobilization Committee; Ben-Gum,
Benishangul-Gumuz; SNNPR, Southern Nations, Nationalities, and Peoples Region.
5.2 Providers of model family training

The training of model families was predominantly provided by HEWs, but
community leaders, including WDA leaders, 1-to-5 network leaders, and women
from other model families were also reported to have participated in training
model families (Figure 5-2).
Figure 5-2. Trainers of model families
Abbreviation: HEP, Health Extension Program.
page- 203
5.2.1 The role of model families
Women who completed MFT were asked whether and how they had participated
in educating their communities on health-related topics. More than half (52.4%)
of women who completed the MFT program reported playing some role in
educating other families, and about one fourth (23%) reported that they were
engaged in sharing their experiences with other families. The most commonly
reported role of women who had completed model family training was serving
as 1-to-5 network leaders (42%) or WDA leaders (25%; Figure 5-3).
Figure 5-3. Role of women who completed model family training

5.3 Participation of community structures in Health

Extension Program implementation
5.3.1 Availability of community structures supporting the HEP

HPs were assessed regarding whether a community structure existed that
supported the HEP and the functionality of its existing structures. Some form
of community structure supporting the HEP was almost universally reported
page-204
by HPs in both agrarian and pastoralist settings. Community structures in

agrarian settings are WDAs and/or 1-to-5 household networks, while those in
pastoralist settings include both WDA structures and SMCs. WDA structures
were reported to be available in 97.0% of agrarian kebeles. Similarly, 92.5% of
pastoralist HPs reported the availability of either a WDA or 1-to-5 network or
an SMC structure that supported the HEP in their respective kebeles (Figure
5-4). In kebeles reporting having WDA structures, the median number of
households per WDA was 32 (IQR: 21-85) in agrarian settings and 37 (IQR:
29-50) in pastoralist settings. The number of households per WDA is close to
the recommended standard of 25 to 30 households per WDA, although there
is considerable variability among kebeles.
Despite the widespread availability of at least 1 WDA structure per kebele,

coverage at the household level is very low. According to the guidelines for
the organization of WDA structures, it is expected that a kebele with 1 000
households will have 30-40 WDA structures. Findings from this assessment
showed that only 25.9% kebeles have a WDA density of more than 30 per
1 000 households (Figure 5-5).
Figure 5-4. Types of community structures supporting the Health Extension

Program, by livelihood
page- 205
Figure 5-5. WDA density, by livelihood
Abbreviations: WDA, Women’s Development Army; hh, household.
5.3.2 Functionality of community structures in agrarian settings

The functionality of the WDA structure at the kebele level was assessed using
the following minimum set of criteria:
1. HP reports the availability of at least 1 functional WDA structure;
2. HP reports the availability of at least 1 functional 1-to-5 network;
3. existence of a list of WDA leaders;
4. existence of any evidence of a work plan for at least 1 WDA; or
5. existence of any evidence of the performance/activities performed by at
least 1 WDA.
The assessment of the existence of at least 1 functional community structure in
support of the HEP based on the above criteria indicates that the functionality
of community structures is a major gap. Despite the universal availability of
at least 1 WDA structure at the kebele level, only 21.5% of HPs in agrarian
settings had at least 1 WDA structure that met the minimum functionality
criteria (Figure 5-6).
page-206
Figure 5-6. Functionality of WDA structures in agrarian settings

Abbreviations: WDA, Women’s Development Army; SMC, Social Mobilization Committee; HP, health
post.
5.3.3 Functionality of community structures in pastoralist settings

Community structures supporting the HEP in pastoralist settings include both
WDA structures and SMCs, which in most of the pastoralist kebeles comprised
different categories of community members, including women’s league leaders,
kebele leaders, religious leaders, and teachers. HEWs were the most commonly
represented group (Figure 5-7).
The existence of at least 1 functional WDA structure at the kebele level was
assessed using the following minimum set of criteria:
1. HP reports the availability of at least 1 functional WDA structure;
2. HP reports the availability of at least 1 functional 1-to-5 network;
3. existence of a list of WDA leaders;
4. existence of any evidence of a work plan for at least 1 WDA; or
5. existence of any evidence of the performance/activities performed by at
least 1 WDA.
For pastoralist settings where SMCs existed, their existence was indicated
by fulfillment of either the previous criteria for the availability of at least 1
functional WDA structure or the following criteria:
page- 207
1. HP declaration that the SMC is functional;

2. HP possession of a mechanism to monitor the performance of the
SMC; or
3. existence of any evidence of an SMC review of their performance on
health-related activities.
The findings show that very few community structures in pastoralist settings
were functional in terms of supporting the HEP. Only 12% of HPs in pastoralist
settings met the minimum criteria for WDA/SMC functionality (either a
functional SMC or at least 1 functional WDA structure).
Figure 5-7. Composition of social mobilization committees in pastoralist kebeles

Abbreviations: HEW, Health Extension Worker; TBA, traditional birth attendant; CHW, community
health worker.
The FGDs and KIIs identified various ways community members engaged with
the implementation of the HEP. Among the most frequently reported means of
community engagement were: cash, voluntary work, and in-kind contributions
for the construction of HPs; building communal latrines; repairing roads for
ambulance services; and constructing HEWs’ residences. Engaging the WDA/
SMCs and 1-to-5 networks has also been identified as a mechanism through
which communities have supported the HEP.
page-208
Active WDA structures were reported to have engaged in different activities

at the household and community levels, including the regular coordination
of community dialogue sessions, both with the guidance of HEWs and by
themselves. FGD participants also reported that the WDAs are considered
health ambassadors: primary health professionals who serve as a bridge
between HEWs and the community and who participate in solving social and
health problems. A community member in an FGD described the role of WDAs
“
as a bridge in the HEP:
The main assembly point for Health Extension Workers and the
community is the development army. The development army is
the best way to implement every activity of the Health Extension
Program, and it is the key means for the success of the health
extension package.
FGD, Male Community Members, Tigray
WDAs’ support was also reported to help reduce the HEWs’ burden, particularly for
the following tasks: promoting environmental hygiene; preventing FGM; reminding
pregnant woman to attend ANC; and identifying, registering and reporting problems
in children and pregnant women to HEWs. WDAs were also reported to support
HEWs by mobilizing communities for vaccination and other outreach and campaign-
based services. One community member explained the role of WDA leaders as follows:
“
Yes, they transmit health messages. They mobilize the community
for immunization and call pregnant mothers for pregnant women’s
conferences. WDAs know who is pregnant in their communities; they
advise them to get ANC and use maternal waiting rooms. They tell
mothers to get vaccinations, which is important for the newborn.
They teach them to share responsibilities with their husband,
because a job burden could lead to preterm delivery. Moreover,
preterm delivery will lead to further costs, as preterm babies need
special care in hot rooms; the mother will also be admitted there.
Therefore, prevention is best. This is how WDAs assist HEWs.
page- 209
The informants also reported that some WDAs had served for a long period by
planning and discussing with and reporting to the HEWs. To create a sense of
ownership among the community, WDA members were selected in public, with
the community considering seriously the candidates’ willingness to serve. After
the establishment of a WDA in the community, women became more aware of
opportunities to participate in HEP activities. HEWs reported that there were
WDA leaders who worked with them and provided services to their respective
groups and who could also function in the absence of HEWs. In pastoralist
societies, together with WDA leaders, men also support HEWs.
5.4 Characteristics of Women’s Development Army leaders
WDA leaders are expected to support the implementation of the HEP by

providing health messages and demonstrating exemplary behaviors. Selecting
WDA leaders is, therefore, an important determinant of their success in
influencing other community members. Comparing the characteristics of WDA
leaders with those of women in the general population revealed that WDA
leaders are more likely to have better health behaviors, higher educational
status, and higher socio-economic status. There remain, however, large gaps in
the health behavior of WDA leaders, given the model role they are expected
to play (Figure 5-8, Figure 5-9, Table 5-2).
Figure 5-8. Educational status of WDA leaders compared with women from the
general population
Abbreviations: WDA, Women’s Development Army; SMC, Social Mobilization Committee.
page-210
Figure 5-9. Wealth quintiles of WDA leaders compared with women from the
general population
Abbreviations: WDA, Women’s Development Army; SMC, Social Mobilization Committee.
Compared to women in the general population, WDA leaders had higher

patterns of use of the health services on some indicators. There were, however,
areas of service use where WDA leaders did not perform better. In general,
WDA leaders did not have service use rates as high as expected from a group
of women who are meant to serve as role models (Table 5-2).
Table 5-2. Comparison of health behavior indicators between women in the

general population and WDA Leaders
Indicator Regular Women WDA Leaders

CPR 44.6% 50.8%
ANC-4 47.9% 62.0%
Facility delivery 55.1% 52.2%
Postnatal care (at least 1 visit) 25.5% 15.5%
Full basic vaccine 35.7% 42.7%
Latrine availability 71.5% 84.2%
Abbreviations: CPR, contraceptive prevalence rate; ANC-4, 4th antenatal visit.
page- 211
The low levels of the targeted HEP behaviors among WDA leaders, compared
to what was expected from them, may be related to the process of selecting
WDA leaders and/or the adequacy of the training or orientation provided to
them. Findings from the qualitative research indicate that the selection of WDA
leaders is mostly the collective responsibility of HEWs, kebele administrators, the
Women’s Affairs office, and the Women’s Forum. The criteria most frequently
mentioned for the selection of WDA leaders, however, were communication
skills, education, community acceptance, a close relationship with the community,
and a willingness to serve as models for others in their health behaviors.
FGD participants and key informants identified a number of challenges related

to community engagement and ownership. The following challenges were mostly
related to (a) volunteerism vs. the expectation of incentives and (b) acceptability-
related issues: conflicts of interest among WDA leaders, acceptability and a lack
of motivation, being perceived as politicians, the lack of service to the expected
level, the lack of follow-up, the lack of incentives and recognition, the lack of
capacity, and poor attention from the government. Moreover, training large
groups of WDA leaders at once has led to compromised quality of training
limiting their ability to deliver their expected roles as community health agents.
In pastoralist society, HPs are often located far from the community; this access
is a challenge for WDA leaders. Men are often reluctant for their wives to serve
as WDA leaders. Cultural resistance to the WDA was also mentioned as a
challenge. WDA leaders expressed disappointment that people perceived them
as working in their own interest. One WDA leader in a pastoralist community
put the challenge as follows:
“
People complain/accuse us that we are not working for them,
but rather for our own benefit….We are deeply concerned for the
people, but they think we are only interested in the per diem….It
is very disappointing, and I have been thinking about resigning
instead of always feeling disappointed….Some understand that we
are benefitting the people, but others think we are self-centered….
The people’s words create hatred and drive you to quit.
FGD, Pastoralist WDA Leaders, Oromia
page-212
page- 213
CHAPTER 6
Information
System and
M&E in the
HEP
page-214
6.Information System and Monitoring
CHAPTER 6
& Evaluation in the Health Extension
Program
INTRODUCTION
M&E is an integral part of the health system. It involves a review and reflection
of the system inputs, processes, outputs, and outcomes of the health system’s
capacity and performance. M&E in the HEP includes both technical and
managerial functions and purposes. Setting goals and defining clear objectives,
targets, inputs, outputs, indicators, and program activities with a good leadership
and governance system are the key characteristics of effective M&E functions.
The Community Health Information System (CHIS) is an information system

especially designed to help manage community health programs, including
the provision of care at HPs. Its purpose is to support decision-making at
the grassroots level. The CHIS is organized around the family folder (FF), a
family‐centered tool designed to help HEWs manage and monitor their work
in educating households and delivering an integrated package of promotional,
preventive, and basic curative health service to families.
The CHIS was first piloted by the Dire Dawa city administration in 2009.
Modification of the system for the agrarian and semi-agrarian contexts was
done in the Awi and Hadiya zones in Amhara and the SNNPR, respectively,
from April to August 2010. According to MoH reports, the CHIS has been
introduced in all 4 largest regions, with the exception of a few pastoralist
woredas in Oromia and the SNNPR. CHIS operationalization was delayed
in pastoralist regions until 2016. Following this, the MoH reviewed different
experiences with CHIS implementation in the pastoralist areas of Afar and
Oromia and, informed by this review, collaborated with partners to design a
data management system for pastoralist communities. In those regions, the
CHIS uses a simplified integrated service recording tool comprising registers, a
tally sheet, and patient cards. Unlike the agrarian CHIS, this simplified system
does not provide for family-centered services and includes neither the household
registration nor the kebele profiling components. The implementation of this
system was begun in March 2017 in the smaller regions and in selected zones
of Oromia and the SNNPR, where pastoralist communities reside.
page- 215
Information System and M&E in HEP

• The CHIS captures only a small portion of the functions of HEP. Currently
reportable indicators of the HEP focus only on outputs, with little chance
for HEWs to track their service processes.
• The HEP HIS was assessed based on 6 functional areas. While none of
these areas had full strength, relative or partial strengths were observed
in the areas of M&E structure, data management, and the availability
of inputs.
• Data accuracy was checked for 10 tracer indicators. The average
verification factor (VF) of all indicators at the national and sub-national
levels, except for skilled delivery, showed over-reporting.
• Nearly one third of HPs over-reported all 10 tracer indicators. Under-
reporting was observed in about 10% of HPs. In total, about 40% of HPs
sent inaccurate data to the next highest level.
• Although most HPs received supportive supervision from catchment HCs,
the support did not meet the standard in terms of frequency, the use of
the checklist, or the provision of feedback or follow-up.
• Across the regions, significant system weaknesses were detected in
information use and data use culture.
6.1 Information system

6.1.1 Adequacy of the HEP HIS/M&E
Given the important role of the CHIS in managing and monitoring the HEP,
it is important to assess its adequacy and the quality and use of the data it
produces, along with other characteristics, such as design, simplicity, and cost.
ADEQUACY OF DESIGN
The CHIS in the rural HEP is designed in line with the package, using family-
centered tools and health cards. About half of the HEP package is on
environmental health and sanitation or health education. As a result, HEWs
spend a significant amount of their time on these services. The CHIS, however,
captures only a small part of HEWs’ routine activities, particularly those related
to environmental health, sanitation, and health education.
page-216
To successfully implement the HEP package, it is recommended that HEWs

work closely with the community development team and 1-to-5 networks. Among
others, the community development team is responsible for collecting and
organizing health-related data, organizing and conducting weekly meetings
to review the progress of the implementation, and preparing and submitting
reports to the HEWs. On the other hand, HEWs are responsible for M&E of the
development teams’ performance. The CHIS focuses only on the activities of the
HEWs and does not have tools to capture the activities of the community team
or the links between the community team and the HEWs. Therefore, the CHIS
is not helpful in monitoring the performance of the community development
team.
ADEQUACY OF INDICATORS
Indicators are set of measures that indicate how a country’s health profile is
changing over time. Having a nationally agreed-upon set of core indicators
are recommended as part of a strong HIS. According to the reformed HMIS,
Ethiopia has a core set of 108 indicators, with subsequent revisions conducted in
2014 and 2017 to meet the data demand for new programs as well as program
modifications. To date, a total of 131 sector-wide indicators have been selected
for use. Of these indicators, about 50 are collected from HPs. Although the
current revision includes most of the key indicators needed at the HP level, a
few important process indicators for the planning and evaluation of kebele
health activities are missing, including: the identification of pregnant women,
the identification of sick newborns, home births, births attended by attendants
other than skilled attendants or HEWs, birth notification, and indicators related
to obstetric and newborn complications.
The recording and reporting tools for HCs have been revised and introduced,
while the HP-level tools have been delayed. Therefore, during the assessment
period, as depicted in the graph below, only 31 of the 51 indicators have data
sources in the existing CHIS, while the remaining 20 indicators have no data
source. These include the following missing indicators: Pneumococcal Conjugate
Vaccine (PCV), Rota, Inactivated polio vaccine (IPV), human papilloma virus
(HPV), and Measles 2 immunizations; iron supplementation for 90 or more days;
nutritional screening for pregnant and lactating women; nutritional screening in
under-5 children; Outpatient Therapeutic feeding Program (OTP) admission
and discharge; and latrines. This finding is also supported by this assessment,
which identified the information system as inadequate for collecting all the
needed indicators. It has been reported that HPs are using non-CHIS recording
and reporting tools due to the fact that the WorHOs, HCs, and the HP itself
require additional data not included in the CHIS.
page- 217
Figure 6-1. Number of indicators at national and HP level and indicators with data
sources, by program category
Abbreviations: HP, health post; NTD, Neglected Tropical Disease; DD, diarrheal diseases; TB,
tuberculosis.
In light of the revised indicator list, the gaps identified in the CHIS, and
feedback from users, the MoH recently revised the agrarian CHIS, including
the health cards and tally sheets. It has also added new cards to the system.
The revised system appears more comprehensive than the previous system. The
HEP care packages and membership status of the households in the WDA and
the 1-to-5 networks have become part of the FF. The revised health cards and
integrated maternal, newborn, and child health (MNCH) cards now include
important data elements that were previously missing, such as immunizations,
maternal nutrition, and postnatal services. In addition, the following 4 new
cards are to be included in the FF: hygiene and sanitation card, nutrition card,
integrated communicable and tropical diseases card, and TB treatment and
follow-up card. At the time of writing, RHBs had begun the printing process,
and training-of-trainers (TOT) was underway.
page-218
ADEQUACY OF SIMPLICITY AND COST
As a component of the reformed HMIS, the design of the CHIS followed the
principles of standardization, integration and simplification, which includes
user-friendliness, a minimal data burden, and alignment with the capacity of
the intended users.
Although the FF follows a family-centered approach and helps monitor services

at the individual and household levels, its processes, such as retrieving cards,
filling in information, updating it at least yearly, and updating the cards after
providing outreach services, are time-consuming. According to the time-motion
study conducted in the 4 largest regions in 2015, HEWs spend an average
of 13% of their time on record-keeping, managing FFs, and reporting. The
capacity of HEWs, their workload, and the English-language nature of the
materials affect the user-friendliness of the CHIS. In a 2015 study6 conducted in
the Gurage zone, only two thirds of HEWs had a positive view of the CHIS and
believed that the CHIS tools were easy to use. Likewise, in the same assessment,
many HPs were using non-CHIS tools because they considered them complex,
time-consuming and requiring of more training than their staff had. Overall,
the design of the CHIS should be appropriate to the capacity and workload of
the HEWs, as well as feasible to update periodically.
Compared to registers, the CHIS tools, such as cards, are more costly. Therefore,
there were challenges in revising and printing them. During the 2014 and 2017
indicator revision processes, while new indicators were included and some
older ones were modified, the revision of CHIS tools was delayed until April
2019 because printing and implementation of the revised version was time-
consuming and costly.
6.1.2 Determinants of data availability and information use
The survey assessed the HIS through the lens of the 5 key functional areas
customized from the WHO’s health system assessment tool for HCs and
hospitals. The areas assessed included: (a) M&E structure, capability and
functionalities; (b) inputs or resources; (c) data collection and reporting forms,
tools, and guidelines; (d) data-management process; and (e) information or
data use for improvement. Each of the areas was in turn weighted based
page- 219
on composite sub-components. This section reviews the agrarian CHIS (a FF-

based system), which is being implemented in the 4 largest regions and Harari.
This analysis is based on information gathered from 194 HPs.
M&E STRUCTURE, FUNCTIONS AND CAPABILITY
Since 2016, the MoH has paid considerable attention to the development and
implementation of the CHIS in agrarian regions in order to standardize the
recording and reporting processes, minimize the burden on HEWs, and improve
the quality of data and service. The assessment found, however, that 1 in 4 HPs
did not implement the standard system CHIS. Compared to other regions,
implementation scores were the lowest in Amhara (69.9%), followed by the
SNNPR (75.1%) and Oromia (77.3%; Table 6-1).
Capacity-building is a cornerstone of the effective implementation of the CHIS

and a functional M&E structure. While in-service training is recommended for
all health workers at the HP (including nurses and midwives), one fifth of HEWs
have not received CHIS training. The highest rate of HEWs not receiving CHIS
training were in the SNNPR (37%), followed by Tigray (29%) and Harari
(28%).
The soundness of the M&E structure includes the availability of annual plans and
community involvement during planning and performance review. During the
assessment, an annual plan was available and observed in more than 80% the
assessed HPs, with the exception of the SNNPR (75%). The assessment found
that community involvement is relatively higher during the HEP performance
review than during HP planning (52.5% vs. 36.8%).
page-220
Table 6-1. M&E structure, functions and capabilities of CHIS for the HP survey
Agrarian Regions
Variables
Tigray Amhara Oromia SNNPR Harari Total
CHIS (FF) implementation

Yes 93.3 69.9 77.3 75.1 100 75.5
No 6.7 30.1 22.7 24.9 0.0 24.5
Compile kebele info profile
Yes 96.7 79.9 73.1 53.0 91.3 70.2
No 3.3 20.1 26.9 47.0 8.7 29.8
# of CHIS trained HEWs
None 28.5 5.4 17.8 37.0 28.3 20.3

1 16.8 18.6 25.5 39.2 15.0 27.2
2 37.7 57.7 35.4 21.3 47.5 37.1
3 and above 17.1 18.4 21.3 2.5 9.2 15.4
Availability of HP annual plan (EFY 2011)
Yes, observed 100 90.0 84.1 74.5 100 83.5

Yes, not observed 0.0 10.0 9.4 5.9 0.0 8.3
No 0.0 0.0 6.5 19.6 0.0 8.2
Community involvement in HP planning

Yes 46.7 17.9 41.1 45.0 45.1 36.8
No 53.3 82.1 58.6 55.0 54.9 63.2
Community involvement in performance review of HEP activities

Yes 87.9 66.6 40.9 56.7 9.5 52.9
No 12.1 33.4 59.1 43.3 90.5 47.1
Abbreviations: CHIS, Community Health Information System; HP, health post; FF, family folder; HEW,
Health Extension Worker; EFY, end of fiscal year; HEP, Health Extension Program, SNNPR, Southern
page- 221
Figure 6-2 Percent of HPs implemented CHIS (FF)
Abbreviations: CHIS, Community Health Information System; FF, family folder; SNNPR, Southern
Figure 6-3. Percentage of Health Extension Workers receiving CHIS training

Abbreviations: CHIS, Community Health Information System; SNNPR, Southern Nations, Nationalities,
and Peoples Region.
To summarize the M&E structure and functionality and measure its strength
at the regional and national levels, an M&E structure and functions index
was computed from the above 6 indicators. The assessment revealed that, on
page-222
average, about two thirds (68%) of the components measured in the index
were fulfilled. The highest index scores were observed in Harari and Tigray
(>80%), and the lowest (58%) in the SNNPR (Figure 6-4).
Figure 6-4. M&E Structure and functions index, by region

Abbreviations: M&E, monitoring and evaluation; SNNPR, Southern Nations, Nationalities, and Peoples
Region.
INPUTS OR RESOURCES FOR THE CHIS
The physical resources necessary for CHIS implementation include standard

shelves, a tickler box, and a FF pouch. Although most (>85%) HPs have at
least 1 shelf for the CHIS, 57% have standard shelving, and nearly three fourths
(73.9%) have shelving considered adequate.
A tickler file system is designed to help HEWs monitor patients and easily
trace defaulters. About 80% of the HPs have a tickler box. It is expected that
HPs will update the FF on at least an annual basis. Thus, there should be an
FF pouch for each new household. Less than half (44.5%) of HPs, however,
had an adequate FF or pouch during the assessment period. As designed,
the CHIS requires support from the kebele administration during and after
implementation. Nevertheless, most HPs (76.4%) assessed have not received
such support.
page- 223
Table 6-2. Input or resources for CHIS of the health post survey
Regions
Variables Total
Tigray Amhara Oromia SNNPR Harari
Availability of shelf for CHIS
Yes – standard 96.5 50.7 66.8 39.8 90.5 56.7
Yes – not standard 0.0 39.9 18.7 54.8 9.5 32.8
No 3.5 9.4 14.6 5.5 0.0 10.5
Availability of adequate shelf for CHIS
Yes 93.1 44.2 77.5 92.2 100 73.9

No 6.9 55.8 22.5 7.8 0.0 26.1
Availability of tickler file box
Yes 96.5 82.7 75.0 81.4 73.0 79.3
No 3.5 17.3 25.0 18.6 27.0 20.7
Availability of adequate FF/ pouch
Yes 90.3 36.4 48.4 39.3 73.0 44.5

No 9.7 63.6 51.6 60.7 27.0 55.5
Receive kebele administrative support to strengthen CHIS
Yes 39.9 15.1 25.7 25.7 50.0 23.6

No 61.1 84.9 74.3 74.3 50.0 76.4
Abbreviations: CHIS, Community Health Information System; FF, family folder; SNNPR, Southern
A composite indicator, the input index, was computed to determine whether the
required inputs were available to implement the CHIS as designed. The input
index score at the national level was 63% (i.e., 63% of the required inputs were
observed). Similar to the M&E structure index, higher scores were reported in
Harar (92%), followed by Tigray (83%). Lower scores were computed in Amhara
and the SNNPR, with index scores of 54% and 57%, respectively (Figure 6-5).
page-224
Figure 6-5. CHIS Input or Resource index, by region

Abbreviation: SNNPR, Southern Nations, Nationalities, and Peoples Region.
DATA COLLECTION AND REPORTING FORMS, TOOLS, AND

GUIDELINES
This section covers the availability and use of all the standard data collection
and reporting forms, tools, and guidelines of the CHIS. The design of CHIS
includes the following:
• 3 types of health cards (men’s health card, women’s health card, and
an integrated MNCH card);
• 4 types of tally sheets (service, disease, drug, and FP); and
• 5 types of reports (service, disease, quarterly, annual, and public health
emergency management (PHEM).
In addition, a CHIS recording and reporting manual is given to HEWs during

their training. The CHIS design does not include registers at the HP other than
an iCCM register. The CHIS design also limits reporting to a single channel to
minimize the burden of data collection and reporting and improve its consistency.
CHIS cards are mandatory. They are used to keep records of the clients or
patients who have visited the HP. This assessment also reviewed the adequacy
of the CHIS cards at the HP and found that on average 3 out of 5 HPs had
adequate cards (Table 6-3).
page- 225
The purpose of the tally sheets at the HP is not only to simplify reporting but
also to serve as the sole primary data source for checking the consistency and
accuracy of reporting. High-quality reporting is not feasible without tally sheets
in the agrarian CHIS. A large proportion of HPs lacked service and disease
tally sheets (31.1% and 50.7%, respectively). Availability was lower for other
types of tally sheets. The findings were even lower when assessing the use of
cards and tally sheets. Merely 2 in 5 (39.4%) HPs had evidence of their use
during the assessment period, while the remaining (60.6%) HPs either did not
use the tools or lacked evidence of their use.
HPs should use the standard forms when reporting to higher levels. After several
years of CHIS implementation, a majority of the HPs consistently used the
standard reporting forms. At the national level, 86.7% and 70.3% of HPs had
been using the standard monthly service and disease report forms, respectively.
The least consistently used report (59%) was the annual report (Table 6-3).
Regarding disease classification, the recently revised list of reportable diseases

from HP is used by only half of the HPs. Although the CHIS recording and
reporting manual is given to HEWs during their training and is recommended
to be kept at the HP to be used as a reference, only one third (36.2%) of HPs
had the manual during the assessment.
A significant number (58.7%) of the HPs assessed use non-CHIS registers, and
67.3% of HPs submit non-CHIS reports. Non-CHIS tools were most common in
Amhara. Although CHIS is designed to capture a wide range of HEP services,
common non-CHIS tools found being used during the assessment were related
to FP, malaria, TB, HIV, EPI, and ANC. In addition, non-CHIS reporting forms
were being used for growth monitoring, OTP, deworming, hygiene and sanitation,
health education, WDA performance, community conversation, CBHI, and
some maternal health indicators, such as the identification of pregnant women.
The major reasons given for the use of non-CHIS or alternative data sources
include the following:
• WorHOs and HCs require additional data;
• HPs need additional data;
• the CHIS tools are not easy to understand or use;
• the CHIS does not meet all the requests;
• the CHIS is time-consuming;
• there is a shortage of registers and reporting forms; and
• the HEWs lack training on CHIS.
page-226
Table 6-3. Availability of data collection and reporting forms, tools and
guidelines for CHIS, health post survey
Regions
Variables
Availability of adequate CHIS cards
Health card (Men) 77.0 45.1 65.2 52.3 45.3 57.3
Health card (Women) 80.0 53.2 61.3 57.7 95.3 59.0
Integrated MCH card 75.9 51.5 63.0 61.6 63.5 60.3
Availability of adequate
Tally sheets
Service delivery tally 93.3 76.8 71.7 54.1 27.0 68.9
Disease tally 86.5 53.4 43.9 50.0 27.0 49.3
Drug availability tally 64.1 37.5 18.3 20.0 13.5 24.9
FP dispensing tally 52.0 58.0 25.5 36.7 27.0 37.2
Availability of standard reporting forms, consistently
Service delivery report 100 98.9 86.9 73.8 91.3 86.7

Disease report 91.4 86.7 62.9 65.4 82.5 70.3
Quarterly report 83.5 96.0 63.4 62.9 68.3 71.8
Annual report 73.3 78.1 48.6 57.0 68.3 58.8
PHEM report 97.0 98.9 61.1 55.6 100 70.0
Routine use of standard data collection tools (cards, tally sheets)
Yes – observed 90.1 45.0 39.1 28.4 54.0 39.4

Yes – not observed 3.3 11.6 23.4 29.6 27.8 21.5
No 6.7 43.4 37.5 42.0 18.2 39.0
Use national disease classification
Yes 80.3 55.4 22.4 85.5 81.8 49.3
No 19.7 44.6 77.6 14.5 18.2 50.7
Availability of CHIS recording and reporting manual
Yes 71.0 34.9 40.6 25.4 68.3 36.2

No 29.0 65.1 59.4 74.6 31.7 63.8
page- 227
Use registers other than CHIS
Yes 28.8 90.1 52.4 45.3 35.7 58.7
Non-CHIS registers used
Family planning 58.4 90.3 61.5 23.1 13.3 64.1

iCCM 58.8 86.6 63.2 20.3 62.2 62.8
Malaria 23.5 83.4 13.4 15.5 13.3 39.8
Tuberculosis 59.0 85.4 49.6 20.3 37.8 56.8
HIV 35.2 82.2 21.9 12.6 13.3 42.4
EPI 55.2 76.6 57.8 15.5 24.5 55.9
ANC 57.8 85.8 58.4 15.5 24.5 59.6
Submit reports other than CHIS
Yes 25.8 87.4 58.3 56.6 27.0 63.7
Non-CHIS reports produced
Family planning 25.1 76.6 49.3 37.0 50.0 55.3

iCCM 51.9 91.6 60.0 32.9 67.6 63.8
Malaria 13.4 76.1 21.7 28.5 67.6 41.1
Tuberculosis 38.5 78.1 48.6 26.1 50.0 52.8
HIV 38.5 75.4 45.0 22.3 35.3 49.4
EPI 44.9 64.5 40.1 22.3 50.0 44.0
ANC 39.2 72.0 41.8 24.7 60.7 47.6
Abbreviations: CHIS, Community Health Information System; MCH, maternal-child health; FP, family
planning; iCCM, Integrated Community Case Management; EPI, Expanded Program on Immunization;
ANC, antenatal care; PHEM, Public health emergency management; SNNPR, Southern Nations,
Like the other functional areas, a data collection and reporting tools index
was calculated based on the above indicators. The result revealed that at the
national level only 60% of tools were available at HPs. Regionally, the 2 largest
regions, the SNNPR and Oromia, had 50% of the CHIS tools and forms, while
a large majority (80%) of the tools were in place in Harari and Tigray (Figure
6-6).
page-228
Figure 6-6. CHIS recording and reporting tools and forms index, by region
Abbreviations: CHIS, Community Health Information System; SNNPR, Southern Nations, Nationalities,
and Peoples Region.
DATA MANAGEMENT PROCESS
Data management in CHIS covers all aspects of data handling, from how the FF
is arranged to how data quality is checked and documented. In addition, in this
section we review how the FF is being used and updated. The recommendation
is to arrange the FF by gote and household number. Accordingly, almost all
(92.7%) HPs order the FF this way. More than a fourth of HPs, however, are
not using FFs to keep family health information as designed.
It is expected that the HEWs will update the FF on at least a yearly basis and
update the health cards after each service delivery. Updating the FF following
service delivery is not a routine practice at HPs, however, 38.9% of HPs were
routinely updated after a service delivery event, and about one third of HPs
never update them. The following reasons were given for not updating the FF:
• a shortage of time due to the HEWs’ high workload;

• a shortage of cards or FFs;
• HEWs’ belief that the FF is unsuitable for use and/or unimportant; and
• a lack of knowledge of how to use the FF.
page- 229
In addition, security problems, a lack of commitment, and poor supervision

were other reasons reported.
Lot Quality Assurance Sampling (LQAS) is a recommended self-assessment

technique for checking data consistency at the health facility level, including
HPs. It is expected to be done monthly before a report is sent to the next highest
level. Since the implementation of CHIS, there is evidence that only about 20%
of HPs have ever conducted a data consistency check using LQAS; nearly half
(48.6%) of these HPs had done the LQAS within the last month.
As part of data management, the survey assessed how data sources (registers
and tally sheets) and reports were documented. A larger majority (>80%) of
HPs had copies of the last 3 months’ reports, while about half kept EFY 2011
source documents. A lower level of documentation was reported in the SNNPR
than in other regions.
page-230
Table 6-4. Data management process in CHIS, health post survey
Region
Variables Total
Tigray Amhara Oromia SNNPR Harari
Arrange FF based on Gote & household #
Yes 96.9 100 83.7 96.5 100 92.7

No 3.1 0.0 16.3 3.5 0.0 7.3
Use FF to keep family health info
Yes 96.5 69.0 57.0 96.5 100 71.9
No 3.5 31.0 43.0 3.5 0.0 28.1
Update FF regularly upon service delivery
Yes – always 77.6 45.9 29.6 38.6 58.7 38.9
Yes – Sometimes 19.1 34.7 39.8 11.6 36.5 27.5
Not at all 3.3 19.3 30.6 49.8 4.7 33.6
Ever done DQ check using LQAS
Yes – observed 64.7 27.9 14.2 19.7 31.0 20.7
Yes – not observed 3.3 16.1 13.1 31.8 27.0 18.5
No 32.0 56.0 72.7 48.6 42.1 60.8
Time LQAS conducted
In less than a month time 32.6 25.5 25.0 4.8 0.0 18.4
Last 1 month 57.8 47.5 6.8 33.8 38.4 30.2
2 to 3 months 4.5 12.2 3.4 24.7 23.3 13.4
4 to 6 months 0.0 5.5 19.0 4.2 30.2 11.0
> 6 months 5.1 9.2 45.9 32.4 8.2 29.0
Kept copy of reports sent to HC
Yes – for all 3 months 100 98.4 86.0 62.6 100 83.2
Yes – but not for 3 months 0.0 1.6 9.7 20.8 0.0 10.4
No 0.0 0.0 4.3 16.6 0.0 6.4
Availability of source documents (second quarter of EFY 2011)
Tally sheet 87.9 72.1 43.3 41.1 50.8 51.2
Family folder 93.3 38.0 47.1 48.6 73.0 46.9
Abbreviations: CHIS, Community Health Information System; FF, family folder; DQ, data quality,;
LQAS, Lot Quality Assurance Sampling; EFY, end of fiscal year; SNNPR, Southern Nations,
page- 231
The data management process index was computed based on the above 7
indicators. It revealed that the national index score was 64% (i.e., only 64% of
the data management procedures and activities were practiced). The lowest
index score was reported in Oromia (50%), and the highest was reported in
Harari and Tigray.
Figure 6-7. Data-management process index, by region

6.1.3 Information use
The ultimate objective of the CHIS and the broader HIS is to improve evidence-
based decision-making through local-level data use to improve the health of
the population. As part of the assessment of information use, the participation
of HPs in review meetings at the HC or WorHO was checked. HPs are expected
to participate in PHCU reviews conducted at the HCs on a monthly basis and
in the woreda review meeting at least once per quarter. The findings show that
a large majority (93.7%) of HPs participated in review meetings at either the
HC or the WorHO.
While nearly 90% of HPs have conducted a performance review at some

point, only about 41.7% had conducted a review within the month prior to
the assessment. In the SNNPR, only half of HPs had conducted a review.
The meeting minutes book is one means of verifying whether and when a
health facility has reviewed its performance. Evidence from our assessment of
meeting minutes books showed that slightly more than half (55.3%) of HPs had
page-232
conducted a performance review; the finding is lower in the Oromia and the
SNNPR (44.7% and 38.4%, respectively). Another finding was that about one
fourth (27%) of HPs document their meetings and follow up on the execution
of decisions made at the review meetings.
Information use and dissemination guidelines require HPs to construct and

display tables and graphs for selected key indicators of the HEP. The display is
also to be updated on a monthly basis, once service delivery data is summarized
and ready for reporting. During the assessment, the percentage of HPs with
updated infographics was low overall (46.5%), but relatively high (>80%) in
Amhara and Harari, and notably low in Tigray and the SNNPR (25.6% and
22.6%, respectively).
We calculated an index of information use based on the above 7 components.

Information use was found to be low across the regions, ranging from an index
score of 27% in Oromia to 60% in Harari. The national average index score
was also very low (42%; Figure 6-8, Table 6-5).
Figure 6-8. Information use index, by region
page- 233
Table 6-5. Information use at HPs

Regions
Variables
Participation of HPs in review meetings held at HC or WorHO
Yes – at HC 3.5 26.4 11.7 32.8 19.0 20.6

Yes – at WorHO 3.5 7.8 7.6 23.1 4.7 11.7
Yes – at HC and WorHO 93.1 60.3 78.8 28.9 76.2 61.4
No 0.0 5.5 2.0 15.2 0.0 6.3
Performance review
Ever done performance
95.3 98.4 90.8 55.5 100 89.3
review
There is evidence for actual PR 63.6 76.3 44.7 38.4 73.0 55.3
Duration of recent PR
Less than a month 17.8 12.5 20.9 0.0 21.2 15.8

Last month 35.9 39.2 19.2 17.1 13.8 25.9
2 to 3 months 17.6 22.6 9.3 10.0 15.0 13.9
4 to 6 months 4.3 0.0 7.5 11.4 7.5 5.4
> 6 months 0.0 0.0 6.2 33.2 0.0 7.0
Don’t know 24.4 25.7 37.0 28.4 42.5 31.9
HP document and follow execution of decisions
Yes – documented 13.1 55.6 16.9 0.0 27.0 27.2

Yes – no documentation 45.7 24.3 29.6 82.9 27.0 36.1
No 41.2 20.1 53.5 17.1 46.1 36.6
Implement community score card
Yes 50.2 28.6 27.1 18.9 8.7 26.3

No 49.8 71.4 72.9 81.1 91.3 73.7
Display HEP performance info
Yes – updated 25.6 84.6 42.5 22.6 81.8 46.5

Yes – not updated 9.9 11.0 19.1 45.8 18.2 24.0
No 64.6 4.5 38.5 31.7 0.0 29.5
Abbreviations: HP, health post; HC, health center; WorHO, woreda health office; PR, performance
review; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page-234
6.1.4 Summary of status of the HEP HIS by functional areas
As depicted in the spider graph below, at the national level no single component
of the HIS has achieved a score of 75% or higher. Scores for M&E structure, data
management, and the availability of inputs were highest among all indicators
(index scores of 68.1%, 63.6%, and 63.4%, respectively), while the score for
information use was substantially lower (42%). At the regional level, the 4
largest regions scored low, with the exception of Tigray (>80%). In comparison
to the other regions, the index scores in the SNNPR were the lowest, ranging
from 27% to 58% (Figure 6-9, Figure 6-10).
Figure 6-9. Spider graph, CHIS functional areas, national level
Abbreviations: CHIS, Community health information system; M&E, monitoring and evaluation.
page- 235
Figure 6-10. Spider graph, CHIS functional areas, by region
Abbreviations: CHIS, Community health information system; M&E, monitoring and evaluation.
6.2 Data Quality Check

Data quality should be ensured at all levels of the health system to ensure that
the data used by the health sector for decision-making are timely, complete, and
accurate. The quality of the reported data is dependent on the underlying data
management and reporting system; thus, the MoH emphasizes strengthening
the HIS to produce better-quality data.
Though data quality includes many factors, this assessment considered its 2
main characteristics: (a) data consistency and (b) data accuracy. Assessing
data quality, particularly data accuracy or consistency, can help identify areas
for improvement in the data management and reporting system and help
understand the contribution of the different efforts to strengthen the HIS and
improve data quality in general.
According to the WHO’s RDQA manual, data accuracy will be measured

through computing result verification ratio (RVR) or verification factor (VF).
RVR at the HP is calculated as the proportion of recounted values from original
data sources (including health cards, registers, and tally sheets) over the value
reported to the next highest level (i.e., to the PHCU or WorHO). Likewise, the
page-236
average verification ratio of the HPs was computed as a proportion of the sum
of all recounted values from the original data sources over the sum of values
reported to the next highest level.
According to the Measure Evaluation project data accuracy standard, an

acceptable accuracy level is between 0.9 and 1.1. A verification factor <0.9
(90%) signifies over-reporting, while a VF>1.1 (110%) signifies under-reporting
(Table 6-5).
Table 6-6. Verification factor – standard
Verification factor-standard RVR/VF Interpretation

<0.9 (<90%) Over-reporting
0.9-1.1 (90%-110%) Acceptable range
>1.1 (>110%) Under-reporting
This section of the survey is based on 343 HPs from both agrarian and pastoralist
areas.
DATA QUALITY: RESULT VERIFICATION RATIO
A total of 10 indicators were chosen for the accuracy check. The indicators were
on the topics of maternal and child health and community access and included
the following:
1. # of women receiving contraceptives
1. # of women with 1st ANC visit
2. # of women with 4+ ANC visits
3. # births attended by skilled health personnel
4. # of children <1 y who received Penta 1
5. # of children <1 y who received Penta 3
6. # of children <1 y who received Measles 1st dose
7. # of women who received TT vaccination
8. # of children <5 y who were screened for Sever Acute Malnutrition
(SM)
9. # of model households graduated
Data accuracy for each indicator were calculated as a proportion of the
recounted values from the available data sources over the value reported to
the next higher level.
page- 237
The result verification factor was computed at the regional, sub-national,

agrarian and pastoralist, and national levels. The average VF at the national
and sub-national levels demonstrates that all indicators except skilled delivery
were over-reported; the verification ratios ranged from 0.26 (screening for
acute malnutrition) to 0.84 (model household graduation). This means that
the number recounted from an original data source is less than was reported
to the next-highest level. For example, the aggregated number of women who
received contraceptives, as recounted from the original data source, was 18 522;
the aggregate number sent to the HC in the relevant report, however, was
26 176, making the verification ratio 0.71.
A significant level of variation was noted by region. The highest levels of data
accuracy among maternal health and child health indicators were observed
in Somali and Afar, where the RVR mostly ranged from 0.9 to 1.0. Under-
reporting was common in Benishangul-Gumuz and for a few indicators in
Gambela. Under-reporting means that the HEWs provided some services that
were not reported to the next level.
In general, however, over-reporting was the norm across the 4 largest regions,
particularly in Oromia, where the verification factor was usually below 0.5.
For instance, in one place in Oromia, the recounted number from the original
data source for pregnant women who attended their first ANC was 394, while
the report from that place was 1 571, a value about 4 times higher than the
actual performance. In short, the findings suggest that the HEWs often submit
inaccurate reports.
page-238
Figure 6-11. Result verification ratio at national and sub-national levels, by

indicator
Abbreviations: ANC, antenatal care; HH, household.
Figure 6-12. Result verification ratio of child health indicators at national and
sub-national levels, by indicator
Abbreviation: TT, tetanus-toxoid.
page- 239
Figure 6-13. Result verification ratio at regional level, by maternal health

indicator and region
Abbreviations: ANC, antenatal care; Ben_G, Benishangul-Gumuz; SNNPR, Southern Nations,
Figure 6-14. Result verification ratio at regional level, by child health

indicators
Abbreviations: ANC, antenatal care; Ben_G, Benishangul-Gumuz; SNNPR, Southern Nations,
page-240
Note that at times the aggregate result verification ratio may appear unrealistic,
particularly when there are extreme results (either very high or very low) in the
data set. Thus, to cross-check the result, we also computed the proportion of
HPs that over-reported, under-reported, or reported with an acceptable result
verification ratio. With this calculation, the accuracy level in the majority (52%
to 63%) of HPs was within the acceptable range. A significant proportion
(nearly one third) of HPs, however, were over-reporting. Under-reporting was
not common across the indicators: less than 10% of HPs reported fewer services
than they provided (Table 6-6).
Table 6-7. Health posts’ level of data accuracy, by indicators
% of HPs
Indicators Accurate Under-
Over-report
report report
# of women that received contraceptives 52.8% 37.2% 10.0%
# of women that received ANC first visit 59.7% 30.9% 9.4%
# of women that received 4 ANC visits 62.7% 29.0% 8.2%
# of children <1 y who received Penta 1 dose 60.1% 30.2% 9.6%
# of children <1 y who received Penta 3 dose 60.1% 32.4% 7.5%
# of children <1 y who received measles 1 dose
st
58.8% 35.5% 5.7%
# of women who received TT vaccination 56.5% 35.9% 7.6%
# of children <5 y who were screened for acute
54.2% 39.0% 6.9%
malnutrition
Abbreviations: HP, health post; ANC, antenatal care; TT, tetanus-toxoid.
The assessment also compared the level of accuracy of HPs in agrarian and
pastoralist areas. The result showed that a higher percentage of HPs (60-75%)
in pastoralist areas had accurate data than did those in agrarian areas (50-
60%). The proportion of HPs in agrarian areas that over-reported (30-45%)
was higher than that in pastoralist areas (less than 30%). The finding was
similar across the indicators (Figure 6-12).
page- 241
Figure 6-15. HPs’ level of accuracy, sub-national level, by indicator

Abbreviations: CAR, Contraceptive Acceptance Rate; TT, tetanus-toxoid.
The assessment also assessed the data quality by checking a randomly sampled
case from the tally sheet for the presence of the record in the respective health
cards or other available data sources, such as registers. As shown below, fewer
than 50% of cases had documented evidence in other sources. This means that
a service was recorded on the tally sheet but not in the cards and registers, an
indirect indication of over-reporting (Figure 6-13).
Figure 6-16. Record or data consistency between tally sheet and health cards
or registers, by indicator
Abbreviations: SAM, Severe Acute Malnutrition; ANC, antenatal care; TT, tetanus-toxoid.
page-242
REASONS FOR LOW DATA ACCURACY
The following are possible reasons for low data accuracy. As per this study:
• HEWs believe that report preparation is time-consuming;
• HEWs believe that the report contains redundant reportable data
elements or that the format asks for documentation of irrelevant activities;
• HEWs are not trained on the CHIS tools, hold a related belief that
the report forms are complex, and/or have difficulty understanding the
English language;
• HEWs are unable to update records due to a shortage of recording and
reporting tools, including cards, and the failure of supervisory units to
provide reporting formats to HPs in a timely manner;
• The reportedly high workload of HEWs affects the quality of both their
record-keeping and the services they provide;
• Supervisors have an inadequate level of supervision, a lack of commitment,
and/or limited knowledge of CHIS and provide irregular and untimely
feedback;
• The HC and WorHOs demand additional data because CHIS is not
considered sufficiently comprehensive; or
• The supervising units sometimes produce false reports, forcing HEWs to
produce false reports to fill the gap.
The findings from the qualitative survey also indicate that, despite the fact that
most of the program participants from WorHOs, partner organizations, and
HEWs report using different data-verification mechanisms (i.e., completeness,
timeliness, accuracy), data quality has become an issue even among the partner
organizations who participated in this HEP assessment. Explaining the problem
of over-reporting and the resultant problems in data quality, a HEP supervisor
“
from Tigray said:
We have health professionals who have witnessed few facts: a toilet

was said to have been constructed; children have been vaccinated;
mothers had given birth in the HCs and HPs; etc. In reality, these
may not happen on the ground. This is a clear indication of false
reporting.
KII, HEP Supervisor, Tigray
Another respondent from Tigray explained that data inconsistency results from
the HEWs’ use of non-standardized forms, their inability to compile data using
a standard register, and a failure to record all of the activities performed:
page- 243
“
They [HEWs] follow up their own daily activities separately….The
problem we have is that there is a big difference between their
register and the HMIS register…which we evaluate as a reason
that can create inaccurate data reports. As they stay at work and
become tired, they may not compile the data that they worked on
in the field. We believe that there may also be reports that are not
registered.
KII, HC Head, Tigray
Participants agreed that there were problems related to the quality of reporting,
such as false reporting, lack of indicators for some outcomes, and the use of
“
inaccurate denominators while calculating some indexes. One participant said:
For BCC, we don’t have that indicator, but we have asked the
government to have that because they are working on prevention
and demand creation….But if you see the indicators, they are
meaningless….For example, they say that we have reached 3563
communities in this quarter….How did they count it?…To add about
the denominator accuracy…they tell you they vaccinated 120%....
The denominator they use is based on a projection. So, they have
to use the census data….They want to be accurate, but HEWs are
not allowed to use their own data.
FGD, Federal-Level Partners
6.3 Monitoring and evaluation

This section reviews the findings regarding the following processes: HP planning,
supervision and feedback, performance reviews, and the involvement of
community members in the overall monitoring and evaluation (M&E) processes
of the HP or kebele.
6.3.1 Planning and implementation
The existence of a health facility plan is a fundamental requirement for monitoring

and evaluating the HEP’s performance and thereby making evidence-based
decisions to improve its performance. The woreda-based plan was developed
page-244
with the support of the MoH, the RHB, and the ZHDs and is expected to be
cascaded down to the HCs. HPs are to prepare their own kebele-level plans
with the support of their respective PHCU staff. Community representatives’
involvement during kebele planning at the PHCU is recommended, although
the guidelines do not clearly state how or who will be involved in this process.
The findings show that nearly 4 in 5 (79.9%) of the assessed HPs had an
2011 EFY plan. Region-wide variations were observed. All HPs in Tigray and
Harari had a plan, but in Somali and Gambela the plan was available in only
11.8% and 38.4% of HPs, respectively. In general, plan availability was higher in
agrarian areas than pastoralist areas (83.4% vs. 45.9%; Table 6-8).
Table 6-8. Availability of health post plan (2011 EFY)

Availability of plan
Background characteristics
Yes, observed Yes, not observed No
Region
Tigray 100 0.0 0.0
Afar 77.2 9.2 13.6
Amhara 90.0 10.0 0.9
Oromia 84.8 8.9 6.3
Somali 11.8 5.2 82.9
Benishangul-Gumuz 93.2 6.8 0.0
SNNPR 75.0 5.6 19.4
Gambela 38.4 28.2 33.4
Harari 100 0.0 0.0
Livelihood
Agrarian 83.4 8.4 8.2
Pastoralist 45.9 3.6 50.5
National 79.9 7.9 12.2
Regarding the plan-preparation process, 40% of HPs work together with the
HCs and WorHOs to prepare the plan; 37.3% report that the plan is prepared
by their respective HC and WorHO and communicated to the HPs, and the
remaining 22.7% of HPs prepare the plan alone and submit it to their respective
HCs and WorHOs (Figure 6-17).
page- 245
Figure 6-17. Health post annual plan preparation process
Abbreviations: HP, health post; HC, health center; WorHO, woreda health office.
Most participants in the qualitative research in both agrarian and pastoralist

areas agree that HEWs receive their general direction and schedule from
higher officials but prepare monthly, weekly, and daily plans themselves to
facilitate their work and make decisions regarding the HEP activities without
external influence. One HEW respondent explained their planning process thus:
“
Our plan comes from the woreda, I mean the yearly plan, and
we then sit and share the work to be done in the HP. And we will
develop the plan on a weekly and monthly basis.
FGD, HEWs, Kurmuk woreda, Benishangul-Gumuz
This assessment showed that about a third (35.5%) of HPs involved the
community in planning in the following ways:
• involving them in situational analysis,
• attending presentations,
• evaluating and commenting on the previous year’s performance,
• suggesting activities to include in the plan, attending consultative
meetings, preparing workshops with cabinet members, and
• participating in final approvals.
page-246
Kebele administrators and WDA leaders are commonly engaged in the

planning process, while the following groups have had minimal involvement:
elderly and other community groups, including the Agriculture Development
Army, Women’s Affairs, school directors, the kebele cabinet, steering committee
members, the kebele manager, the kebele House of Representatives, and
women’s and youth associations (Table 6-9).
Table 6-9. Community involvement in health post planning

Region
Variables
Agrarian Pastoralist National
Community involvement in HP planning
% of HPs involved community in planning 36.4 20.3 35.5
Community groups involved in HP planning
WDA leaders 36.5 6.0 35.1
1-to-5 network leaders 25.7 9.0 24.9
Kebele administrators 49.4 36.6 48.8
Elders 21.8 22.5 21.9
Religious leaders 27.4 19.5 27.0
Other 13.3 6.1 13.0
Abbreviation: HP, health post.
Relatively higher community involvement in the planning was reported in

Gambela and the 4 largest regions (except Amhara), with less involvement in
Amhara, Benishangul-Gumuz, Afar and Somali (<20%; Figure 6-18).
page- 247
Figure 6-18. Community involvement in health post Planning

Abbreviations: Ben_Gumuz, Benishangul-Gumuz; SNPPR, Southern Nations, Nationalities, and Peoples
Region.
Some additional challenges raised by participants are:

• delays in the cascading woreda plan; no separate plan for the HP;
• political instability;
• a lack of commitment;
• a lack of human resources;
• a lack of logistics, stationery, and resources;
• a lack of incentives for WDA leaders;
• language barriers with the community;
• lengthy planning time;
• a high workload; and
• planning for irrelevant indicators.
Unrealistic targets were commonly reported in Amhara, Benishangul-Gumuz,
Oromia, and Harari. The denominator was mentioned as a major problem in
Gambela and Amhara, centralized decisions were the main challenge reported
in Afar and the SNNPR, and the knowledge gap in planning was found to be
critical in Somali (Table 6-10).
page-248
Table 6-10. Reported challenges in health post planning
Regions
Challenges
Gambela
Amhara
SNNPR
Oromia
Somali
Harari
Tigray
Ben/G
Total
Afar
Unrealistic targets
37.1 36.4 79.2 61.5 0.0 67.1 37.6 19.3 69.3 55.5
from WorHO/HC
Lack of correct
21.9 48.7 51.3 16.0 0.0 42.3 12.1 51.7 39.5 22.9
denominator
Lack of alignment
between local & 17.7 33.1 32.8 43.1 0.0 18.2 23.9 51.7 9.7 32.8
woreda priorities
Centralized
0.0 40.3 31.6 18.6 0.0 28.3 41.8 29.0 24.6 25.7
decision-making
Lack of interest
in participating in
25.7 5.6 20.7 21.0 0.0 3.5 11.3 35.5 19.3 17.3
planning among
community groups
The plan doesn’t

reflect day-to-day 2.6 29.4 14.5 39.8 1.4 0.0 8.5 6.4 9.7 22.8
activities
Lack of knowledge
and/or skills on
45.5 30.8 10.5 20.1 90.6 7.7 8.2 35.5 5.2 19.7
planning
Other 18.5 0.0 10.1 10.9 9.4 14.0 37.8 6.4 21.0 17.2
Abbreviations: Ben/G, Benishangul-Gumuz; SNPPR, Southern Nations, Nationalities, and Peoples

Region; WorHO, woreda health office; HC, health center.
The performance of the HEWs was evaluated by different groups of actors:

community members, kebele administrators, the kebele council, HCs, WorHOs,
and the HEWs themselves. The evaluation mechanism used is review meetings.
According to the revised HEP implementation guide, at the kebele level, the
page- 249
HEWs, development team and 1‐to‐5 networks should meet every 2 weeks
to discuss the implementation of the HEP, specifically focusing on health
services delivered at the facility level (these include delivery, HIV testing during
pregnancy, vaccination services, malaria prevention and control activities,
and other essential services relevant to that particular site). At the meetings,
participants are to identify challenges and gaps, such as attitudes, skills, and
inputs, then develop solutions to the issues and challenges raised and discussed.
Accordingly, community engagement in HEP performance reviews was reported

in nearly half (49.7%) of the HPs (Table 6-11). Community participation in the
reviews was found to be lower in pastoralist areas than agrarian ones (20.0%
vs. 52.7%). In addition, the level of participation by WDA leaders and kebele
administrators was higher than that of the 1-to-5 network leaders, elders, or
religious leaders. Other community groups involved in the HEP performance
review are agriculture and education leaders, kebele cabinet members,
development agents, school directors, kebele representatives, kebele civil
servants, youth chairs, members of health committees of the HP selected from
the community, and women’s leagues.
Table 6-11. Community involvement in Health Extension Program performance

review
Variables Agrarian Pastoralist National
Community involvement in HEP PR

% of HPs that involve the community in HEP
52.7 20.0 49.7
PR
Community groups involved in HEP PR
WDA leaders 85.6 62.2 84.8
1-to-5 network leaders 64.9 68.7 65.1
Kebele administrators 99.7 100 99.7
Elders 63.4 39.8 62.5
Religious leaders 69.5 41.3 68.4
Other 41.0 25.7 40.5
Abbreviations: HEP, Health Extension Program; PR, performance review.
Remarkable differences were observed between regions with regard to community

participation in the performance reviews of the HEP. The highest participation
was found in Tigray, followed by Amhara, the SNNPR, and Benishangul-Gumuz
page-250
(88%, 67%, 56%, and 52%, respectively). On the other hand, lower (41% or less)
involvement was observed in Oromia and all developing regional states except
Benishangul-Gumuz (Figure 6-19).
Figure 6-19. Community involvement in Health Extension Program performance

review
Community engagement in the performance evaluation was found to be

generally low, as described by most participants in the qualitative assessments.
“
A female community member from Raya Kobo (Amhara) affirms this:
Despite invitations from concerned bodies to participate in the

kebele council to evaluate the overall performance of HEWs, we
have never been involved in any sort of council meetings…because
of the fear of being regarded as jobless. A woman who participates
in public meetings is culturally considered to be jobless. We feared
being labeled with this kind of naming.
6.3.2 Supportive supervision

Strong and continuous supportive supervision is considered critical for the
success of the HEP. The transformation agenda has also integrated supportive
supervision of the major components of the 4 transformation agendas.
Supportive supervision is a process of guiding, helping, mentoring, training, and
encouraging staff to perform their duties and provide high-quality healthcare
services. It is conducted by both supervisors and supervisees to solve problems
page- 251
jointly at the point of service delivery. Supportive supervision emphasizes

two-way communication, with the aim of continuously improving the quality
of healthcare delivery. It provides an opportunity to increase ownership and
accountability, helping health workers at the operational level to progress and
address challenges in their work.
According to the revised HEP implementation guide, 3 types of supervision

should be done to support HPs. They are detailed below.
SUPERVISION BY THE HP FOCAL PERSON FROM THE CATCHMENT

HC
The health professionals at HCs support their assigned HPs on a weekly basis
using a checklist to guide the supervision process. The supervision activity focuses
on the implementation of all HEP packages and should address implementation
challenges, such as attitudes, skills, and program inputs. The director of the
PHCU leads, supports, and closely follows up on this activity.
SUPERVISION BY WORHO
The WorHO is expected to provide supportive supervision of its HCs and

HPs every month, using a standard checklist to guide the supervision activity
and recommend solutions for the issues and challenges it identifies. It also is
expected to follow up on the implementation of the solutions for the issues and
challenges raised and discussed.
JOINT SUPPORTIVE SUPERVISION
On a quarterly basis, the ZHD, RHB, and MoH are expected to jointly support
and supervise HPs in clusters, based on the program’s implementation status and
guided by a standardized checklist. The joint integrated supportive supervision
and follow-up is considered instrumental in delegating responsibility among
the parties involved and improving understanding about the status of program
implementation. Joint supervision should reach many zones, woredas, HCs, and
HPs in a short period of time.
page-252
HEWs were asked whether they had received supportive supervision from
the catchment HC and WorHOs within the last 6 months. The results showed
that more than three fourths (77.8%) of HPs received supervision from their
catchment HC, but less than half (46.9%) received it from their WorHOs.
According to the standard, the HCs should supervise the catchment HPs on a
weekly basis and WorHOs on a monthly basis. Only 1 in 5 (22%) HPs, however,
were being supervised on a monthly basis by HCs. Monthly supervision by the
WorHO appears negligible (Figure 6-20).
The qualitative assessment supports the above findings. Although it is

recommended that HC staff conduct supportive supervision weekly, a participant
“
from Oromia region reported that:
Not weekly but once in 2 weeks….To tell you the truth, no weekly visit
due to the burden here. Once in 2 weeks, sometimes even once in
3 weeks…but we follow up from the health center using motorcycles
and through phone calls.
KII, HEP Supervisor, Oromia
Similarly, WorHO staff members conduct supportive supervision quarterly. In

agrarian settings, most informants reported that the frequency of supportive
supervision by the WorHO was quarterly and by the RHB every 6 months. A
respondent from Assosa, Benishangul-Gumuz explained that:
“
The supportive supervision is done by quarters, at the bureau
level. We adjust it like this most of the time. On the woreda level,
samething, as at the zonal level, it is the same, it’s done one time ...
but in woredas if there isn’t any other work. They have to go in
monthly.
KII, RHB Head
page- 253
Figure 6-20. Number of supervisions to health posts in 6 months, by supervisor

type
Abbreviations: HEW, Health Extension Worker; WorHO, woreda health office.
It was reported that nearly four fifths (79.6%) of supervisors from HCs and 72.1%
from WorHOs used a checklist during supervision. Hygiene and environmental
sanitation, disease prevention and control, maternal health, child health, FP,
and health education were addressed in at least 80% of the supervision events;
on the other hand, only two thirds (65.8%) of supervisors included the CHIS in
their supervision support.
One key component of supervision is feedback. The findings reveal that a

relatively higher percentage of supervisors (70.5%) from the HC provided
feedback compared to supervisors from WorHOs (59.7%). A large majority
(84.5%) of HC supervisors provided either written or both written and oral
feedback, while a significant percentage (36.2%) of WorHO supervisors provided
only oral feedback. In pastoralist areas, a large proportion of feedback was
provided orally, while the feedback provided in agrarian settings was either
written or both oral and written (Figure 6-21). The HPs reported that they
received written feedback, and the assessment team was able to document
that the HC and WorHO provided them with written feedback 87.8% and
85.9% of the time, respectively. The content of the written feedback was also
verified. The main topics discussed included gaps in performance, causes of
these gaps, actions to be taken by HEWs, and actions to be taken by HCs.
page-254
Table 6-12. Availability and characterization of supervision of HPs
Variables Agrarian Pastoralist National
Supervision in the last 6 months

HEWs from catchment HC 78.7 68.6 77.8
Woreda Health office staff 46.4 51.8 46.9
Use of checklist during supervision
Receive feedback after supervision
Availability of written feedback
HWs from catchment HC 88.1 76.4 87.8
Abbreviations: HEW, Health Extension Worker; HC, health center.
Figure 6-21. Feedback type, by supervisor type and livelihood

Abbreviation: WorHO, woreda health office.
page- 255
The assessment also verified the knowledge of supervisors and the value of
supervision from the HEWs’ perspective. The findings are that a plurality
(>95%) of both HC and WorHO supervisors are considered either very or
somewhat knowledgeable. Similarly, a large majority (>90%) of HEWs agreed
that supportive supervision is either very or somewhat helpful.
Table 6-13. Perception of HEWs on knowledge of supervisors and usefulness of

supervision of HPs, by type of supervisor
Livelihood
Variables National
Agrarian Pastoralist
Knowledge of supervisors
Supervisor from catchment HC
Very knowledgeable 41.4 70.1 43.8
Somewhat knowledgeable 55.2 23.9 52.7
Not knowledgeable 3.3 6.0 3.6
Supervisor from WorHO
Very knowledgeable 55.0 80.8 57.6
Somewhat knowledgeable 40.6 15.7 38.0
Not knowledgeable 4.4 3.5 4.3
Usefulness of the supervision
Supervisor from catchment HC
Very helpful 70.3 76.0 70.8
Somewhat helpful 26.7 21.0 26.2
Not helpful 3.0 3.0 3.0
Supervisor from WorHO
Very helpful 75.7 87.9 77.0
Somewhat helpful 15.6 11.2 15.2
Not helpful 8.6 0.9 7.8
Abbreviations: HP, health post; HEW, Health Extension Worker; HC, health center.
page-256
Findings from the qualitative study show that the HCs’ supportive supervision
of HEWs usually focuses on technical areas, such as medical consultation,
training on new guidelines or issues, or assigning additional professionals from
the HC to accompany HEWs during campaigns, home visits, immunizations,
and outreach activities. It was also reported that the HCs provide HEWs with
resources. HEWs reported that most of their supervisors from the HCs are the
professionals from whom they expect to receive supportive supervision. The
following statements from an HEP supervisor in Afar and an HC head from
“
Oromia, respectively, confirm these reports:
When I go for a supportive supervision visit, I check the HEW’s

registers, monitor the performance of the vaccination program…
check the minutes of the discussions they had with the community,
and check the quality of the work, as well as the quality of the
data. We also discuss ways to improve the quality of care and
quality of data. The HC staff’s supportive supervision is based on
the supervisor’s specialty.
HEP Supervisor, Afar
“
Our supportive supervision may be technical assistance or
providing support to HEWs during immunization campaigns, social
mobilization, and home visits.
HC Head, Oromia
Program staff from the WorHOs, on the other hand, reported that the practice
of supportive supervision was weak and not uniform from place to place. The
supportive supervision is reported to differ in its quality and frequency due
to supervisors’ limited capacity. In addition, population dynamics (pastoralist
and agrarian) and the proximity of HPs to the WorHO or HCs influence the
frequency, depth, and quality of supportive supervision. One of the program
staff from a woreda in Oromia reflected on the lack of uniformity and consistency
of the supportive supervision:
page- 257
“
According to the standard, WorHO to HC and HC to HP are
expected to conduct supportive supervision on a weekly basis.
When we see the actual performance, it is not uniform and is not
implemented according to HEP standards.
“
HEP Staff, Oromia
Another reflection from an HEW in Harari is worrisome:

Nobody monitored our work, and no one visited us. Sometimes the
HC manager visits us and asks us how we are working…however, we
are informed that a supervisor is assigned, but we have not seen
anyone so far.
HEW, Harari
Generally, the culture of providing written feedback is weak. Participants in the

qualitative assessments confirm that this is a general problem. A statement
“
from an HEW in Semen Mecha, Amhara is a good example:
Well, supervisors from the woreda office are professionals who

provide feedback after inspecting the HPs, but we sometimes
wonder what they think of the work we do within the timeframe we
set in the plan. We think that this gap is created due to the limited
support that the woreda office is providing us; they come once a
year or once every 6 months.
The participants in the qualitative assessments described the most

common reasons for the weakness of supportive supervision by HCs
and WorHOs as: inadequate budget, supervisors’ limited knowledge
and skills, a shortage of training, and a lack of commitment.
HEW, Semen Macha, Amhara
page-258
6.3.3 Reporting and PHCU review
HPs are expected to compile their performance reports based on the HMIS
monthly report forms (service and disease) and send a hard copy of the report
to their respective catchment HCs. The findings reveal that a majority (92.1%)
of HPs send a written report to the supervisory institution. In addition, more than
50% communicate their performance review to the HC and WorHOs by phone
and/or during review meetings. The supervising institutions send feedback to
the HP through written feedback (62.7%), telephone communications (36.6%),
physical visits to the HP (51.9%), and review meetings (67.9%).
The major challenges of performance reporting to supervisory institutions and

sending feedback to HPs include:
• the time needed for report preparation (60.2%);
• the complexity of the report forms (51.0%);
• parallel reporting requirements (47.3%);
• a lack of a system for sending reports (46.9%); and
• a lack of training on how to report (41.4%).
A shortage of forms was also raised as an issue by nearly one third (30.4%) of
HPs. Some additional challenges that were raised relating to reporting include:
the forms are not provided to HPs in a timely manner; the forms are in English,
which is difficult to understand; there is a lack of orientation on the forms; the
report is bulky, with too much detail or large page numbers; the reportable data
elements are redundant; there are frequent changes to the sanitation report;
the reports include irrelevant activities that are not done at HPs; the forms
contain unnecessary sections (for example, OPD reporting); the mobile network
is occasionally down; HEWs have a large workload, particularly at HPs with
only 1 HEW; a lack of mobile battery life hinders the immediate reporting of a
reportable disease; transport is lacking to send and receive reports; there is a
shortage of stationery materials; supervisors are not knowledgeable about how
to supervise; tally sheets are incomplete; and reports include multiple copies
and different formats.
Similarly, participants were asked about challenges related to communicating

feedback to HPs. To this end, more than 3 in 5 (≥60%) HPs claimed that
feedback was not sent regularly or in a timely manner. A significant percentage
(35.1%) of respondents also reported that the feedback focuses on faults and is
unclear (30%; Figure 6-22). Additional challenges regarding feedback include:
page- 259
• The same feedback is given for everyone and every supervision;

• Supervisors lack commitment;
• Sometimes the HC or WorHO forces HEWs to produce inaccurate
reports;
• Knowledgeable supervisors do not participate in supervision;
• The feedback does not include HPs’ strengths;
• The feedback does not compare the HP’s performance with its own
plan but rather with the HC’s plan;
• The feedback is not helpful;
• The WorHO produces false reports and pressures HEWs to fill the
gaps;
• The feedback contains irrelevant information that does not directly
focus on the work of the HP;
• HEWs are told about problems when they go to the woredas for other
purposes; and
• The forms are in English and therefore difficult to understand.
Figure 6-22. Challenges in communicating feedback between supervisory

institutions and HP
Abbreviations: HC, health center; WorHO, woreda health office; HP, health post.
According to the PHCU guide, HCs should have a PHCU review meeting with
all catchment HPs on a monthly basis. The report shows that 89.5% of HPs
had attended at least 1 review meeting in the last 2 years. Less than half
(48.1%) of HPs, however, had participated in a review meeting, conducted at
page-260
either the HC or the WorHO, in the last month. The remaining 45.3% of HPs
had attended a review meeting in the previous 2 to 6 months, and 7.6% had
attended a review meeting in the previous 2 years.
The findings from the qualitative assessments show there are multiple systems
in place for monitoring and evaluation the HEP activities. Among these systems
are:
• integrated supportive supervision (ISS);
• follow-up with checklists;
• regular review meetings;
• other observations; and
• ad hoc telephone and written feedback.
“
Such mechanisms were reported by a HC head from Tigray:
Our monitoring system includes integrated supportive supervisions,

monitoring activities using checklists, inventory control, and review
meetings. Both the woreda health office and HC professionals
are conducting these monitoring and evaluation activities of the
HPs. Our monitoring and evaluation is supported by reports and
a feedback system. When we receive reports from the HPs and
HEWs, we review the reports and provide feedback to them,
including feedback if we find false reports.
HC Head, Tigray
The qualitative assessment indicates that HEWs are receiving both technical
and administrative support, mainly from HC and WorHO staff. WorHOs are
focusing mainly on technical, administrative, and financial support (budgeting).
The offices also conduct review meetings with HCs and HPs. A WorHO head
from Amhara affirms these roles and responsibilities of the office in terms of
“
supportive supervision:
To strengthen the links among the HP, HC, and woreda health
office, we perform cluster meetings with the HC and HP. Every
month, the HC has a review meeting with HPs in the catchment
areas. The HC also holds a conference with the HPs and community
representatives. The woreda health office, in turn, holds monthly
review meetings after the HC organizes its meetings and we
conduct supportive supervision visits.
page- 261
CHAPTER 7
Coverage of
HEP-Related
Services
page-262
7 Coverage of Health Extension
CHAPTER 7
Program-Related Services
Over the years, the Ethiopian Health System has made tremendous strides in
improving access to and coverage of health services. This is witnessed in terms
of reduced maternal, neonatal, and infant mortality, reduced incidence of
major infectious diseases, and improved coverage of immunization, improved
water, sanitation, and hygiene (WaSH) practices. The HEP has been the
principal vehicle for expanding access to essential health service packages
to all Ethiopians, with a focus on women and children. It has also been the
primary vehicle driving improvements in hygiene and sanitation practices.
Because the HEP is the main mechanism for reaching most vulnerable people
across the country, it is important to track the progress of the main performance
indicators of the health sector and measure the coverage of essential services
contributing to the achievement of these targets. This section presents findings
of the National Assessment of the HEP on the levels of coverage of essential
health services, the role of the HEP as source of information and service, and the
implications of providing services through the HEP for the quality of care. This
section is divided into subsections on WaSH, disease prevention and control,
family health services, maternal and neonatal health, child health, vaccinations,
and infant and child feeding.
7.1 Coverage of hygiene and sanitation services

The HEP includes 7 specific hygiene and environmental health packages that
can enable the community to lead a healthy and more prosperous lifestyle.
The packages are:
(a) Water Supply Safety Measures,
(b) Latrine Construction, Usage, and Maintenance,
(c) Solid and Liquid Waste Management,
(d) Personal Hygiene and Building and Maintaining a Healthy House,
(e) Food Hygiene,
(f) Control of Insects, Rodents, and Other Biting Species, and
(g) Healthy Home Environment.
page- 263
Coverage of HEP-Related Services
In rural settings, the HEP is the primary source of interventions in hygiene and
sanitation matters in the community. Results in this regard have a strong direct
relationship to the community’s level of implementation of the HEP.
Summary of key findings

The coverage of hygiene and sanitation facilities and practices is very low
and exhibits substantial variability across livelihood, geographic, and socio-
economic categories (Table 7-1).
page-264
Table 7-1. Key indicators in hygiene and environmental health
Percentage of households with:

characteristic
Background
separated from
drinking water
waste disposal
disposal sites
handwashing
An improved
Use of liquid
critical times
“Low” indoor
Appropriate
Appropriate
living rooms
solid waste
practice at
sanitation
pollution*
Improved
Livestock
source of
facilities
sites
National 71.4 20.0 11.6 10.7 10.8 62.0 72.6
Livelihood
Agrarian 72.6 20.7 11.1 10.5 10.6 62.3 73.1
Pastoralist 46.1 5.2 23.1 13.4 14.9 53.1 61.8
Region
Tigray 72.5 15.1 8.5 22.2 12.5 61.4 94.6
Afar 86.7 6.1 12.8 6.4 6.7 62.7 51.6
Amhara 80.2 16.5 10.7 9.3 17.9 53.6 82.0
Oromia 70.8 25.5 14.0 11.1 9.5 72.1 68.1
Somali 31.4 4.7 42.6 24.3 25.7 52.5 86.3
Benishangul-
92.6 12.4 9.1 15.5 7.2 68.0 91.1
Gumuz
SNNPR 64.5 15.4 5.0 7.4 2.5 48.5 63.1
Gambela 63.4 13.6 19.4 11.9 4.5 37.6 98.3
Harari 83.1 27.2 26.1 14.1 11.6 51.8 96.3
Wealth quintile
Lowest 54.7 8.1 12.5 8.4 7.7 45.4 59.5
Lower 64.7 17.8 8.6 6.8 8.3 42.8 59.3
Middle 76.2 19 8.2 12.1 8.3 59.1 68.7
Higher 76.0 24.5 12.9 9.2 11.6 72.4 80.5
Highest 82.0 28 15.5 16.5 17.3 81.6 88.6
page- 265
7.1.1 Water Supply Safety Measures

In rural Ethiopia, 71.4% of households have access to improved water sources,
with a wide difference between agrarian (72.6%) and pastoralist (46.1%)
households. Among improved water sources, the main sources of drinking
water were public tap or standpipe (25.5%), tube well or borehole (19.9%), and
protected spring (16.3%). Among the unimproved water sources, unprotected
springs (18.1%) served as a source for agrarian households, and unprotected
dug wells (24.7%) for pastoralist households (Table 7-2).
In this assessment, the timeframe for the availability of water from improved
sources was 2 weeks prior to the survey. Among the households surveyed,
54.4% of households in pastoralist areas had access to improved water sources
for at least 1 day in that timeframe, as did 25.2% of the households in agrarian
areas.
There was a significant difference in access to improved water sources among

regions, with the highest access in Benishangul-Gumuz (92.6%) and the lowest
in Somali (31.4%). Access to improved source increases with wealth quintile
(54.7% in the lowest and 82.0% in the highest wealth quintiles (Table 7-3).
page-266
Table 7-2. Basic and limited drinking water sources
Percentage of households with basic and with limited drinking water, by source
Households
Characteristic
Source of drinking water
Improved source 72.6 46.1 71.4
Piped into dwelling 0.2 0.0 0.2
Piped into yard/plot 4.0 0.2 3.9
Piped to neighbor 1.6 1.2 1.5
Public tap/standpipe 25.4 27.8 25.5
Tube well/borehole 20.5 6.7 19.9
Protected dug well 3.9 7.7 4.1
Protected spring 16.9 2.3 16.3
Rainwater 0.0 0.2 0.0
Unimproved source 27.4 53.9 28.6
Unprotected dug well 2.0 24.7 2.9
Unprotected spring 18.1 8.4 17.7
Surface water 7.0 16.8 7.5
Tanker truck 0.1 2.8 0.2
Cart with small tank 0.2 1.1 0.2
Other 0.1 0.1 0.1
Time to obtain drinking water (round trip)
Water on premises1 0.2 0.1 0.2
30 minutes or less 59.8 33.0 58.6
More than 30 minutes 40.0 66.9 41.2
Percentage with basic drinking water

45.8 23.4 44.8
sources2
Percentage with limited drinking water
26.8 22.7 26.6
sources3
Number of households (unweighted) 4421 2009 6430
Includes water piped to a neighbor and those reporting a round-trip collection time of zero
1
minutes
2
Defined as drinking water from an improved source, provided that either water is on the premises
or round-trip collection time is 30 minutes or less. Includes safely managed drinking water, which is
not shown separately.
3
Drinking water from an improved source, provided that round-trip collection time is more than 30
minutes
page- 267
Table 7-3. Drinking water, by livelihood, region, and household wealth quantile
Percent distribution of households, by drinking water source, percentage of households

with basic drinking water service, and percentage with limited drinking water service, by
livelihood, region and wealth quintile
Percentage
Improved Unimproved Percentage
with basic Unweighted
Background source of source of with limited
drinking number of
characteristic drinking drinking drinking
water households
water water water service2
service1
National 71.4 28.6 44.8 26.6 6430
Livelihood
Agrarian 72.6 27.4 45.8 26.8 4421
Pastoralist 46.1 53.9 23.4 22.7 2009
Region
Tigray 72.5 27.5 38.5 34.0 607
Afar 86.7 13.3 42.4 44.3 399
Amhara 80.2 19.9 51.2 28.9 1060
Oromia 70.8 29.2 49.1 21.7 1319
Somali 31.4 68.6 12.8 18.6 798
Benishangul-Gumuz 92.6 7.4 73.5 19.1 406
SNNPR 64.5 35.6 31.0 33.4 1009
Gambela 63.4 36.6 51.7 11.7 417
Harari 83.1 16.9 44.9 38.3 415
Wealth quintile
Lowest 54.7 45.3 34.8 19.9 1343
Lower 64.7 35.3 43.5 21.2 1291
Middle 76.2 23.8 47.2 29.0 1278
Higher 76.0 24.0 46.3 29.7 1274
Highest 82.0 18.0 50.4 31.6 1244
1
Defined as drinking water from an improved source, provided that either water is on the premises
or round-trip collection time is 30 minutes or less. Includes safely managed drinking water, which is
not shown separately.
2
Drinking water from an improved source, provided round-trip collection time is more than 30
minutes
Abbreviation: SNNPR, Southern Nations, Nationalities, and People Region.
At the national level, 3 out of 5 households (58.8%) had a round trip of less
than half an hour to get water. Only 0.2% were able to get water on their
page-268
premises. Only one third (33.1%) of pastoralist areas had this access, while
66.9% traveled more than 30 minutes to obtain water.
The assessment showed that the percentage of households with basic drinking
water1 service was 44.8% (45.8% in agrarian and 23.4% in pastoralist areas),
while 26.6% of the households (26.8% in agrarian and 22.7% in pastoralist
areas) had access to limited drinking water source.2
There is regional variability in the levels of access to basic and limited drinking
water services. Somali (12.8%), the SNNPR (31%), and Tigray (38.5%) had
percentages of basic drinking water service lower than the national average
(44.8%), while Benishangul-Gumuz (73.5%) had the highest. Somali (18.6%)
and Gambela (11.7%) had the lowest access to limited drinking water service.
The percentage of households with basic drinking water service increases
consistently with wealth quintile; 34.8% in the lowest quintile have basic service,
but 50.4% in the highest wealth quintile do.
This assessment showed that a majority (89.4%) of the households were

not using any method to treat their drinking water, with no significant
difference between agrarian (89.6%) and pastoralist areas (85.0%). Among
the households that treated their drinking water, 8.6% used an appropriate
treatment method (8.3% in agrarian and 4.1% in pastoralist areas). Among the
methods of treatment, bleaching (adding liquids containing chlorine) was the
most commonly used method (7.2%). Bleaching was used as a treatment
method by 13.2% of pastoralist households and 6.9% of agrarian
households that treated their drinking water (Table 7-3).
1 Drinking water from an improved source, provided that either water is on the premises or round-
trip collection time is 30 minutes or less.
2 Drinking water from an improved source, provided that round-trip collection time is more than 30
minutes.
page- 269
Table 7-4. Water treatment methods

Percentage of households using various methods to treat drinking water and percentage
using an appropriate treatment method (2019)
Households
Water treatment method
Boiling 1.6 0.6 1.5
Adding bleach/chlorine 6.9 13.2 7.2
Straining through cloth 2.2 1.3 2.2
Filtering with ceramic, sand or another filter 1.1 1.1 1.1
Solar disinfection 0 0.1 0.0
Letting it stand and settle 0.5 0.8 0.5
Other
No treatment 89.6 85.0 89.4
Percentage using an appropriate treatment
8.3 14.1 8.6
method1
Number of households (unweighted) 4 413 2 008 6 421
Note: Respondents may report multiple treatment methods, so the sum of treatments may exceed
100%.
1
Appropriate water treatment methods are boiling, bleaching, filtering, and solar disinfecting.
7.1.2 Construction, Use, and Maintenance of Sanitary Latrine
The availability of improved facilities, in the form of a ventilated improved pit

(VIP) latrine, pit latrine with a slab, or composting toilet, was 20.0% (20.7% in
agrarian and 5.2% in pastoralist communities). Only 3.9% of households owned
a VIP latrine, while 16.0% owned a pit latrine with a slab. The percentage
ofhouseholds with unimproved sanitation facilities was 51.6%. Among
sanitation facilities, unimproved pit latrines without a slab or open pits
were used bymore than half (51.4%) of households (Table 7-4).
page-270
Table 7-5. Availability of toilet or latrine facility

Percent distribution of households with a toilet or latrine facility and facility location, by
livelihood
Households
Type and location of toilet/latrine facility
Improved sanitation facility
Ventilated improved pit (VIP) latrine 4.0 1.8 3.9
Pit latrine with a slab 16.6 3.1 16.0
Composting toilet 0.1 0.4 0.1
Unimproved sanitation facility
Pit latrine without slab/open pit 53.1 14.5 51.4
Other 0.2 0.0 0.2
Open defecation (no facility/bush/field) 26.1 80.4 28.5
Location of the facility (N=3194)
In own dwelling 1.5 1.8 1.5
In own yard/plot 91.0 77.4 90.8
Elsewhere (< 500 meter) 6.8 19.8 7.0
Elsewhere (> 500 meter) 0.7 1.0 0.7
Number of households with sanitation facility (unweighted)
Percentage with basic sanitation service1 18.5 2.5 17.8
Percentage with limited sanitation service 2
3.3 2.8 3.3
Number of households (unweighted) 4421 2009 6430
Defined as the use of improved facilities not shared with other households. Includes safely
1
managed sanitation service, which is not shown separately.

2
Defined as the use of improved facilities shared by 2 or more households
The percentage of households with an improved sanitation facility varies among

regions, with the highest percentage found in Harari (27.2%) and Oromia
(25.5%) and the lowest in Somali (4.7%) and Afar (6.1%). The percentage
of unimproved sanitation facilities was 73.0% in the SNNPR and 69.9% in
Benishangul-Gumuz. The lowest percentage of households with unimproved
sanitation facilities were found in Somali (4.7%) and Gambela (4.8%). Among
households in the lowest wealth quintile, 8.1% had improved and 38.1% had
unimproved sanitation facilities. Similarly, of the households in the highest
wealth quintile, 28.0% had improved and 61.5% had unimproved sanitation
facilities (Figure 7-1, Table 7-6).
page- 271
Figure 7-1. Sanitation facilities, by wealth quintile
The largest group (90.8%) of households with access to latrines had them
in their own yards or plots; only 1.5% had them in their own dwellings. In
pastoralist communities, 19.8% of latrines were located elsewhere but within
500 meters of their dwelling.
No latrines were found in 28.5% of households (26.1% of agrarian and 80.4% of

pastoralist households), indicating a high prevalence of open defecation. The
regions with the highest proportion of households using open defecation were
Somali (91.2%), Gambela (81.7%), and Afar (80.6%). The SNNPR (11.6%),
Benishangul-Gumuz (17.8%), and Oromia (19.3%) had the lowest levels of open
defecation. The proportion of households practicing open defecation decreases
with wealth quintile: 53.8% of the households in the lowest quintile used open
defecation, but only 10.6% of households in the highest did. Overall, 17.8% of
households had access to basic3 sanitation services and only 3.3% had access
to limited4 sanitation services. This difference is significant in accessing basic
sanitation service in agrarian (18.5%) and pastoralist (2.5%) communities. Both
basic and limited sanitation services consistently increase with wealth quintile
(Table 7-6).
3 Defined as the use of improved sanitation facilities not shared with other households.
4 Defined as the use of improved sanitation facilities shared between 2 or more households.
page-272
Table 7-6. Sanitation facilities, by background characteristics
Type of sanitation Percentage Percentage

Unweighted
Background Improved Unimproved with basic with limited
Open number of
characteristic sanitation sanitation sanitation sanitation
defecation households
facility facility service service
National 20.0 51.6 28.5 17.8 3.3 6430

Livelihood
Agrarian 20.7 53.3 26.1 18.5 3.3 4421
Pastoralist 5.2 14.5 80.4 2.5 2.8 2009
Region
Tigray 15.1 11.7 73.2 13.1 2.7 607
Afar 6.1 13.3 80.6 3.4 2.9 399
Amhara 16.5 41.1 42.4 13.0 4.6 1060
Oromia 25.5 55.3 19.3 23.9 2.7 1319
Somali 4.7 4.1 91.2 1.4 3.4 798
Benishangul-
12.4 69.9 17.8 10.5 2.2 406
Gumuz
SNNPR 15.4 73.0 11.6 13.4 2.7 1009
Gambela 13.6 4.8 81.7 7.3 7.3 417
Harari 27.2 24.1 48.7 23.4 6.4 415
Wealth quintile
Lowest 8.1 38.1 53.8 7.0 1.3 1343
Lower 17.8 47.0 35.2 16.0 2.5 1291
Middle 19.0 53.4 27.7 17.2 2.6 1278
Higher 24.5 55.1 20.4 22.7 3.1 1274
Highest 28.0 61.5 10.6 23.9 7.0 1244
Handwashing facilities were available in only 6.7% of households: 3.7% were

fixed handwashing facilities, and the remaining 3% were portable. Among
households with handwashing facilities, water was available in 67.7%, and
soap was observed in 44.4% of them (Table 7-6).
According to the joint definition by the WHO and the United Nations Children’s
Fund (UNICEF), a household is considered to have a basic handwashing
service when it has a handwashing facility with soap and water on the premises
page- 273
page-274
and a limited handwashing service when it has a handwashing facility without soap or water on the
premises.54 Using this definition, the percentage of households with a basic handwashing service was 2.7%
(2.8% in agrarian areas and 0.8% in pastoralist areas) and those with a limited handwashing service
were 4% (4.1% in agrarian areas and 0.9% in pastoralist areas; Table 7-7).
Table 7-7. Availability and adequacy of handwashing facilities

Percentage of households for whom the place most often used for washing hands was observed, by whether the location
was fixed or mobile, and among the households for whom the place for handwashing was observed, the percentage with
water available, and the percentage with soap available; percentage of households with a basic handwashing facility and
percentage with a limited handwashing facility, according to background characteristics (2019)
and:
Among
in which:
number of
households
Unweighted
those having
handwashing
was observed
Percentage of
households for
facilities, those
washing hands
whom place for
handwashing facility2
Total
limited handwashing facility3
was mobile
for handwashing was observed
Percentage of the household with a
was a fixed place

Water is available
Soap is available1
Number of households for which place
Place for handwashing

Place for handwashing
Percentage of the household with a basic
Background characteristic
National 3.7 3.0 6.7 67.7 44.4 370 2.7 4.0 6 430
Livelihood
Agrarian 3.8 3.1 6.9 67.7 44.4 339 2.8 4.1 4 421
Pastoralist 1.0 0.4 1.3 69.7 38.9 31 0.5 0.9 2 009
Region
Tigray 1.9 0.2 2.1 41.5 25.0 12 1.5 1.5 607
Afar 0.0 0.4 0.4 0.0 0.0 1 0.0 0.4 399
Amhara 3.7 3.7 7.4 52.2 26.9 78 1.7 5.7 1 060
Oromia 3.5 2.6 6.1 76.5 51.6 62 3.0 3.2 1 319
Somali 0.2 0.0 0.2 45.6 64.2 5 0.1 0.1 798
Benishangul-
11.3 8.5 19.8 73.0 34.7 68 6.6 13.2 406
Gumuz
SNNPR 4.9 3.7 8.6 72.9 54.1 79 4.0 4.6 1 009
Gambela 6.2 3.2 9.4 78.8 72.6 43 5.8 3.6 417
Harari 6.9 0.0 6.0 82.0 65.4 22 3.4 2.6 415
Wealth quintile
Lowest 1.5 1.3 2.8 47.2 28.7 22 0.8 2.0 1 343
Lower 2.0 1.5 3.5 83.3 42.8 38 1.4 2.1 1 291
Middle 2.2 1.6 3.8 58.7 47.3 44 1.6 2.2 1 278
Higher 4.9 3.3 8.1 76.8 52.6 106 3.8 4.3 1 274
Highest 7.3 6.9 14.2 64.0 41.3 160 5.2 8.9 1 244
Number
235 135 370 370 149 221 6 430
(unweighted)
1
Soap includes soap or detergent in bar, liquid, powder or paste form.
2
The availability of a handwashing facility on premises with soap and water.
3
The availability of a handwashing facility on premises without soap and/or water.
page- 275

7.1.3 Personal Hygiene
The personal hygiene practices of household members were assessed by

verifying whether household members practiced hand- and face-washing,
washed their clothes, and wore shoes or sandals. In addition, the handwashing
practice at critical times was assessed.
Three fourths of households reported that they washed their hands, faces,
and clothes and wore shoes or sandals, with only a minor difference between
agrarian (75.2%) and pastoralist (68.5%) communities. The practice was
variable across regions, with the lowest rate in the SNNPR (56.8%) and the
highest in Amhara (86.6%).
Assessment of women’s handwashing practice at critical times (after using the

toilet, before food preparation, before eating, and before feeding children,
including breastfeeding) showed that they most commonly practice it before
feeding (93.3%), after feeding (80.3%), before food preparation (69.6%), and
after using the toilet (59.8%). The least commonly reported critical time that
handwashing was practiced was before breastfeeding (15.8%). The overall
percentage of women with appropriate handwashing at critical times was
11.6% (11.1% in agrarian and 23.1% in pastoralist households; Table 7-8).
page-276
Table 7-8. Handwashing education and practice
Percentage of households that received health education about handwashing, when they received the education, and their hand-
washing practice, by background characteristics
Percentage of women who wash their hands
Background
characteristic
about
Workers
Other
feeding
women who
handwashing
some work
Percentage of
before food
women with
education from
received health
preparation
at which HEWs
appropriate
delivered health
after waking
handwashing
after feeding
before breast
Health Extension
up from sleep
education for the

Percentage of
before prayers
Before feeding
The mean duration

critical times of
handwashing at
1
household members
after using toilet
after completing
Unweighted
number of
households
National 52.7 15.2 93.2 80.3 69.6 15.8 59.8 1.2 26.7 2.1 5.5 11.6 6430
Livelihood
Agrarian 53.9 15.3 93.4 80.4 69.8 15.1 60.0 1.2 25.5 2.2 5.6 11.1 4421
Pastoralist 27.4 12.0 90.0 78.2 65.9 32.5 57.1 0.5 53.2 0.6 2.5 23.1 2008
Region
Tigray 65.7 9.8 97.1 65.1 79.0 13.2 22.3 8.8 1.8 1.1 11.9 8.5 607
Afar 54.6 3.5 82.3 64.3 54.9 25.8 49.9 0.0 52.9 0.0 2.0 12.8 398
Amhara 49.2 10.8 91.4 80.8 70.1 16.6 51.3 0.4 30.7 0.6 3.8 10.7 1060
Oromia 49.6 20.2 97.0 87.4 70.3 17.1 67.9 0.0 35.7 1.1 3.3 14.0 1319
Somali 9.9 5.9 88.8 83.8 79.0 54.5 59.7 0.3 70.5 0.0 0.2 42.6 798
Benishangul-
73.0 11.2 97.2 95.5 62.2 11.8 72.2 0.8 17.9 2.0 3.9 9.1 406
Gumuz
SNNPR 65.5 11.1 87.1 67.6 64.6 8.4 62.5 3.1 2.7 7.1 11.5 5.0 1009
Gambela 42.5 3.2 84.6 55.8 58.7 30.0 62.9 0.0 1.5 0.4 0.9 19.4 417
Harari 27.6 15.4 99.4 88.6 74.0 33.4 83.4 0.4 46.3 0.0 0.8 26.1 415
Wealth quintile
Lowest 43.7 12.9 89.5 67.5 60.6 17.3 50.9 1.0 29.1 0.8 4.2 12.5 1343
Lower 49.2 13.2 93.9 82.9 70.2 13.6 54.8 1.2 25.9 1.9 4.8 8.6 1291
Middle 52.2 14.4 92.9 78.5 67.5 12.7 56.2 1.0 24.5 4.4 8.1 8.2 1277
Higher 56.5 17.2 93.7 82.1 69.9 15.8 64.6 1.3 25.0 1.9 5.5 12.9 1274
Highest 60.2 16.6 95.7 88.6 78.6 20.0 70.4 1.4 29.6 2.1 4.5 15.5 1244
page- 277

1
Duration of health education in minutes.
7.1.4 Solid and Liquid Waste Management
Only 10.7% of households reported using appropriate sites for domestic solid-
waste disposal. Open areas were the most commonly used domestic waste
disposal sites (79.7%). Of the open areas, 67.5% were undesignated and 12.2%
were designated open areas. Communal and household pits were used by only
6.3% of the households. Only 6% of households used composting as a domestic
waste disposal method. The percentage of households with appropriate waste
disposal practices did not differ significantly between agrarian (10.5%) and
pastoralist (13.4%) communities. Variability was high across the regions,
ranging from 6.4% in Afar and 7.4% in Benishangul–Gumuz to 22.2% in Tigray
and 24.3% in Somali (Table 7-9).
Similarly, only 10.8% of households used liquid waste disposal pits. Pastoralist
households used them at a slightly higher rate (14.9%) than did agrarian
communities (10.6%). The use of liquid waste disposal pits varied widely among
regions, with the lowest use in the SNNPR (2.5%), followed by Gambela (4.5%),
and the highest use in Somali (25.7%), followed by Amhara (17.9%). The use of
liquid waste disposal pits increases with wealth quintile (Table 7-9).
page-278
Table 7-9. Liquid and solid waste disposal
Domestic (solid) waste disposal site and percentage of the households with appropriate
solid waste disposal and use of liquid waste disposal site, by background characteristics
appropriate solid waste disposal

Percentage of households using
Percentage of households using

liquid waste disposal sites Percentage of households which dispose their domestic
(solid) waste at
sites
household bin
communal pit
undesignated
household pit
for garbage
designated
/ container
open area
open area
collection
burn it
bury it
Other
National 10.8 5.7 0.6 1.1 12.2 67.5 2.7 9.4 0.8 10.7
Livelihood
Agrarian 10.6 5.9 0.4 0.9 12.4 67.8 2.8 9.0 0.8 10.5
Pastoralist 14.9 2.0 4.7 5.8 9.0 59.7 0.9 17.8 0.1 13.4
Region
Tigray 12.5 16.5 2.1 1.3 19.6 58.0 2.3 0.2 0.0 22.2
Afar 6.7 0.0 5.2 0.9 19.4 73.2 0.3 1.0 0.0 6.4
Amhara 17.9 2.9 0.6 1.5 14.6 63.2 4.3 12.5 0.5 9.3
Oromia 9.5 7.8 0.1 0.0 9.1 74.5 1.9 5.2 1.3 11.1
Somali 25.7 2.7 8.8 11.1 10.9 32.3 1.6 32.4 0.1 24.3
Benishangul-
7.2 12.8 1.2 0.5 5.0 69.4 0.9 9.5 0.7 15.5
Gumuz
SNNPR 2.5 2.3 0.4 1.8 14.4 63.4 2.8 14.8 0.1 7.4
Gambela 4.5 2.0 3.8 5.5 4.2 60.4 0.4 22.0 1.7 11.9
Harari 11.6 11.7 1.6 0.5 22.7 60.0 0.4 3.2 0.0 14.1
Wealth quintile
Lowest 7.7 3.7 0.4 1.8 11.6 71.6 2.3 7.8 0.7 8.4
Lower 8.3 3.6 0.6 0.5 11.4 74.2 2.0 7.4 0.3 6.8
Middle 8.3 5.4 0.7 0.7 12.2 65.0 3.7 10.7 1.6 12.1
Higher 11.6 4.6 0.8 1.4 11.7 69.7 1.9 9.3 0.7 9.2
Highest 17.3 11.0 0.5 0.9 14.2 57.7 3.8 11.7 0.3 16.5
page- 279
7.1.5 Healthy Home Environment
A healthy home environment was assessed in terms of households’ ability to

avoid indoor air pollution and whether they had separate rooms for different
services and separated their livestock. Among the households covered by
this assessment, 58.6% had 1 sleeping room, where all the members of the
household slept. Only 17.7% of the members of pastoralist households have
more than 1 sleeping room.
The percentage of households with at least 1 livestock animal, herd animal,

other farm animal, or poultry was 84.2%. Among the households with animals,
62.0% separated their animals from their living rooms, with a significant
difference in this practice between agrarian (53.1%) and pastoralist (62.3%)
communities. The percentage increases with wealth quintile. Gambela (37.6%)
and the SNNPR (48.5%) had a relatively lower percentage, and Oromia
(72.1%) had a remarkably higher percentage of households that separated the
animals from their living rooms (Table 7-10).
page-280
Table 7-10. Sleeping rooms and separating livestock inside the house
Number of sleeping rooms and percentage of households with livestock, herd animals, other
farm animals, or poultry, and whether they were separated from living rooms, by background
characteristics
Percentage
of households Percentage
Percentage
having at least of households Unweighted
of households Unweighted
Background 1 livestock that number of
in which its number of
characteristic animal, herd separated households
members slept households
animal, other their livestock with animals
in 1 room
farm animals, or from their
poultry living room
National 58.6 6395 84.2 62.0 5154
Livelihood
Agrarian 57.6 4400 84.8 62.3 3614
Pastoralist 82.3 1995 72.8 53.1 1540
Region
Tigray 59.0 607 89.9 61.4 547
Afar 85.6 395 88.3 62.7 348
Amhara 80.3 1053 84.9 53.6 908
Oromia 46.8 1313 88.7 72.1 1165
Somali 91.4 794 53.9 52.5 450
Benishangul-
49.6 406 82.2 68.0 304
Gumuz
SNNPR 52.7 1009 75.2 48.5 871
Gambela 52.4 403 39.3 37.6 197
Harari 70.8 415 87.5 51.8 364
Wealth quintile
Lowest 87.5 1328 70.0 45.4 870
Lower 70.7 1282 84.4 42.8 1014
Middle 58.8 1271 87.1 59.1 1087
Higher 45.5 1272 89.6 72.4 1111
Highest 37.5 1242 87.2 81.6 1072
The most common source of fuel for cooking in rural households was found to
be wood (92.6%), which is an air pollutant. Only a very few households (0.6%)
were using fuel sources with a low risk of air pollution (electricity and biogas).
page- 281
In the households where food was cooked, a majority of the respondents

(63.8%) reported having a separate kitchen for cooking, and 8.6% cooked their
food outdoors. The remaining respondents (27.5%) reported that they cooked
their food inside their living room. This result differed significantly between
pastoralist (38.2%) and agrarian (27.1%) communities and across regions
(ranging from 48.8% in Afar to 1.7% in Gambela). Among households that
cooked their food within their homes, 12.5% had a separate room dedicated to
food cooking (Table 7-11, Figure 7-2).
The percentage of households with low indoor air pollution (i.e., those cooking
their food outside of their living rooms or using fuel with a lower risk of pollution)
was 72.1%. Households in Afar had a higher level of indoor pollution (49.4%)
than did other regions.
Figure 7-2: Place of cooking among households, by region
page-282
Table 7-11. Type of fuel used and indoor air pollution
Place for cooking food, type of fuel used for cooking, and percentage of households at risk of indoor air pollution, by back-
ground characteristics
Place of cooking food Type of fuel used for cooking
indoor
Crop
Separate
dung
grass
house
Wood
the home
with “low”
pollution*
Straw/
Biogas
Animal
Percentage
residual
Background
Outdoor
Charcoal
Kerosene
Inside the
Electricity
characteristic
kitchen inside
of households
In separate
No cooking
living home
National 27.5 63.8 8.6 12.5 0.5 0.1 0.2 1.2 92.6 0.7 3.0 1.7 0.0 72.6
Livelihood
Agrarian 27.1 64.6 8.3 12.5 0.5 0.1 0.2 1.1 92.4 0.7 3.2 1.8 0.0 73.1
Pastoralist 38.2 44.8 17.0 11.3 0.1 0.0 0.0 1.8 97.7 0.3 0.2 0.0 0.0 61.8
Region
Tigray 5.6 55.7 38.7 30.9 0.7 0.3 0.0 4.0 83.5 1.8 0.3 9.4 0.0 94.6
Afar 48.8 24.2 27.0 7.2 0.8 0.0 0.0 2.3 96.9 0.0 0.0 0.0 0.0 51.6
Amhara 18.2 61.7 20.1 21.1 0.8 0.2 0.0 2.3 81.9 1.6 11.0 2.4 0.0 82.0
Oromia 32.1 67.3 0.6 10.9 0.5 0.0 0.5 0.3 97.3 0.1 0.1 1.2 0.0 68.1
Somali 13.7 56.0 30.3 61.8 0.0 0.0 0.0 3.6 95.5 0.9 0.0 0.0 0.0 86.3
Benishangul- 8.9 84.6 6.5 7.1 0.0 0.0 0.4 0.8 98.1 0.0 0.0 0.8 0.0 91.1
Gumuz
SNNPR 37.0 61.6 1.4 7.7 0.1 0.0 0.0 0.6 98.5 0.7 0.0 0.0 0.1 63.1
Gambela 1.7 27.7 70.6 0.0 0.0 0.0 0.0 2.3 92.5 5.3 0.0 0.0 0.0 98.3
Harari 3.7 85.0 11.3 0.0 1.9 0.0 0.3 1.6 86.8 9.3 0.0 0.0 0.0 96.3
Wealth quintile
Lowest 40.9 43.4 15.7 12.0 0.3 0.3 0.7 0.0 93.1 1.4 2.1 2.1 0.1 59.5
Lower 40.7 44.1 15.2 8.7 0.5 0.0 0.2 0.0 91.9 0.8 3.8 2.7 0.0 59.3
Middle 31.3 60.6 8.1 8.0 0.1 0.1 0.0 0.5 93.4 0.5 3.9 1.4 0.0 68.7
Higher 19.5 76.2 4.4 10.5 0.4 0.0 0.0 1.0 94.3 0.9 2.3 1.0 0.0 80.5
Highest 11.9 84.8 3.3 40.6 1.2 0.0 0.4 3.8 90.0 0.2 2.9 1.6 0.0 88.6
*Households were assumed to have a “low” risk of indoor air pollution if electricity or biogas were used as cooking fuels and the place
of cooking was not within the living home.
page- 283

7.1.6 Food Hygiene and Control of Insects, Rodents, and Other

Biting Species
This assessment examined the abnormal presence of insects, rodents, and
biting species. In addition, it examined households’ practice of covering food
items for protection from flies.
Among the households in this survey, 75.4% covered or protected their food
items from flies at the time of data collection. The food items were covered or
protected in more than three fourths (76.5%) and half (51.6%) of agrarian and
pastoralist households, respectively. This result ranged from 34.2% to 94.3% in
Somali and Tigray, respectively (Table 7-12).
Most households (63.7%) reported the abnormal presence of flying species

(mosquitoes, fleas, and cockroaches); other species (insects, stinging species,
and rodents) were reported as rare (0.7%). Nearly two thirds of households
(63.3%) in agrarian areas and 79.85% of households in pastoralist areas had
at least 1 type of insect, stinging species, or rodent. The particular species of
which there was an abnormal presence varied little by wealth (61.8% in the
lowest and 60.8% in the highest wealth quintiles (Table 7-12)).
page-284
Table 7-12. Protection of food from flies and abnormal presence of insects and
rodents
Percentage of households in which food items are protected from flies and percentage of
households with an abnormal presence of insects, rodents, and other biting species, by back-
ground characteristics
Percentage of households which had Percentage
Percentage
abnormal presence of rodents or insects of households
of
with abnormal
households
Background presence of at
where food
characteristic Flying Stinging least 1 type of
is covered Insects Rodents
species species insects, rodents
or protected
or other biting
from flies
species
National 75.4 0.5 63.7 0.0 0.2 64.0
Livelihood
Agrarian 76.5 0.5 63.1 0.0 0.1 63.3
Pastoralist 51.6 0.6 77.9 0.0 1.9 79.8
Region
Tigray 94.3 0.0 31.4 0.0 0.4 31.7
Afar 76.2 0.0 37.1 0.0 0.0 37.1
Amhara 81.8 0.2 40.3 0.0 0.0 40.3
Oromia 73.5 0.7 77.9 0.0 0.0 78.2
Somali 34.2 0.0 90.1 0.0 4.1 94.2
Benishangul-
89.2 0.0 79.0 0.0 0.2 79.2
Gumuz
SNNPR 70.3 0.8 68.6 0.1 0.5 68.8
Gambela 75.8 0.4 77.6 1.3 0.0 77.6
Harari 77.1 0.3 89.3 0.0 0.0 89.3
Wealth quintile
Lowest 62.1 0.9 66.4 0.0 0.4 66.8
Lower 71.3 1.4 65.3 0.0 0.3 66.2
Middle 76.8 0.1 64.6 0.0 0.2 64.6
Higher 79.0 0.3 62.3 0.0 0.1 62.3
Highest 84.9 0.0 60.7 0.1 0.1 60.8
page- 285
7.1.7 Health education on Hygiene and Environmental Health
The HEP is the main mechanism for improving hygiene and sanitation within
households in rural Ethiopia. To determine the frequency of the HEWs’
engagement with households, we assessed their role in providing health
education on handwashing.
In the current study, 52.7% of women, 45.2% of men, and 23.6% of youth
girls had ever received health education from HEWs on handwashing, with
significant variability between pastoralist and agrarian settings. The most
recent point at which residents had received health education from HEWs was
on average 15.2 months prior to the survey for women, 15.8 months for men,
and 15.8 months for youth girls (Figure 7-3).
Figure 7-3. Households receiving health education on handwashing by HEWs,

by region
Abbreviation: SNNPR, Southern Nations, Nationalities, and Peoples Region; HEW, Health Extension
Worker.
page-286
7.2 Coverage of disease prevention and control

interventions
The Disease Prevention and Control umbrella package includes 3 specific
sub-packages: malaria prevention and control, HIV/AIDS and tuberculosis
prevention and control, and first aid. The HEP is expected to create community
awareness, improve knowledge, and increase the use of services related to
these programs. HEWs are also involved in the provision of services, including
malaria diagnosis and treatment, the distribution of ITNs, HIV testing and
counseling, and community treatment of TB. This section presents findings of
the 2019 National Assessment on Coverage and Outcomes of the HEP related
to disease prevention and control interventions regarding these 3 diseases—
malaria, HIV, and TB, with a focus on the coverage of interventions, disease
occurrence, and the role of the HEP in the prevention and control of these
diseases.
Summary of findings
• Half of the households in malarious areas own at least 1 ITN.
Households with at least 1 ITN mostly have adequate numbers of ITNs;
84% of the households with at least 1 ITN meet the criteria of having
at least 1 ITN for every 2 persons in the household.
• Only 27% of the household members in malarious areas and 47% of
those who own at least 1 ITN sleep under an ITN. Only 43% of the
ITNs owned are used by household members. Only half of children and
one third of pregnant women slept under an ITN the night before data
collection.
• A majority of women, men, and youth girls (more than 84%) have
awareness about HIV/AIDs, with significant variation between
pastoralist (as low as 61%) and agrarian communities. Only 13% of
women and 21% of men and youth girls, however, have comprehensive
knowledge about HIV/AIDS.
• A majority of household members are aware of TB. Women have
a lower level of awareness than men or youth girls, but only 9% of
women, 15% of men, and 19% of youth girls know TB prevention
methods.
page- 287
• In relative terms, HEWs serve as a good source of information about

communicable diseases, but they provided services to a low percentage
of households. HPs are less likely (only 10%) to serve as the first point
of contact for household members with symptoms of TB, while HCs are
the most commonly used facilities (60%).
• HPs are very often bypassed. Only 11% and 9% of patients who
had been sick in the last 1 month get referral from HPs or HEWs,
respectively. The most common reason for bypassing the HP was the
unavailability of the services.
• HEP is currently playing a very limited role in the prevention and
control of NCDs and in addressing the mental health needs of
community members.
7.2.1 Malaria Prevention and Control
The assessment showed that 51% of households had at least 1 ITN, with a mean
ownership of 1.1 ITN per household. Having at least 1 ITN was relatively higher
in pastoralist areas (63.1%), the Somali (96.3%), Benishangul-Gumuz (90.8%),
and Harari (89.2%) regions, and the highest wealth quintile (64.8%). Coverage
was lowest in the SNNPR (21.6%) and Afar (47.4%). The highest average
number of ITNs owned per household was in Somali (mean=2.41), followed by
Benishangul-Gumuz (mean=2.02). In pastoralist areas, the average number
of ITNs was 1.5 per household, compared to 1.0 in agrarian areas (Table 7-13,
Table 7-14).
page-288
Table 7-13. Household possession of ITNs
Percentage of households with at least 1 insecticide-treated net (ITN); average number of

ITNs per household; and percentage of households with at least 1 ITN per 2 persons who
stayed in the household last night, by background characteristics
Percentage of
Number of
Average households with
households with
Households number at least 1 ITN for
Background Number of at least 1 person
with at of ITNs every 2 persons
characteristic households who stayed in
least 1 ITN per who stayed in the
the household
household household last
last night
night
National 51.0 1.1 4522 84.2 4516
Livelihood
Agrarian 50.2 1.0 2988 84.6 2985
Pastoralist 63.1 1.5 1534 78.7 1531
Region
Tigray 67.2 1.5 403 78.8 403
Afar 47.4 0.8 266 73.7 265
Amhara 58.6 1.1 616 89.1 616
Oromia 60.8 1.3 607 82.3 605
Somali 96.3 2.4 693 74.9 691
Benishangul-
406 87.0 405
Gumuz 90.8 2.0
SNNPR 21.6 0.5 745 82.8 745
Gambela 68.4 1.6 417 84.8 417
Harari 89.2 1.7 369 85.2 369
Wealth quintile
Lowest 48.9 0.9 952 86.9 951
Lower 42.0 0.8 875 87.5 874
Middle 45.5 1.0 900 83.6 899
Higher 51.5 1.0 893 83.6 893
Highest 64.8 1.5 902 80.6 899
The adequacy of ITNs (at least 1 ITN for every 2 household members) was
assessed among the households that reported ownership of at least 1 ITN.
There were enough ITNs at the household level in malarious area of the
page- 289
country (Table 7-14). If all ITNs possessed by a household were used, 100% of
the people could have access to ITNs, provided that the number of household
members was less than 3. This access declines as the number of household
members increases, and reaches 69.1% when the number of household members
is more than 7. Had all the ITNs been used by the population in malarious
areas, on average, 84.3% of the household members would potentially have
slept under ITNs.
Table 7-14. Access to an insecticide-treated net (ITN)
Percent distribution of the de facto household population, by number of ITNs the household
owns, and percentage with access to an ITN, according to number of people who stayed in
the household the night before the survey
Number of people who stayed in the household the night before the survey
Number of
ITNs 7 8+ Total
1 2 3 4 5 6
1 4.9 7.4 4.7 5.6 7.4 4.3 2.4 3.1 5.0
2 16.5 12.9 18.2 19.6 17.6 15.6 15.1 19.7 17.2
3 2.3 3.7 5.8 7.1 9.1 8.8 7.2 8.1 7.2
4 2.7 2.7 2.3 1.7 1.8 1.8 2.1 1.5 2.0
5 14.7 9.4 10.9 10.6 10.8 6.0 6.1 3.7 8.6
6 9.8 12.8 9.3 6.9 7.5 3.3 5.0 5.7 7.1
7 1.1 1.2 1.7 2.0 0.9 0.9 1.7 1.0 1.4
8+ 48.1 49.8 47.1 46.5 44.8 59.4 60.3 57.1 51.6
Number of
139 434 656 742 751 635 515 644 4516
households
Percentage
with access to 100.0 100.0 95.4 94.4 75.0 80.2 75.2 69.1 84.3
an ITN1
1
Percentage of the de facto household population who could sleep under an ITN if each ITN in the
household were used by up to 2 people.
As possession of an adequate number of ITNs in the households was not

necessarily evidence of their use, actual use was also assessed. The proportion
of people who slept under an ITN the previous night among the population
in malarious districts was 27.1%, and the proportion of people who slept
under ITNs among households who had at least 1 ITN was 46.7%. The use
of ITNs was relatively higher in pastoralist areas than agrarian ones and
among higher socio-economic groups, and was highest in Gambela, followed
by Somali, while the lowest use was in the SNNPR. Among the ITNs owned
page-290
by households, fewer than half (43.3%) were used the night prior to the survey.
Use was relatively higher in pastoralist areas (34.1%), among people in the
highest wealth quartile (33.2%), and in Gambela (62.2%) and Benishangul-
Gumuz (52.4%; Table 7-15).
Table 7-15. Use of ITNs and existing ITNs used in the household
Percentage of the de facto household population who slept the night before the survey
under an insecticide-treated net (ITN); and among the de facto household population
in households with at least 1 ITN, percentage who slept under an ITN the night before
the survey; percentage of ITNs that were used by anyone the night before the survey, by
background characteristics
Among the household Among population in
population in malarious households with at least
districts: 1 ITN: Percentage
Background Percentage of existing Number
characteristic Percentage ITNs used of ITNs
who slept
who slept Number Number last night
under an
under ITNs of persons of persons
ITN last
last night
night
National 27.1 4113 46.7 2841 43.3 6603
Livelihood
Agrarian 26.6 2676 46.3 1903 26.6 4342
Pastoralist 34.1 1437 52.6 938 34.1 2261
Region
Tigray 9.5 379 13.3 294 9.5 671
Afar 39.3 253 74.0 150 39.3 257
Amhara 25.4 543 41.2 361 25.4 761
Oromia 36 593 58.8 266 36.0 552
Somali 49.5 663 50.9 628 49.5 1639
Benishangul-
52.4 396 57.5 365 52.4 882
Gumuz
SNNPR 14.3 564 44.0 200 14.3 507
Gambela 62.2 395 87.2 287 62.2 693
Harari 34.5 327 38.3 290 34.5 369
Wealth quintile
Lowest 23.3 875 42.9 643 23.3 1457
Lower 21.7 773 40.8 552 21.7 1229
Middle 22.9 804 42.0 513 22.9 1180
Higher 31.0 820 54.5 543 31.0 1274
Highest 33.2 841 49.6 590 33.2 1463
page- 291
Table 7-16. Use of ITNs by children and pregnant women

Percentage of children under age 5 and pregnant women who, the night before the survey,
slept under an insecticide-treated net (ITN); and, among children under age 5 and pregnant
women in households with at least 1 ITN, percentage who slept under an ITN the night
before the survey, by background characteristics
Children Among pregnant Among pregnant
under age 5 Children
in
under age 5
households women aged women aged 15-49
in all with at least 1 ITN 15-49 in all in households with
households households at least 1 ITN1
under an ITN
under an ITN
under an ITN
Percentage
Percentage
Percentage
Percentage
Number of
Number of
of children
of children
under ITN
last night
last night
last night
last night
who slept
who slept
who slept
who slept
pregnant
pregnant
Number
Number
women
women
National 31.8 5 673 52.5 4 399 19.2 2872 35.8 2 430
Livelihood
Agrarian 31.3 2 779 52.5 2 021 16.8 778 35.6 477
Pastoralist 35.7 2 894 52.3 2 378 27.7 2094 31.1 1 953
Region
Tigray 13.5 260 18.8 198 22 23 48.3 11
Afar 47.8 212 80.9 130 35.4 18 85.2 8
Amhara 37.3 271 50.8 215 26.4 61 45 37
Oromia 36.7 612 62.1 236 21.8 117 37.5 38
Somali 47.8 2 195 48.9 2 085 29.6 2 027 30.2 1 929
Benishangul- 62.7 311 67.4 295 59.7 32 63.8 30
Gumuz
SNNPR 15.8 563 46.5 190 4.8 189 26.1 24
Gambela 67.3 658 87.2 517 40.5 64 75.3 45
Harari 40.2 591 44.5 526 17.3 341 19.7 308
Education
No N/A N/A N/A N/A 18.2 2 544 34.6 2 175
education
Primary N/A N/A N/A N/A 15.3 163 32 129
Secondary N/A N/A N/A N/A 29.9 114 51.9 88
More than N/A N/A N/A N/A 8.8 51 11.5 38
secondary
Wealth quintile
Lowest 26.7 1 731 42.6 1 486 15.2 1 246 25.5 1 111
Lower 26.7 1 172 46.6 944 35.4 657 54.7 571
Middle 21.4 949 42.3 662 13.7 399 27.3 321
Higher 33.2 957 57.2 655 17.3 303 40 207
Highest 43.6 864 62.6 652 19.5 267 38.3 220
N/A: Not Applicable
page-292
In the household use of ITNs, priority should be given to pregnant women and
children under 5 years of age. Among under-5 children residing in malarious
areas, 31.6% slept under an ITN the previous night. Similarly, 52.5% of children
under 5 whose families owned at least 1 ITN had slept under one the previous
night. Children from families in the highest wealth quintile were more likely to
sleep under an ITN relative to other children. Among all children in malarious
districts, the likelihood of sleeping under an ITN the previous night was relatively
higher in the pastoralist areas (35.7%) and Gambela (67.3%), followed by
Benishangul-Gumuz (62.7%; Table 7-16).
Similarly, 19.2% of all pregnant women and 35.8% of pregnant women whose
household possessed at least 1 ITN had spent the previous night under an ITN.
The patterns of ITN use across livelihood, region, maternal education, and
wealth quintile vary among all pregnant women and pregnant women from
households possessing at least 1 ITN (Table 7-16).
7.2.2 HIV/AIDS Prevention and Control
In this assessment, 83.8% of women, 94.6% of men, and 94.3% of youth girls
reported ever having heard about HIV/AIDS. Respondents from Somali had
the lowest level of awareness about HIV/AIDS, with 46.2%, 62.1%, and 62.1%
of women, men, and youth girls, respectively, reporting everhearing about
HIV/AIDS. The highest level was observed among respondentsfrom Tigray
(Table 7-17).
Table 7-17. Awareness of HIV: Women, men, and youth girls
Percentage of women, men, and youth girls who had ever heard of HIV or AIDS, by
Women Men Youth girls
Background Have
Have heard Number Have heard Number
characteristic heard of Number
of HIV or of of HIV or of youth
HIV or of men
AIDS women AIDS girls
AIDS
National 83.8 5183 94.6 4030 94.3 803
Age
15-19 67.1 194 94.7 17 93.1 633
20-24 85.2 621 93.7 196 98.3 170
25-29 86.9 992 97.2 521 Na N/A
30-34 85.8 778 98.7 618 Na N/A
35-39 88.1 790 93.0 590 Na N/A
page- 293
40-44 83.8 415 96.6 568 Na N/A

45-49 85.5 293 96.9 384 Na N/A
50+ 78.1 1100 91.2 1136 Na N/A
Marital status
Currently married 84.7 4225 94.7 3970 94.1 67
Divorced 87.4 257 98.8 18 95.5 34
Widowed 76.9 514 77.9 20 100.0 3
Separated 68.4 149 100.0 7 100.0 10
Never married 87.8 38 96.6 15 94.2 689
Livelihood
Agrarian 84.8 3886 95.2 3022 95.0 625
Pastoralist 60.9 1297 79.8 1008 72.5 178
Region
Tigray 97.2 586 99.3 405 95.4 111
Afar 82.2 326 95.4 251 87.4 52
Amhara 91.5 967 95.6 573 94.2 159
Oromia 82.6 1063 96.2 1072 95.5 149
Somali 46.2 341 62.1 211 62.1 51
Benishangul-
94.3 370 99.2 337 100.0 50
Gumuz
SNNPR 76.3 812 90.8 697 94.6 143
Gambela 87.3 367 89.5 143 94.1 34
Harari 85.5 351 95.1 341 95.0 54
Education
No formal
80.8 3700 91.2 2144 72.7 106
education
Attended grade
87.1 645 96.8 656 92.7 123
1-4
Attended grade
94.1 566 97.3 760 97.1 357
5-8
Attended grade
96.6 272 99.9 470 98.7 217
9+
Wealth quintile
Lowest 72.2 915 86.5 570 90.6 116
Lower 78.4 995 92.3 744 90.7 133
Middle 84.5 1040 95.8 819 97.7 173
Higher 88.8 1111 96.9 965 92.9 174
Highest 92.3 1122 97.7 932 96.5 207
N/A=Not applicable; the numbers are unweighted.

page-294
In this study, comprehensive knowledge about HIV/AIDS was defined as

knowing that the consistent use of condoms during sexual intercourse and
having just 1 uninfected faithful partner can reduce the chance of acquiring
HIV, knowing that a healthy-looking person can have HIV, and rejecting the
2 most common local misconceptions about the transmission or prevention
of HIV (that HIV can be transmitted by mosquito bites and that a person
can become infected by sharing food with a person who has HIV). By this
definition, this assessment found that only 13.0% of women had comprehensive
knowledge about HIV/AIDS (13.1% in agrarian and 7.1% in pastoralist areas).
The lack of knowledge about the transmission and prevention methods of HIV
and misconceptions contributed to the observed low level of comprehensive
knowledge. Women’s comprehensive knowledge differed notably among
regions, from 0.0% and 3.0% in Somali and Benishangul-Gumuz, respectively,
to 15.4% and 21.7% in Oromia and Gambela, respectively. Men and youth girls
had an almost equal level of comprehensive knowledge about HIV/AIDS, at
20.9% and 20.4%, respectively. Both men (10.9%) and youth girls aged 15 to
19 (17.9%) had the lowest levels of comprehensive knowledge compared to
older men and youth girls. Overall, women have a low level of comprehensive
knowledge of HIV or AIDS across all regions (Table 7-18, Table 7-19).
page- 295
Table 7-18. Knowledge of HIV prevention methods: Women, men, and youth girls
page-296
Percentage of women, men, and youth girls who, in response to prompted questions, stated that people can reduce their risk of getting HIV
by using condoms every time they have sexual intercourse and by having 1 sex partner who is not infected and has no other partners, by
girls
to one
to one
to one
women
partner2
partner2
partner2
Background
partner1,2
partner1,2
partner1,2
characteristic
uninfected
uninfected
uninfected
intercourse
intercourse
intercourse
Number of

and limiting
and limiting
and limiting
to 1 uninfected
to 1 uninfected
to 1 uninfected
Using condoms
Using condoms
Using condoms
Limiting sexual
Limiting sexual
Limiting sexual
Using condoms1
Using condoms1
Using condoms1
Number of men
Number of youth
sexual intercourse
sexual intercourse
sexual intercourse
National 51.2 70.5 43.4 69.4 80.7 60.4 72.2 78.2 60.3
Age
15-19 57.0 70.8 46.4 194 95.4 91.1 89.1 199 70.9 78.3 59.0 633
20-24 60.2 78.2 53.5 621 73.1 86.1 67.5 564 76.9 78.0 64.7 170
25-29 56.9 74.9 48.6 992 70.2 83.4 63.9 860 N/A N/A N/A N/A
30-34 54.8 72.3 48.5 778 76.0 85.2 67.4 632 N/A N/A N/A N/A
35-39 56.3 71.3 47.7 790 72.6 79.9 62.4 643 N/A N/A N/A N/A
40-44 56.8 74.5 49.5 415 70.9 76.9 58.4 295 N/A N/A N/A N/A
45-49 45.1 62.7 35.8 293 73.0 81.7 64.1 190 N/A N/A N/A N/A
50+ 35.3 61.9 27.4 1 100 61.9 78.9 53.5 597 N/A N/A N/A N/A
Livelihood
Agrarian 51.8 71.1 43.9 3 886 70.1 81.4 61.0 3 022 73.1 79.0 61.1 625
Pastoralist 33.2 50.1 27.8 1 297 46.4 59.5 39.4 1 008 36.5 47.8 29.1 178
Region
Tigray 68.9 88.0 63.6 586 87.2 95.4 84.1 405 78.6 91.7 72.1 111
Afar 41.5 52.5 35.0 326 48.4 59.6 38.7 251 62.1 66.7 41.3 52
Amhara 53.7 72.6 45.1 967 74.8 79.9 63.9 573 75.6 77.4 61.2 159
Oromia 50.8 72.9 43.9 1 063 68.9 84.1 61.0 1 072 68.9 80.0 58.3 149
Somali 10.5 27.6 9.7 341 17.5 28.0 16.9 211 5.6 17.7 5.1 51
Benishangul- 42.5 69.3 32.0 370 66.1 79.7 57.6 337 80.8 81.4 68.0 50
Gumuz
SNNPR 45.2 58.3 35.4 812 63.6 72.0 51.9 697 75.4 75.3 61.8 143
Gambela 70.0 81.1 62.5 367 88.3 83.4 78.3 143 87.0 86.2 73.2 34
Harari 54.4 78.0 51.0 351 72.1 88.2 67.0 341 76.3 83.8 75.1 54
Education
No formal 45.4 66.8 38.3 3 700 61.5 77.3 52.2 2 761 62.3 60.2 46.5 106
education
Attended 55.9 73.4 47.5 645 71.6 78.2 60.5 560 64.2 74.2 53.1 123
grade 1-4
Attended 67.8 79.9 56.7 566 77.8 83.9 69.3 465 70.3 78.1 56.3 357
grade 5-8
Attended 77.7 89.8 71.1 272 79.7 91.1 73.5 194 82.2 85.5 74.0 217
grade 9+
Marital status
Currently 52.8 71.3 45.0 4 225 69.6 80.7 60.5 3 949 56.3 83.6 49.3 67
married
Divorced 52.2 75.1 46.1 257 51.3 82.1 51.3 7 82.3 85.2 80.7 34
Widowed 36.3 60.5 28.4 514 54.2 52.8 31.6 18 0.0 0.1 0.0 3
Separated 39.2 60.8 29.0 149 19.9 87.5 7.5 2 78.5 100.0 78.5 10
Never 39.7 85.7 34.6 38 96.2 99.0 96.2 4 73.3 78.2 60.6 689
married
Wealth quintile
Lowest 41.5 61.1 33.9 915 63.0 71.6 50.3 308 68.5 47.2 34.7 116
Lower 48.3 69.3 40.3 995 68.7 79.1 61.9 451 65.8 83.9 56.7 133
Middle 51.7 67.8 44.0 1 040 70.0 80.6 60.2 471 77.0 82.5 68.7 173
Higher 51.2 72.4 42.9 1 111 67.4 82.7 59.2 595 73.3 78.2 60.3 174
Highest 59.5 78.0 52.2 1 122 75.3 84.7 66.4 592 72.3 85.3 66.4 207
N/A=Not applicable
1
Using condoms every time they have sexual intercourse
2
Partner who has no other partners
page- 297

Table 7-19. Comprehensive knowledge of HIV: Women, men, and youth girls
Percentage of women, men and youth girls who say that a healthy-looking person can have HIV and who,
in response to prompted questions, correctly reject local misconceptions about transmission or prevention
of HIV, and percentage with a comprehensive knowledge about HIV, according to age, livelihood and
region
Percentage of respondents who say that:
say that a healthy-

looking person can
have HIV and who
knowledge about
reject the 2 most
a comprehensive
Percentage with
Percentage who
misconceptions1
common local
characteristic
Total number
Background
cannot become
HIV cannot be
HIV cannot be
transmitted by
transmitted by
looking person
mosquito bites
can have HIV
with a person
who has HIV
supernatural
sharing food
infected by
A healthy-
A person
HIV2
means
WOMEN
National 52.8 41.4 74.0 75.8 19.5 13.0
Age
15-19 47.9 30.5 68.3 60.9 10.3 3.1 194
20-24 54.0 54.9 83.4 89.2 26.6 18.4 621
25-29 56.6 48.0 77.4 80.7 24.7 17.0 992
30-34 55.2 48.8 78.0 85.1 25.0 19.0 778
35-39 49.0 43.8 78.7 80.0 18.2 13.2 790
40-44 58.9 34.2 73.5 74.3 18.2 12.4 415
45-49 61.4 31.9 69.9 69.1 18.8 9.1 293
50+ 46.4 28.9 62.6 59.8 10.7 5.3 1 100
Livelihood
Agrarian 53.4 41.6 75.0 76.7 19.7 13.2 3 886
Pastoralist 33.7 33.9 44.4 47.7 13.3 7.1 1 297
Region
Tigray 69.1 18.9 64.0 69.0 11.3 10.2 586
Afar 48.5 57.2 53.5 73.7 25.2 11.2 326
Amhara 66.1 34.5 67.9 81.7 22.3 13.7 967
Oromia 50.5 47.6 81.6 73.2 20.2 15.4 1 063
Somali 15.9 10.1 10.9 11.4 1.8 0.0 341
Benishangul- 28.6 22.8 42.2 74.1 4.7 3.0 370
Gumuz
SNNPR 34.4 46.5 73.8 74.1 17.0 7.7 812
Gambela 66.0 41.0 62.0 80.1 24.6 21.7 367
Harari 52.9 40.8 40.8 65.2 18.1 11.9 351
MEN
National 56.3 51.5 81.4 84.0 27.0 20.9
Age
15-19 91.8 20.7 88.8 35.6 12.1 10.9 17
20-24 59.0 49.5 87.6 81.9 28.2 23.6 196
25-29 56.9 55.0 81.2 84.6 26.8 21.7 512
page-298
30-34 56.4 56.3 84.6 85.0 29.5 24.0 618

35-39 57.5 57.9 80.9 86.1 31.2 24.5 590
40-44 55.8 53.8 85.5 86.3 27.5 21.7 568
45-49 56.3 47.5 85.1 91.1 30.2 23.1 384
50+ 55.1 45.7 76.1 79.3 22.7 16.1 1 136
Livelihood
Agrarian 56.7 51.7 82.2 84.8 27.2 21.1 3 022
Pastoralist 41.8 45.6 55.1 56.2 20.3 14.6 1 008
Region
Tigray 76.8 33.7 78.0 77.0 23.1 21.2 405
Afar 45.0 53.7 48.0 68.2 21.7 14.0 251
Amhara 70.7 46.7 77.0 90.6 31.9 22.9 573
Oromia 57.6 55.2 87.4 82.2 28.8 24.4 1 072
Somali 15.3 13.8 15.5 14.3 1.0 0.9 211
Benishangul- 31.3 39.4 43.4 85.9 10.2 7.9 337
Gumuz
SNNPR 36.6 52.7 75.5 86.3 20.1 11.3 697
Gambela 74.2 50.9 58.6 80.9 39.1 36.7 143
Harari 64.2 53.1 52.1 76.5 30.4 23.9 341
YOUTH GIRLS
National 58.1 53.4 89.1 86.9 30.1 20.5
Age
15-19 56.0 52.1 88.9 86.2 27.7 17.8 633
20-24 65.3 57.9 89.7 89.1 37.9 29.4 170
Livelihood
Agrarian 58.7 53.8 90.3 87.9 30.4 20.7 625
Pastoralist 33.8 38.1 40.2 45.8 18.6 11.0 178
Region
Tigray 76.8 37.8 81.0 80.8 24.2 18.6 111
Afar 49.5 61.4 38.0 70.7 27.5 18.4 52
Amhara 64.9 57.0 88.9 89.5 29.4 22.6 159
Oromia 59.7 50.3 96.0 90.9 33.0 22.5 149
Somali 7.9 15.8 11.2 11.7 6.3 0.0 51
Benishangul- 26.2 51.3 67.2 95.6 18.3 18.3 50
Gumuz
SNNPR 44.4 61.1 85.6 83.0 29.1 16.0 143
Gambela 76.1 48.3 64.1 64.1 30.5 23.6 34
Harari 58.6 54.7 57.4 77.0 26.3 26.3 54
1
The 2 most common local misconceptions are that HIV can be transmitted by mosquito bites and that a
person can become infected by sharing food with a person who has HIV.
2
Comprehensive knowledge means knowing that the consistent use of condoms during sexual intercourse
and having just 1 uninfected faithful partner can reduce the chance of getting HIV, knowing that a
healthy-looking person can have HIV, and rejecting the 2 most common local misconceptions about the
transmission or prevention of HIV.
Abbreviation: SNNPR, Southern Nations, Nationalities, and Peoples Region. page- 299
The respondents’ knowledge of transmission of HIV from mother to child was

also assessed. Only 41.6% of women, 46.5% of men, and 57.0% of youth girls
knew all 3 ways that HIV can be transmitted from mother to child (during
pregnancy, during delivery, and through breastfeeding). Women’s knowledge
about the mother-to-child transmission (MTCT) of HIV varied across livelihood,
geographic, and age categories. The percentage of men and youth girls who
knew the modes of MTCT of HIV also varied by region. The percentage of
respondents who knew that the risk of MTCT of HIV could be reduced by the
mother’s taking special drugs was 50.9% among women, 53.6% among men,
and 62.7% among youth girls (Table 7-20).
page-300
Table 7-20. Knowledge of prevention of mother-to-child transmission of HIV

Percentage of women, men, and youth girls who know that HIV can be transmitted from mother to child during
pregnancy, during delivery, by breastfeeding, and by all 3 means, and percentage who know that the risk of mother
to child transmission (MTCT) of HIV can be reduced by the mother’s taking special drugs, by age, livelihood and
region
Percentage who know that HIV can be Percentage who

transmitted from mother to child: know that the risk
Background Total
By of MTCT can be
characteristic During During By all 3 number
breast- reduced by mother
pregnancy delivery means taking special drugs
feeding
WOMEN
National 59.1 69.0 62.5 41.6 50.9
Age
15-19 60.0 72.8 62.1 28.5 47.5 194
20-24 65.7 75.1 71.2 48.1 56.3 621
25-29 61.8 73.4 64.5 44.5 50.7 992
30-34 65.8 70.7 68.5 49.0 54.2 778
35-39 56.4 68.9 59.4 39.6 51.0 790
40-44 61.8 73.3 66.5 43.2 63.3 415
45-49 59.6 69.5 64.5 45.6 56.1 293
50+ 50.1 59.3 53.3 34.8 40.8 1 100
Livelihood
Agrarian 59.7 69.8 63.3 42.6 51.9 3 886
Pastoralist 38.0 44.4 39.6 18.7 22.8 1 297
Region
Tigray 56.1 69.5 64.4 40.7 63.4 586
Afar 59.1 63.3 53.2 34.0 36.8 326
Amhara 66.4 75.8 73.9 55.4 65.3 967
Oromia 60.0 68.8 62.0 40.5 49.1 1 063
Somali 21.4 28.5 25.7 9.0 4.3 341
Benishangul- 52.8 66.3 62.0 43.9 31.0 370
Gumuz
SNNPR 48.4 60.9 47.6 29.4 31.7 812
Gambela 72.8 80.8 77.7 55.4 75.7 367
Harari 63.8 67.5 69.4 42.5 53.8 351
MEN
National 56.2 64.8 60.3 46.5 53.6
Age
15-19 39.6 74.6 67.7 44.2 51.6 199
20-24 54.1 70.3 62.4 42.7 55.3 564
page- 301
25-29 59.5 63.7 59.8 46.0 57.8 860

30-34 58.9 67.1 58.8 46.2 50.2 632
35-39 55.8 67.9 61.6 47.7 58.8 643
40-44 57.1 69.8 64.2 50.0 53.8 295
45-49 56.5 65.6 62.3 46.0 58.0 190
50+ 53.7 59.2 57.8 45.2 43.2 597
Livelihood
Agrarian 56.8 65.4 60.9 47.0 54.4 3 022
Pastoralist 38.2 44.8 39.5 29.0 26.0 1 008
Region
Tigray 53.5 69.2 65.0 42.3 70.1 405
Afar 54.7 58.4 47.0 33.5 36.2 251
Amhara 65.4 74.7 77.1 62.2 62.8 573
Oromia 57.0 65.7 59.5 45.6 56.0 1 072
Somali 19.1 26.5 23.9 17.4 4.3 211
Benishangul- 44.7 62.8 62.5 39.7 31.7 337
Gumuz
SNNPR 47.6 53.5 46.4 35.9 37.1 697
Gambela 61.1 79.6 70.1 61.0 76.2 143
Harari 65.7 66.2 71.6 52.8 61.7 341
YOUTH GIRLS
National 67.5 76.5 69.7 57.0 62.7
Age
15-19 67.4 76.2 68.2 56.2 60.1 633
20-24 67.8 77.2 74.8 59.7 71.7 170
Livelihood
Agrarian 68.1 77.2 70.4 57.5 63.6 625
Pastoralist 44.7 46.9 43.1 38.6 26.1 178
Region
Tigray 66.5 71.5 57.6 45.1 64.2 111
Afar 58.0 73.4 59.0 53.3 45.2 52
Amhara 74.4 81.7 76.0 64.9 77.1 159
Oromia 72.3 79.3 73.8 62.9 66.4 149
Somali 25.7 25.7 27.4 25.7 0.0 51
Benishangul- 73.1 79.9 69.0 62.4 38.0 50
Gumuz
SNNPR 53.5 69.4 61.4 42.6 42.3 143
Gambela 68.4 80.2 77.7 61.4 76.0 34
Harari 66.0 75.5 76.2 53.3 49.9 54
page-302
In addition to assessing the knowledge of household members, the assessment

also reviewed the percentage of women who received services related to
HIV testing (counseling, testing, and receiving test results), the content of
the counseling services received, and the type of service providers among all
women aged 15 to 49 who received ANC for their latest pregnancy in the 5
years prior to the survey.
Among women in the reproductive age group who received ANC services
for their latest pregnancy, 53.3% received counseling on HIV during ANC.
The likelihood of receiving this service was relatively higher among women
in the 45 to 49 (70.1%) and 35 to 39 (58.6%) age categories than the other
age categories. The percentage of women who received this service varied
significantly across regions: in Somali, only 2.3% of women received this service,
while 89.0% in Gambela and 70.3% in Amhara did. The counseling service was
considered to have been delivered when the provider discussed all 3 topics
(how babies get HIV from their mothers, how to prevent the transmission of
the virus, and how to get tested for HIV) with the pregnant mother. Among
these topics, 66.3% of the women reported that they received counseling on
getting tested for HIV and 63.9% and 60.6% reported that they had received
counseling about how to prevent getting HIV and how babies get HIV from
their mothers, respectively (Table 7-21). HEWs were mentioned as sources of
information about getting tested for HIV during pregnancy, the prevention
of MTCT of HIV, and how babies can get HIV from their mothers by 23.8%,
35.8%, and 31.1% of women who had ANC for their most recent pregnancy,
respectively (Table 7-22).
The assessment also examined the percentage of women who were offered
an HIV test, who were tested for HIV, and who received their HIV test results.
The findings showed that 61% of pregnant women who had ANC visits were
offered an HIV test. Among these women, 89.6% chose to be tested for HIV.
The great majority (94%) of the tested women got post-test counseling and
received their results. Only a few (2.3%) women in Somali were offered the
test for HIV during their ANC visits by health service providers. Among the
women who were offered the test, none accepted the offer. On the other hand,
99.9% of women in the Gambela and 99.2% of women in Tigray were offered
the test. In almost all other regions, a high proportion of women accepted the
offered test (Table 7-22).
page- 303
Table 7-21. Counseling on HIV during ANC for most recent pregnancy
Among all women aged 15-49 who had a ANC visit for their latest pregnancy in the last
5 years, percentage of women who received counseling on specified components of HIV
counseling and the type of providers of the ANC service, by background characteristics
Percentage who got

Percentage who got
counseling on HIV during

Percentage who received
counseling about the HIV
Total number of women

counseling on HIV
prevention, transmission and
during ANC and who got
Antenatal care1
initiated for testing during
information about:
Unweighted
ANC from:
Getting tested
get HIV from
Other health
their mother
getting HIV
they can do
How babies
Things that
to prevent
remember
for HIV
workers
HEWs
Don’t
No
National 63.9 60.6 66.3 53.3 39.7 33.3 25.3 1.8 1962
Age
15-19 45.8 59.6 48.2 35.5 40.0 28.5 29.2 2.3 68
21-24 65.8 58.5 65.3 51.6 41.5 33.5 22.0 3.1 357
25-29 60.9 60.0 65.1 50.7 37.2 34.8 25.9 2.1 591
31-34 63.4 57.3 62.9 51.6 40.8 28.9 28.5 1.9 399
35-39 67.3 64.5 71.9 58.6 38.3 37.2 24.0 0.5 371
41-44 65.4 62.6 70.1 57.8 43.4 31.8 24.5 0.4 128
45-49 72.8 74.6 72.2 70.1 49.9 25.8 24.3 0.0 48
Marital status
Currently
63.9 60.6 66.1 53.4 40.3 32.3 25.6 1.8 1 843
Married
Divorced 68.3 71.7 78.7 56.3 25.0 57.7 17.3 0.0 41
Widowed 43.3 37.6 48.5 30.5 24.1 36.9 39.0 0.0 36
Separated 83.0 69.4 83.6 69.0 35.4 60.3 0.9 3.5 38
Never
78.8 78.8 100.0 78.8 0.1 99.9 0.0 0.0 4
Married
Livelihood
Agrarian 64.4 61.0 66.9 53.8 39.8 33.7 24.9 1.6 1 541
Pastoralist 45.5 41.9 43.0 34.5 34.9 15.7 41.6 7.9 421
Region
Tigray 77.4 73.5 87.9 68.5 36.3 56.2 7.2 0.3 239
Afar 65.3 65.5 56.6 49.2 42.1 31.5 23.4 3.1 125
page-304
Amhara 76.6 77.2 83.4 70.3 42.7 43.1 13.4 0.8 334
Oromia 55.7 53.4 57.0 45.1 33.6 31.4 32.3 2.7 497
Somali 7.3 13.4 2.3 2.3 2.3 11.0 78.6 8.0 25
Benishangul-
62.2 55.1 84.9 52.9 32.7 53.3 12.1 1.9 168
Gumuz
SNNPR 64.9 54.2 62.0 48.3 52.1 18.8 28.1 1.0 350
Gambela 89.0 91.0 95.2 89.0 31.9 63.3 4.8 0.0 81
Harari 63.2 64.2 79.6 55.9 23.3 59.6 16.3 0.8 143
Education
No formal
61.2 58.7 65.1 53.0 39.5 29.8 29.0 1.7 1 163
education
Attended
65.3 60.4 60.9 45.3 43.5 33.3 21.4 1.8 313
grade 1-4
Attended
67.1 62.3 69.7 56.0 38.3 38.9 21.6 1.2 329
grade 5-8
Attended
73.3 71.2 80.5 68.9 34.7 46.5 15.9 2.9 157
grade 9+
Wealth quintile
Lowest 69.9 61.9 62.0 52.4 47.1 29.4 21.8 1.7 304
Lower 67.3 63.6 71.7 56.4 49.6 28.6 19.7 2.1 322
Middle 58.6 57.1 65.0 50.4 39.9 30.0 29.9 0.2 367
Higher 57.2 55.2 60.7 47.6 35.7 30.7 31.6 2.0 491
Highest 69.7 66.6 72.7 60.6 31.7 45.1 20.6 2.6 478
1
In this context, “counseling” means that someone talked with the respondent about all 3 of the
following topics: (a) babies’ getting HIV from their mothers, (b) preventing the virus, and (c)
getting tested for HIV.
3
The denominator for percentages includes women who did not receive ANC for their last birth in
the past 2 years.
Abbreviations: HEW, Health Extension Worker; SNNPR, Southern Nations, Nationalities, and
Peoples Region.
page- 305
page-306
Table 7-22. Providers of information and counseling on HIV during ANC
Among all women age 15-49 who received ANC services for their latest pregnancy in the last 5 years prior to the survey,
percentage who received counseling on HIV during ANC (by type of provider and components of HIV counseling), by
Percentage who got information Percentage who got counseling Percentage who got counseling
about getting tested for HIV about things to do to prevent about how babies get HIV from
Background during ANC from: getting HIV during ANC from: their mothers during ANC from:
characteristic Other Other Other
Don’t Don’t Don’t
HEWs health No HEWs health No HEWs health No

remember remember remember
workers workers workers
National 23.8 42.5 30.6 3.1 35.8 28.2 30.4 5.6 31.1 29.6 32.6 6.8
Age
15-19 18.8 29.5 49.4 2.3 23.9 21.9 36.7 17.5 35.0 24.6 36.8 3.6
21-24 23.1 42.1 29.2 5.5 37.9 27.9 27.2 7.0 29.2 29.2 33.5 8.1
25-29 18.6 46.4 31.9 3.1 34.6 26.3 33.9 5.3 28.9 31.1 31.7 8.3
31-34 22.7 40.5 33.4 3.4 36.7 26.7 29.4 7.2 34.2 23.1 33.3 9.3
35-39 28.4 43.6 26.3 1.8 36.9 30.3 29.8 2.9 30.8 33.7 32.8 2.8
41-44 33.1 37.0 29.5 0.4 37.1 28.3 33.8 0.8 37.1 25.5 36.2 1.3
45-49 43.9 28.3 27.1 0.7 46.9 25.8 22.7 4.5 43.8 30.8 19.8 5.6
Marital status
Currently
24.0 42.1 30.7 3.2 37.3 26.6 30.6 5.4 32.1 28.5 32.8 6.6
Married
Divorced 15.5 63.2 20.1 1.1 18.7 49.6 28.9 2.9 21.1 50.6 20.0 8.3
Widowed 23.9 24.6 51.5 0.0 17.0 26.4 49.1 7.5 16.5 21.1 62.0 0.5
Separated 21.5 62.1 13.0 3.5 23.5 59.5 13.5 3.5 19.6 49.8 5.4 25.3
Never
0.1 99.9 0.0 0.0 0.1 78.6 0.0 21.2 0.1 78.6 0.0 21.2
Married
Livelihood
Agrarian 23.8 43.1 30.2 3.0 36.5 27.9 30.4 5.3 31.5 29.5 32.4 6.6
Pastoralist 24.1 18.9 48.4 8.6 31.6 13.8 42.0 12.6 27.7 14.2 46.7 11.5
Region
Tigray 19.0 68.9 10.4 1.8 32.3 45.2 17.1 5.4 23.2 50.3 19.9 6.6
Afar 31.2 25.5 36.3 7.1 31.4 33.9 26.3 8.3 34.9 30.6 24.5 10.0
Amhara 30.6 52.8 15.1 1.6 37.4 39.3 20.0 3.3 35.7 41.5 19.3 3.4
Oromia 17.0 40.0 38.2 4.8 31.6 24.1 37.0 7.3 26.9 26.4 39.1 7.6
Somali 2.3 0.0 89.6 8.0 2.3 5.0 69.8 22.9 2.3 11.0 78.6 8.0
Benishangul-
17.7 67.2 13.2 1.9 30.2 32.0 30.2 7.7 22.9 32.1 33.6 11.4
Gumuz
SNNPR 33.6 28.5 37.0 1.0 48.8 16.1 32.0 3.1 39.9 14.3 37.1 8.7
Gambela 25.2 70.0 4.8 0.0 23.8 65.2 11.1 0.0 26.2 64.9 9.0 0.0
Harari 18.5 61.2 16.9 3.5 19.0 44.2 25.0 11.8 18.6 45.6 22.8 13.0
Wealth quintile
Lowest 27.4 34.6 35.6 2.5 43.0 26.9 25.8 4.4 36.2 25.7 30.1 8.1
Lower 30.4 41.3 25.2 3.1 44.3 23.1 23.7 9.0 37.1 26.4 27.8 8.7
Middle 25.8 39.3 33.4 1.6 37.3 21.3 37.6 3.8 33.6 23.4 36.0 7.0
Higher 20.3 40.4 35.6 3.7 32.9 24.3 37.8 5.0 27.0 28.3 39.8 4.9
Highest 18.9 53.8 23.3 4.0 29.4 40.3 25.0 5.3 27.2 39.4 27.1 6.4
Abbreviations: ANC, antenatal care; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page- 307

Table 7-23. Pregnant women counseled and tested for HIV

Among all women age 15-49 who received ANC service for their latest pregnancy in the
last 5 years preceding the survey, percentage who received an HIV test during ANC for
their most recent pregnancy; percentage who received their test results, by background
characteristics
Percentage
Percentage who Percentage who
who received
were offered were tested for Total number of
Background counseling on
an HIV test HIV during ANC women receiving
characteristic HIV during
and got tested and who received ANC
ANC and were
during ANC test results
offered a test1
National 61.0 89.6 94.0 1 962
Age
15-19 47.9 97.4 79.3 68
21-24 65.1 89.6 94.9 357
25-29 57.0 90.4 94.9 591
31-34 61.0 90.7 94.3 399
35-39 63.3 85.3 92.9 371
41-44 68.1 92.9 94.8 128
45-49 51.6 95.6 89.6 48
Marital status
Currently
60.7 89.0 94.6 1 843
Married
Divorced 69.7 98.5 77.3 41
Widowed 49.7 99.7 100.0 36
Separated 80.0 99.7 81.7 38
Never Married 100.0 100.0 100.0 4
Livelihood
Agrarian 61.5 89.6 94.1 1 541
Pastoralist 40.1 86.1 90.9 421
Region
Tigray 83.1 99.2 93.5 239
Afar 47.5 86.1 96.2 125
Amhara 77.8 97.1 93.7 334
Oromia 54.3 83.2 96.9 497
Somali 2.3 0.0 0.0 25
Benishangul-
75.0 96.8 98.7 168
Gumuz
SNNPR 50.5 87.4 87.6 350
Gambela 92.8 99.9 100.0 81
Harari 71.7 89.6 95.1 143
page-308
Education
No formal
59.7 89.6 91.9 1 163
education
Attended grade
55.4 88.7 95.5 313
1–4
Attended grade
63.0 85.5 95.7 329
5–8
Attended grade
80.1 98.5 99.7 157
9+
Wealth quintile
Lowest 58.4 87.8 87.4 304
Lower 65.5 88.4 95.2 322
Middle 53.5 92.3 93.9 367
Higher 55.3 88.2 94.3 491
Highest 72.3 90.7 96.5 478
1
In this context, “counseling” means that someone talked with the respondent about all 3 of the
following topics: (a) babies’ getting HIV from their mothers, (b) preventing the virus, and (c)
getting tested for HIV.
2
Women were asked whether they received an HIV test during labor only if they gave birth in a
health facility.
3
The denominator for percentages includes women who did not receive ANC for their last birth in
the past 2 years.
7.2.3 Tuberculosis Prevention and Control
About 91% of women and 95% of men and youth girls knew that TB can be
transmitted from person to person; 85% of women, 91% of men, and 88% of
youth girls know that the disease is preventable. Generally, people living in
agrarian regions seem to be more knowledgeable about TB prevention than
those residing in pastoralist areas. There is also marked variation in knowledge
about TB across regions. Knowledge about TB tends to increase with level of
education and wealth quintile for all categories of respondents (Table 7-24).
Overall, only a small proportion of women (9%), men (15%), and youth girls
(19%) have comprehensive knowledge of all 3 methods for preventing TB
(opening windows and doors, covering one’s mouth while coughing, and treating
sick patients), and the treatment of sick patients is the most known of the 3
page- 309
methods of prevention across all 3 population groups. Although knowledge

about the 3 methods of TB prevention is very low across all regions, a notable
difference among regions is also observed. Generally, men and youth girls in
agrarian areas are relatively more knowledgeable than their counterparts in
pastoralist areas, while women in pastoralist areas are more knowledgeable
than those residing in agrarian areas. Moreover, all 3 methods of TB prevention
are relatively better known in Harari across all population groups and least
known in Benishangul-Gumuz among women and in Gambela among men
and youth girls (Table 7-25).
page-310
Table 7-24. Knowledge of tuberculosis prevention methods
Knowledge of women, men, and youth girls about modes of transmission and prevention of TB, by background
characteristics
Percentage of those who:
Background
characteristic
of TB
of TB
of TB
Number of
Number of
Number of
respondents
respondents
respondents
be prevented
be prevented
be prevented
Have ever heard

Have ever heard
Have ever heard
person to person
person to person
person to person
Know that TB can

Know that TB can
Know that TB can
Know that TB can
Know that TB can
Know that TB can
be transmitted from
be transmitted from
be transmitted from
National 75.4 90.8 85.2 6 430 89.1 94.7 90.9 4416 83.8 94.6 88.3 900
Age
15-19 62.3 84.1 80.0 281 93.6 98.1 98.1 23 81.8 94.3 87.3 720
21-24 75.1 90.1 86.4 763 89.7 86.6 78.7 211 90.7 95.4 91.4 180
25-29 73.4 91.3 86.5 1 209 90.1 95.6 92.6 565 N/A Na N/A N/A
31-34 73.5 92.1 87.3 943 93.6 96.2 90.4 660 N/A Na N/A N/A
35-39 80.1 94.3 88.9 923 86.6 94.5 91.3 635 N/A Na N/A N/A
41-44 76.7 93.5 84.6 519 92.1 94.9 91.3 619 N/A Na N/A N/A
45-49 77.7 93.3 86.6 353 92.5 95.5 92.6 417 N/A Na N/A N/A
50+ 75.2 86.4 80.2 1 439 85.4 94.2 90.6 1 285 N/A Na N/A N/A
Livelihood
Agrarian 75.3 91.6 86.3 4 421 89.4 95.3 91.6 3157 84.1 95.3 89.0 658
Pastoralist 76.7 73.5 61.2 2 009 82.4 79.1 69.7 1 259 72.5 68.1 61.2 242
Region
Tigray 90.7 87.8 89.4 607 96.0 93.9 91.4 407 82.7 82.2 84.3 117
page- 311

page-312
Afar 82.7 90.8 76.9 399 93.6 88.8 79.4 275 90.8 91.7 81.0 61
Amhara 82.1 90.3 87.0 1 060 90.9 92.5 89.7 603 87.0 91.9 84.9 169
Oromia 73.0 92.9 87.0 1 319 90.3 96.4 92.9 1 139 85.8 98.3 96.1 162
Somali 81.3 59.3 47.8 798 79.0 57.6 45.2 376 67.2 46.7 44.6 97
Benishangul-
90.9 81.1 82.7 406 96.6 90.6 91.9 340 97.6 93.8 94.7 50
Gumuz
SNNPR 66.9 91.2 81.2 1 009 83.7 95.2 89.4 759 77.6 98.7 82.9 150

Gambela 81.6 90.8 87.0 417 84.2 95.1 91.4 157 93.1 93.7 81.0 36
Harari 84.5 95.7 92.2 415 94.8 97.8 93.0 360 93.2 93.8 91.9 58
Education
No formal
73.4 89.3 82.7 5 226 86.2 92.7 87.4 2 702 66.8 93.7 76.8 192
education
Grade 1-4 74.2 92.2 88.2 823 87.4 95.1 93.3 741 75.2 96.2 81.6 160
Grade 5-8 83.7 94.3 90.2 684 92.6 96.0 93.4 838 82.9 94.2 91.6 421
Grade 9+ 92.0 97.6 99.1 310 96.6 99.0 95.4 524 96.3 94.5 89.7 247
Wealth quintile
Lowest 68.7 82.0 74.8 1 343 82.2 89.6 87.0 708 77.9 95.7 77.1 143
Lower 72.9 87.4 83.4 1 291 84.7 94.7 91.8 843 73.6 89.3 83.7 157
Middle 72.9 91.6 85.6 1 278 90.7 93.8 88.5 881 89.6 93.9 86.1 190
Higher 78.3 94.4 89.0 1 274 90.7 95.8 90.9 1 028 80.9 93.1 92.6 191
Highest 82.6 95.0 89.5 1 244 93.6 97.3 94.6 956 90.1 98.2 92.7 219
Abbreviations: TB, tuberculosis; SNNPR, Southern Nations, Nationalities, and Peoples Region.
Table 7-25. Knowledge of women, men, and youth girls about modes of transmission and prevention of
tuberculosis, by background characteristics

Percentage of those who mentioned TB Percentage of those who mentioned TB can be Percentage of those who mentioned
can be prevented by: prevented by: TB can be prevented by:
Background
characteristic
Number
of youth
other
other
other
girls
methods
methods
methods
treatment
treatment
Treatment
ventilation
ventilation
ventilation
& doors for

& doors for
& doors for
while coughing
while coughing
while coughing
covering mouth
covering mouth
covering mouth
all 3 prevention
all 3 prevention
all 3 prevention
Number of men
Opening windows
Opening windows
Opening windows
Number of women
National 18.7 56.1 67.9 8.8 20.4 4140 27.6 63 71.6 14.7 18.8 3414 31.1 64.5 74.3 19 14.6 616
Age
15-19 24.8 46.8 75.8 11.8 11 164 41.8 76.2 61.8 31.5 6.5 18 31.2 64.3 72.4 19.2 12.8 477
21-24 17.4 62.9 67.1 6.7 16.3 501 39.6 59.5 81.7 25.7 7.1 160 30.8 64.9 80 18.5 19.9 139
25-29 24.8 60.8 67.6 12 21.5 781 34 67.2 72.9 20.1 16.3 433 N/A N/A N/A N/A N/A N/A
31-34 18.6 57.4 64.7 8.2 18.7 615 31.4 63.1 74.1 17.6 18 525 N/A N/A N/A N/A N/A N/A
35-39 18.6 57.5 69.6 7.9 18.3 635 29.3 66.3 70.5 13.6 18.5 501 N/A N/A N/A N/A N/A N/A
41-44 14 56.1 66.5 7.1 27.4 334 26.8 65.8 71.9 10.6 17.2 483 N/A N/A N/A N/A N/A N/A
45-49 18.1 54.1 63.9 10.9 25.7 240 26.3 63.6 71.4 15 20.4 337 N/A N/A N/A N/A N/A N/A
50+ 16.4 48.6 70.2 8.3 21.2 870 22.2 58.5 69.7 12.7 21.8 957 N/A N/A N/A N/A N/A N/A
Livelihood
Agrarian 18.5 56.3 67.6 8.7 20.8 3 100 27.6 63.3 71.6 14.8 19 2 633 31.2 64.9 74.1 19.1 14.7 489
Pastoralist 24.3 47.4 77.5 11.4 9.1 1 040 27.6 53.2 73.6 13.5 11.3 781 27.5 43.5 82 17.4 5.3 127
page- 313

page-314
Region
Tigray 14.8 53.6 36.3 13.4 54.6 489 21.4 62.1 45 15.5 52.2 356 23.1 72 44.1 20.9 46.7 80
Afar 26.2 42.6 83 7.3 1.8 246 22.5 46.4 85.7 6.7 2.2 202 19.6 39.9 81 4.9 0 42
Amhara 16.4 46.1 66 7.6 21.2 752 26.2 48.4 66.7 10.7 21.1 487 36.5 59.1 66.7 20.9 6.5 120
Oromia 23.2 65.4 74.8 10.4 12.9 785 32.9 71.9 76.3 18.7 14.1 913 34.8 66.7 82.9 19 9.5 113
Somali 19.9 24.3 98.3 16.6 1.2 344 29.1 32.9 96 23.4 0 151 26.1 26 100 23.3 0 32
BG 2.1 25.1 62 0.7 49.4 307 4.9 41 63.3 3.2 50.1 307 8.3 57.9 60.9 6.7 45.6 48

SNNPR 12.5 52.8 62.9 4.6 27.3 571 16.8 56.1 70.4 8.4 21 550 18.8 66.8 75.6 16.3 26.8 103
Gambela 32.9 54.6 55.8 8.4 19.9 313 34.5 39 61.4 1.3 14.4 124 19.4 70.4 43.6 0 15 30
Harari 49.3 90.4 62.8 28.2 11.9 333 54.7 88.1 61.9 30.5 18.8 324 59.8 90.9 58.6 40.5 6.1 48
Education
None 18.3 52.5 67.5 10.3 19.7 2 904 21.2 56.7 68.8 9.7 18.3 1 721 21.8 71.5 65.4 9.6 3.1 69
Grade 1-4 16.7 56.4 74.4 17.5 20.9 512 23.7 63.3 70.5 11.1 22.9 560 27.2 63 67.8 15.3 20.3 82
Grade 5-8 20.1 64.1 60.9 6.7 24.7 467 28.9 64.1 77.8 17.5 20 691 31.4 58.5 75.5 16.8 14.1 280
Grade 9+ 26.7 79 71.6 9.9 17.8 257 50.9 81 71.6 30.9 13 442 34.3 71.7 77.2 25.7 15.3 185
Wealth quintile
Lowest 16.3 48.4 66 10.4 18.9 727 24.6 47.5 70.5 16.8 21.3 455 39 76.9 68 23.5 3.8 78
Lower 18.4 55 67.9 7.5 18.9 787 27.2 61.6 69.6 12.8 16.1 624 28.8 66.7 63.6 20.8 14 90
Middle 19 55 64.8 8.7 23.4 817 24.4 61.3 68.5 12.4 20.7 686 36.3 55.6 70.7 22.9 16.9 131
Higher 16.4 54.6 69.6 7.5 22.4 881 27.8 65.4 73.4 15 19 816 26.2 66.9 77.1 17.6 13.4 148
Highest 22.6 63.8 70 10.3 17.8 928 31.8 70.6 74.3 16.7 17.7 833 29.9 63.8 80.6 15.3 17.6 169
N/A- not applicable
Abbreviations: TB, tuberculosis; SNNPR, Southern Nations, Nationalities, and Peoples Region.
7.2.4 The role of Health Extension Workers in preventing and con-

trolling communicable diseases
Health extension workers as sources of information in communicable disease
prevention and control
MALARIA PREVENTION AND CONTROL
In assessing the contribution of the HEWs in malaria prevention and control,

households were asked whether they had seen a demonstration of how to use
ITNs and who had conducted it. Among the respondents, 48.7% had ever seen
a demonstration on how to use ITNs. Among the households who had seen
a demonstration of ITN use, most (86.7%) had received the education from
HEWs, 87.3% in agrarian and 66.7% in pastoralist communities. Families or
neighbors were the second most common sources of education on ITN use
(8.1%) and were more common in pastoralist (30.4%) than agrarian (7.4%)
communities. The contribution of HEWs varied significantly among regions:
Somali (36.8%) and Afar (25.8%) had the lowest percentage of households
receiving education from families or neighbors, and Oromia (87.1%) and the
SNNPR (91.2%) had the highest (Table 7-26).
Table 7-26. ITN demonstration, by source of education and background

characteristics
Percentage Among households who saw demonstration of how to use
of households ITNs, percentage who received the demonstration from
Background
who saw Home/family/ Health Total
characteristic Health
demonstration relatives/ post/ Other number of
on ITN use center
neighbors HEWs households
National 48.7 8.1 86.7 4.2 1.0 2106
Livelihood
Agrarian 50.3 7.4 87.3 4.3 1.0 1673
Pastoralist 23.1 30.4 66.7 2.4 0.6 433
Region
Tigray 46.1 8.4 79.3 11.5 0.8 208
Afar 37.6 73.1 25.8 0.0 1.2 116
Amhara 49.5 7.3 86.7 5.9 0.0 330
Oromia 63.8 9.2 87.1 2.2 1.5 351
Somali 6.6 56.4 36.8 6.8 0.0 43
page- 315
Benishangul-
84.4 18.7 80.8 0.5 0.0 354
Gumuz
SNNPR 36.3 3.8 91.2 3.4 1.6 352
Gambela 34.1 8.3 70.2 21.6 0.0 174
Harari 47.9 43.1 51.6 4.5 0.9 178
Wealth quintile
Lowest 37.1 9.6 86.9 2.0 1.5 265
Lower 38.5 5.0 91.5 2.5 1.0 339
Middle 43.7 9.5 84.5 4.9 1.1 450
Higher 55.1 8.6 85.5 5.4 0.5 517
Highest 65.3 7.7 86.6 4.9 0.9 535
Abbreviations: ITN, insecticide-treated net; SNNPR, Southern Nations,

HIV/AIDS PREVENTION AND CONTROL
The provision of HIV counseling and testing for pregnant mothers during
ANC was assessed as part of the contribution of the HEWs in preventing
and controlling HIV/AIDS. The findings of the assessment showed that 39.7%
of the women who had received ANC in their latest pregnancy during the 5
years prior to the survey had received counseling services from HEWs, while
33.3% had received counseling from other health service providers. The HEWs
provided the services to 39.8% of women in agrarian and 34.9% of women in
pastoralist communities (Figure 7-4).
Figure 7-4: Providers of HIV counseling at ANC, by livelihood
Abbreviation: ANC, antenatal care.
page-316
TUBERCULOSIS PREVENTION AND CONTROL
HEWs are expected to provide information and some specified services in the
prevention and control of TB. In this assessment, the roles of the HEWs as
sources of information and providers of services were assessed. HEWs were
the primary sources of information about TB for 34.3% of women, 33.9% of
men, and 12.3% of youth girls. A few additional respondents also mentioned
other HEP staff (e.g., WDA leaders) as sources of information about TB (Table
7-27).
The role of HEWs in TB diagnosis and treatment is to screen households for

symptoms of TB, refer those with symptoms to higher-level health facilities,
and provide treatment for confirmed cases. At the time of the survey, about
5% of households had at least 1 member with cough of >2 weeks, and about
half of those (51%) reported that their symptomatic member had sought
medical care. HCs were the most commonly (60%) used level of care as first
contact for symptomatic members of the households, while HPs were the least
commonly (10%) used. The use of HPs as first contact for symptomatic TB
patients was far higher in pastoralist areas (34.9%) than in agrarian areas
(9.7%; Table 7-29, Table 7-30).
In the last 3 years, about 3% of households reported having had at least 1

member diagnosed with TB. HEWs were reported to have provided assistance
to those households in different ways, ranging from referring patients for
diagnosis (24.7%) to providing isoniazid preventive therapy (IPT) for children
under 5 (1.1%). Only 18% and 13% of those households received TB screening
and treatment follow-up services, respectively, from HEWs (Table 7-28).
Among households who reported having a member diagnosed with TB, only
27.2% mentioned that the diagnosed patients were receiving treatment. The
majority receiving treatment were being treated in either a HC (49.6%) or
a hospital (38.5%). HPs were the rarest source (2.9%) of treatment of TB
patients (Table 7-31).
page- 317
Table 7-27. Source of information about tuberculosis (TB)
Percentage of women, men, and youth girls receiving information about TB, by their main
source of information and background characteristics
Percentage of those who got information about TB from:
Health Extension Health Extension
characteristics
Program Program
Family/ Relative/
Background
Public Meeting
No of Women
Mass Media
HC/Hosp.
HC/Hosp.
Neighbor
Other
Community
Community
other than
other than
Workers,
Workers,
leaders
leaders
leaders
leader
HEWs
HEWs
WDA
WDA
WDA
WDA
WOMEN MEN
National 34.3 1.4 1.0 12.5 11.4 5.2 27.3 6.8 5 142 33.9 1.6 2.8 20.4
Livelihood
Agrarian 34.5 1.3 0.9 12.8 11.6 4.9 27.1 6.9 3 576 33.9 1.4 2.7 20.5
Pastoralist 30.6 4.7 1.9 6.6 6.9 12.7 32.0 4.6 1 566 35.4 7.6 7.4 17.4
Region
Tigray 38.3 2.2 0.8 17.8 3.7 2.5 28.1 6.6 551 39.8 2.2 5.3 29.5
Afar 29.9 2.9 0.6 30.5 5.3 3.4 26.8 0.6 327 28.2 2.1 1.9 47.3
Amhara 29.8 1.2 0.8 16.4 4.0 3.7 37.0 7.2 866 32.0 2.0 4.0 27.6
Oromia 33.9 1.3 0.7 8.1 21.4 7.6 19.9 7.1 956 31.4 0.6 1.6 10.4
Somali 27.8 7.4 1.9 0.5 6.2 18.5 37.4 0.3 637 48.1 21.1 14.6 14.2
BG 36.9 0.4 3.4 10.1 1.0 6.1 36.0 6.2 363 45.4 1.0 2.5 25.7
SNNPR 42.5 1.2 1.9 16.0 2.7 1.4 27.8 6.6 730 40.0 2.5 3.5 35.8
Gambela 40.5 0.9 2.1 24.9 4.8 2.6 20.6 3.8 356 34.4 0.5 2.1 45.8
Harari 21.6 0.4 1.4 28.5 21.2 1.1 22.7 3.2 356 15.4 2.6 6.8 38.6
Education
None 34.7 1.3 0.7 12.8 9.8 5.9 29.7 5.2 3 748 34.8 1.5 2.5 17.6
Grade 1-4 35.2 2.5 1.9 10.8 15.2 5.0 24.4 5.0 600 31.9 3.2 2.4 21.4
Grade 5-8 31.8 0.8 0.8 11.9 12.5 3.2 23.3 15.8 531 33.8 1.1 2.9 22.3
Grade 9+ 33.1 2.1 1.8 15.5 19.8 2.0 12.1 13.5 263 33.9 0.7 4.6 25.6
Abbreviations: BG, Benishangul-Gumuz; HEW, Health Extension Worker; WDA,

Women’s Development Army;HC, health center; Hosp, hospital; TB, tuberculosis.
page-318
Health Extension
Program
Family/ Relative/
Family/ Relative/
Public Meeting
Public Meeting
No of Youth girls
Mass Media
Mass Media
No of Men
HC/Hosp.
Neighbor
Neighbor
HEWs
Other
Other
Community
other than
Workers,
leaders
leaders
WDA
WDA
MEN YOUTH GIRLS
35.0 18.6 28.6 7.0 3 954 12.3 1.4 2.2 8.1 24.9 10.8 36.4 32.7 742
35.4 18.5 28.4 7.1 2 889 11.7 1.2 2.0 8.0 25.2 10.3 36.3 33.3 563
25.2 23.4 33.8 6.8 1 065 31.8 8.8 8.5 9.5 14.7 27.2 39.3 10.5 179
17.3 16.1 30.7 8.9 392 8.1 0.0 6.3 7.7 12.6 3.8 45.9 46.9 96
22.4 21.2 40.5 0.6 249 14.3 0.0 4.4 14.9 9.0 12.9 75.2 7.4 52
12.7 14.7 37.4 7.1 545 14.3 0.0 0.0 9.7 13.3 5.8 29.3 38.1 146
51.7 21.8 23.3 5.8 1 018 9.8 2.8 2.4 3.3 37.5 18.0 33.8 21.4 125
17.3 28.5 39.0 0.3 294 35.2 11.6 14.3 12.0 18.1 44.3 36.2 0.0 63
12.0 18.1 38.1 9.7 328 36.5 0.0 0.0 0.0 7.9 0.0 24.6 42.7 49
21.0 14.6 31.8 10.0 649 13.5 0.4 2.7 14.1 22.8 4.5 47.2 44.0 124
4.4 4.6 33.2 5.6 135 27.6 0.0 0.0 56.7 2.9 2.9 19.8 5.7 34
67.7 15.5 33.6 8.9 344 16.1 4.9 2.5 34.8 61.9 12.4 40.8 28.0 53
26.4 19.8 35.1 3.2 2 135 11.0 6.5 1.6 4.3 20.9 10.5 52.7 18.6 103
36.2 18.8 30.9 5.7 618 13.3 4.4 4.1 4.2 29.9 21.9 43.5 12.2 109
42.2 18.3 22.9 10.8 740 12.1 0.2 0.7 8.9 16.1 10.0 32.5 39.6 320
50.9 14.9 12.8 15.6 461 12.3 0.3 3.4 9.6 35.5 6.9 34.1 36.2 210
page- 319
Table 7-28. The role of HEWs in TB diagnosis and management
Percentage of households who had at least 1 TB patient and those reporting having had
assistance from HEWs in the past 3 years, by household background characteristics
Among households who reported to have had at least 1 TB patient, percentage

of those who were assisted by HEWs on:
Characteristic
Background
to test for HIV

members for
Referral for
Monitoring
medication
households
Counseling
Treatment
Education
household
Screening
diagnosis
Screened
Number
children
under 5
Anti-TB
IPT for
Health
supply
for TB
Total
% of
TB
National 3.0 5143 17.8 24.7 12.0 13.1 11.2 18.5 6.7 1.1 138
Livelihood
Agrarian 3.0 3576 17.4 24.2 11.8 13.0 11.3 18.0 6.7 0.8 102
Pastoralist 1.6 1567 36.4 47.1 19.0 18.8 10.0 37.0 9.0 15.5 36
Region
Tigray 1.3 551 0.0 28.5 0.0 0.0 0.0 0.0 0.0 0.0 6
Afar 3.3 327 28.5 40.1 13.7 26.6 12.8 58.6 15.5 28.5 12
Amhara 2.2 866 32.0 46.6 32.8 32.0 20.7 32.0 26.8 3.1 18
Oromia 2.7 956 21.4 22.5 8.6 8.6 16.3 16.9 1.5 0.3 24
Somali 1.4 637 52.2 70.1 13.5 14.8 2.5 22.4 13.5 11.0 10
Benishangul-
3.1 363 6.6 27.3 15.9 19.3 12.0 34.0 3.5 0.0 9
Gumuz
SNNPR 5.7 730 4.9 11.4 3.2 6.9 0.2 12.1 0.0 0.0 31
Gambela 5.6 356 32.7 38.2 38.2 32.0 32.0 38.2 32.0 23.1 22
Harari 1.9 356 19.1 0.0 19.1 19.1 19.1 19.1 0.0 0.0 6
Wealth quintile
Lowest 3.9 1031 20.9 27.5 15.6 25.5 15.6 20.5 3.6 0.1 34
Lower 1.9 996 3.9 14.1 11.5 4.0 3.0 11.5 9.0 1.5 22
Middle 2.9 1000 0.7 8.0 0.5 0.8 0.5 4.1 0.5 0.2 34
Higher 3.6 1046 26.3 33.6 11.9 12.0 13.3 22.7 8.5 2.7 23
Highest 2.5 1070 27.9 31.9 20.1 19.7 19.4 28.7 12.6 0.3 25
Abbreviations: TB, tuberculosis; HEW, Health Extension Worker; SNNPR, Southern Nations,
Nationalities, and Peoples Region; IPT, isoniazid preventive therapy.
page-320
Table 7-29. Tuberculosis (TB) symptoms and health seeking
Percentage of households having a member with symptoms or a diagnosis of TB and care-

seeking behavior, by household background characteristics
Among households
Percentage
Percentage having a member
of Number of
households with cough of > 2
households households
currently weeks, percentage
Background which get Number of with at least
having at least of those who
characteristic regular households 1 member
1 member with reported to have
screening having cough
cough for > 2 had their member
service for for > 2 weeks
weeks sought care from
TB by HEW
health facility
National 5.2 5.1 5 143 51.0 236
Livelihood
Agrarian 5.4 5.1 3 576 52.4 176
Pastoralist 1.9 4.8 1 567 20.2 60
Region
Tigray 3.4 8.8 551 57.0 48
Afar 2.4 2.2 327 41.7 11
Amhara 6.7 5.7 866 59.0 47
Oromia 6.0 3.6 956 62.7 38
Somali 0.4 5.0 637 7.2 29
Benishangul-
24.1 4.5 363 89.4 14
Gumuz
SNNPR 1.9 6.7 730 25.4 29
Gambela 9.8 3.4 356 37.6 14
Harari 3.0 1.6 356 71.9 6
Wealth quintile
Lowest 1.7 9.0 1 031 51.7 57
Lower 3.4 2.9 996 47.7 30
Middle 6.2 5.6 1 000 45.3 55
Higher 6.4 5.0 1 046 43.3 55
Highest 7.1 3.9 1 070 69.8 39
Abbreviations: TB, tuberculosis; HEW, Health Extension Worker; SNNPR, Southern Nations,
page- 321
Table 7-30. Place of first contact for tuberculosis (TB)-related care and reasons for bypassing health posts
Among households with a member
page-322
currently having cough of > 2 Among those who sought care, percentage of those who
weeks duration and sought care, mentioned the reason for not seeking care from HEWs/HP first
percentage of those who reported to was because:
have had their member sought care
first in:
HEW/HP
HP
HC
refer
Hosp.
Other
Other
closed
HEWs
HEWs
service
Private
HP was
anyway
Total who sought care
have no
capacity
to go, as
No need
available
at the HP
HEW was
to provide
is close by
HC/ Hosp.
% of those referred from

Service not
the specific
clinic/ hosp.
unavailable
Background
characteristics
National 10.1 59.8 15.8 12.6 1.7 9.0 2.6 1.3 49.6 12.3 6.8 21.6 6.0 113

Livelihood
Agrarian 9.7 60.7 15.4 12.5 1.8 8.9 2.3 1.2 49.4 12.4 6.8 21.8 6.0 100
Pastoralist 35.0 8.9 40.5 15.6 0.0 12.2 27.3 5.4 62.8 0.0 0.0 0.0 4.6 13
Region
Tigray 6.4 64.9 22.0 6.8 0.0 4.0 0.0 0.0 27.6 18.5 0.0 26.7 27.1 29
Afar 100.0 0.0 0.0 0.0 0.0 N/A Na N/A N/A N/A N/A N/A N/A 3
Amhara 4.6 80.9 8.2 6.3 0.0 6.9 3.3 3.0 37.4 13.1 0.0 37.1 6.0 27
Oromia 14.5 43.8 18.0 23.7 0.0 15.7 2.7 0.0 60.5 12.7 15.6 8.6 0.0 20
Somali 15.4 0.0 0.0 84.6 0.0 0.0 100.0 0.0 0.0 0.0 0.0 0.0 0.0 2
Benishangul- 21.3 45.1 7.8 22.2 3.6 6.6 0.0 0.0 3.6 59.3 6.9 25.4 4.9 11
Gumuz
SNNPR 15.1 42.2 27.9 0.0 14.8 0.0 0.0 0.5 86.5 0.0 12.5 0.0 0.4 11
Gambela 24.7 75.3 0.0 0.0 0.0 38.0 0.0 0.0 87.3 0.0 0.0 12.8 0.0 6
Harari 0.0 77.5 0.0 22.5 0.0 0.0 28.8 0.0 36.9 34.3 0.0 0.0 0.0 4
Wealth quintile
Lowest 8.5 74.9 8.2 8.4 0.0 8.4 4.7 0.0 52.5 9.1 0.0 25.3 8.5 26
Lower 2.9 53.7 32.9 10.6 0.0 0.1 2.7 0.0 62.5 2.7 0.0 19.3 12.8 13
Middle 25.6 60.4 13.8 0.1 0.0 5.6 0.0 0.4 66.0 5.4 20.2 1.7 6.3 28
Higher 3.6 53.6 14.0 28.7 0.1 15.6 4.6 6.3 16.7 25.8 17.7 28.8 0.1 23
Highest 7.7 49.1 20.8 15.1 7.3 10.1 0.1 0.0 55.0 14.6 0.0 26.1 4.2 23
Abbreviations: TB, tuberculosis; HEW, Health Extension Worker; Hosp., hospital; HC, health center; HP, health post; SNNPR, Southern
Table 7-31. Places of tuberculosis (TB) treatment
Treatment percentage of households whose TB-diagnosed member was on treatment and place
of treatment, by household background characteristics
Percentage Percentage of households reporting that
of their TB-diagnosed member was being
households treated in/by:
Background whose TB- Number of
Total
characteristic diagnosed respondents Private
member HEWs/
HC Hospital clinic/ Other
was in HP
hospital
treatment
National 27.2 113 2.9 49.6 38.5 4.3 4.7 21
Livelihood
Agrarian 27.6 100 2.8 49.5 38.6 4.3 4.7 19
Pastoralist 4.0 13 22.7 77.3 0 0 0 2
Region
Tigray 5.8 29 0 57.5 42.5 0 0 3
Afar 9.3 3 100 0 0 0 0 1
Amhara 21.1 27 0 45.0 25.1 13.8 16.1 5
Oromia 43.1 20 4.7 57.3 38.0 0.0 0 6
Somali 0.0 2 N/A N/A N/A N/A N/A N/A
Benishangul-
17.2 11 0 0 0 100 0 1
Gumuz
SNNPR 17.2 11 0 2.5 97.5 0 0 3
Gambela 24.7 6 0 100 0 0 0 1
Harari 11.8 4 0 100 0 0 0 1
Wealth quintile
Lowest 36.4 26 0 80.6 6.9 0 12.5 6
Lower 32.6 13 0 7.9 92.1 0 0 2
Middle 26.2 28 15.5 32.9 30.0 21.5 0 8
Higher 19.6 23 0 48.3 51.7 0 0 2
Highest 20.9 23 0 30.4 67.8 1.8 0 3
Abbreviations: TB, tuberculosis; HEW, Health Extension Worker; HP, health post; SNNPR, Southern
page- 323
HEWS’ ROLE IN THE MANAGEMENT OF ILLNESSES,

INJURIES, AND DEATH
To assess the role of HEWs in the management of illnesses, injuries, and

death, the occurrences of these health events were assessed. In addition, the
perceived causes of morbidity and mortality and the health-seeking behavior
of household members were also assessed.
OCCURRENCE OF ILLNESSES, INJURIES AND DEATH
Among the households covered by the assessment, 18.8% had at least 1 member
of their household who was sick: 19.0% in agrarian and 12.8% in pastoralist
communities. The occurrence of self-reported illness in the households differed
among regions, ranging from 10.7% in Afar and 14.4% in Amhara to 24.6% in
Benishangul-Gumuz and 32.4% in Gambela. It also increased with the level of
wealth. For acute illnesses, the mean number of days the household member
felt sick before the survey was 9.9 days. As to the respondents’ perceptions,
the most common causes of illness was chronic illnesses (26.4%), followed by
common cold (23.8%), diarrhea (10.2%), and other acute illnesses (6.7%; Table
7-32).
An injury or accident had occurred to at least 1 household member in 3.2%

of the households in the 1 year prior to the survey. The highest percentage of
injuries and accidents occurred in Benishangul-Gumuz (8.0%), followed by the
SNNPR (5.0%). The most common accident or injury was physical violence
or assault (31.7%), followed by fire or burning (21.2%) and an accidental fall
(17.9%). Physical violence or assault decreases with an increase in wealth
quintile, from 47.5% among those in the lowest quintile to 4.4% in the highest
(Table 7-33).
page-324
Table 7-32. Morbidity among household members
Percentage of household members with illness, perceived cause of illness, treatment seeking, and type of services received in the past 12 months, by
Perceived causes of sickness Percentage of persons who received
month
sought
illnesses)
Drug
Other
Other
Background
cause)
Advise
persons
Malaria
illnesses
Characteristic
Percentage of
treatment was
Diarrhea
stay before the

Procedure
sick in the last 1

members of the
(no known
family for whom

Number of
Pneumonia
Mean number of
survey (for acute
Other acute
Percentage of sick
households with at
days the symptoms
Common cold
least 1 member was

Chronic Illness
Total
National 18.8 9.9 69.4 23.8 8.4 10.2 2.4 26.4 6.7 22.2 91.1 4.1 4.0 0.8 822
Livelihood
Agrarian 19.0 8.7 69.6 23.8 7.6 10.0 2.3 26.8 6.8 22.7 888 91.0 4.2 4.0 0.7 660
Pastoralist 12.8 13.1 62.1 20.9 35.3 16.7 6.3 12.6 2.0 6.3 276 93.0 0.6 3.6 2.8 162
Region
Tigray 24.0 13.4 62.1 37.9 15.0 7.7 6.7 9.6 12.1 10.9 139 92.6 3.1 2.6 1.8 85
Afar 10.7 11.5 72.7 16.3 18.1 25.9 12.9 23.6 0.0 3.3 35 97.6 0.0 2.4 0.0 24
Amhara 14.4 9.5 69.0 23.0 0.8 13.7 5.0 19.8 7.8 30.0 141 85.9 8.1 4.6 1.5 100
Oromia 22.0 13.8 72.9 25.5 3.8 6.2 1.3 31.8 7.0 24.5 223 93.4 3.9 2.6 0.1 168
Somali 15.0 17.4 50.6 25.2 51.3 5.7 5.2 10.2 0.2 2.3 137 98.6 1.4 0.0 0.0 71
Benishangul- 24.6 6.0 89.5 11.9 18.4 23.6 12.7 5.6 16.3 11.5 118 95.4 2.4 0.8 1.5 107
Gumuz
SNNPR 16.5 9.1 63.0 14.8 23.4 18.8 0.9 26.0 3.1 12.9 165 88.1 0.5 9.4 2.0 103
Gambela 32.4 4.9 81.0 0.7 83.7 10.1 1.4 3.6 0.0 0.6 146 99.8 0.2 0.0 0.0 120
Harari 15.0 10.4 75.4 23.1 13.0 14.9 2.2 26.6 4.4 16.0 60 99.1 0.0 0.9 0.0 44
Wealth quintile
Lowest 15.3 10.5 46.4 10.4 6.8 21.0 2.3 35.7 5.7 18.2 243 95.2 0.2 1.0 3.6 143
Lower 13.5 9.0 68.0 34.0 6.8 9.5 2.9 23.1 3.5 20.2 190 92.3 2.2 5.5 0.0 134
Middle 16.5 8.0 67.8 26.7 9.2 11.5 1.9 15.0 10.6 25.2 234 87.5 2.5 9.5 0.5 166
Higher 24.0 11.8 72.4 23.3 8.1 6.2 2.5 32.3 7.4 20.3 236 91.6 4.1 3.9 0.4 173
Highest 23.1 10.6 80.5 24.0 10.1 8.3 2.6 24.1 5.4 25.5 261 90.8 7.3 1.1 0.9 206
page- 325

Table 7-33. Injury and type of accident
page-326
Percentage of household with at least 1 member injured in an accident in the past 12 months, by type of accident and background character-
istics
Type of accident
Total
fall
bite
an accident
Background
Fire/
tree/
Road
cattle
characteristic
traffic
Other
Percentage of
assault
Animal
burning
building
households with
accident
Fall from
Violence/
Poisoning
Kicked by
at least 1 member
Accidental
injured or killed in
National 3.2 7.4 31.7 21.2 6.3 17.9 2.4 3.5 4.5 5.1 171

Livelihood
Agrarian 3.3 7.2 32.1 21.2 6.2 17.9 2.4 3.4 4.5 5.0 132
Pastoralist 0.8 18.7 1.4 19.7 15.4 19.0 1.5 5.6 5.1 13.6 39
Region
Tigray 3.5 3.0 37.3 0.0 12.6 47.1 0.0 0.0 0.0 0.0 21
Afar 2.9 10.5 0.0 20.1 25.7 6.2 4.6 17.2 15.7 0.0 26
Amhara 1.5 15.2 5.7 0.0 5.5 27.6 9.0 7.7 20.3 9.0 14
Oromia 3.4 6.3 23.2 37.6 9.4 12.4 2.5 5.0 3.7 0.0 35
Somali 0.7 0.0 3.6 15.5 17.6 29.3 0.0 0.0 0.0 34.0 9
Benishangul-
8.0 2.8 17.6 9.7 20.6 26.7 2.3 0.0 20.3 0.0 28
Gumuz
SNNPR 5.0 6.8 55.5 8.3 0.0 17.2 0.0 0.0 0.0 12.2 23
Gambela 1.7 1.9 67.3 6.4 9.0 0.0 15.5 0.0 0.0 0.0 8
Harari 1.5 18.0 26.2 30.8 0.0 25.0 0.0 0.0 0.0 0.0 7
Wealth quintile
Lowest 5.8 0.0 47.5 27.7 0.1 13.5 0.0 0.0 4.1 7.0 30
Lower 2.8 3.5 32.4 14.3 12.5 29.8 7.2 0.2 0.1 0.0 34
Middle 3.5 5.1 40.3 22.3 3.4 16.4 5.0 0.0 0.2 7.3 42
Higher 1.1 1.3 16.6 31.1 34.5 15.5 0.2 0.1 0.7 0.0 24
Highest 3.3 26.3 4.4 12.0 2.6 17.3 0.1 16.3 15.0 6.1 41

HEALTH-SEEKING BEHAVIOR OF HOUSEHOLD MEMBERS AND

REFERRALS FROM HPS
Of the family members who reported having been sick, treatment was sought
for 69.4%, with a slight difference between agrarian (69.6%) and pastoralist
(62.1%) communities. Only half of the sick people in Somali (50.6%) sought
treatment from health facilities. On the other hand, the vast majority of sick
people in Benishangul-Gumuz (89.5%) and Gambela (81.0%) sought treatment.
Health-seeking behavior showed an increase with wealth (Figure 7-5).
Figure 7-5: Household occurrence of illness and health seeking, by wealth index
The assessment consistently showed that HPs were not the first choice of
treatment for illnesses experienced by household members. Only 1 out of 5
(10.8%) sick individuals received services from the HP. Among those who
sought treatment from health facilities, nearly half (48.3%) received treatment
from HCs, while 22.4% received care from government hospitals. There was
a significant difference in the type of health facility from which households
in agrarian and pastoralist areas sought care. A majority of patients in Afar
(78.7%) and Gambela (54.6%) sought treatment from HPs, while patients in
Amhara (4.6%), Harari (8.7%), Oromia (10.4%), and Tigray (10.6%) received
services from HPs much less often. Pharmacies served as sources of treatment
for 6.5% of patients in Somali (6.5%). A great majority of the patients (91.1%)
received drugs from the health facilities, and very few received advice (4.2%),
procedures (4.0%) or other (0.8%) services (Table 7-34).
page- 327
Among the individuals who received services from health facilities, only 6.5%
reported that they were referred to the higher-level facility from the HP or
an HEW (6.5% in agrarian and 9.5% in pastoralist communities). A relatively
higher percentage of sick individuals from Gambela (23.1%) and Benishangul-
Gumuz (24.6%) received referral services from HPs. Respondents were asked
about their reasons for bypassing the HP without getting a referral. The
majority believed that the specific service they needed was unavailable at the
HP (34.0%) and that HEWs did not have the capacity to provide the specific
service they needed (33.4%). In addition, a notable percentage of respondents
identified their reasons for bypassing the HP as the availability of a hospital
or HC in the vicinity (13.6%), absence of HEWs’ (7.8%), and the closure (at the
moment) of the HP (6.7%; Table 7-34).
page-328
Table 7-34. Place of treatment and referral from health facilities among household members
Among household members with illness and treatment sought from health facilities, place of treatment, ability to get a referral from
Health Extension Workers, and reason for not getting a referral in the past 12 months, by background characteristics
Place of Treatment Reasons for not having referral from Health post /
HEWs
Background
Characteristic
HEWs
Other
Other
HEWs
Private
anyway
close by
without referral
Hospital
Hospital
Total number of
available
available
from health post/

Hospital is
Percentage of sick
Service not
no capacity
unavailable
Health post
HEWs have
Government
Private clinic
patients bypass HPs
Health post /
was closed or
who received referral

HEW was not
to provide the
Health Center
Health center/
as HEWs refer
specific service
No need to go,
members of the family

National 10.8 48.3 22.4 2.3 14.8 1.6 6.5 6.7 7.8 34.0 33.4 4.3 13.6 0.2 822
Livelihood
Agrarian 10.3 48.7 22.5 2.3 14.9 1.5 6.5 6.6 7.6 34.0 33.6 4.3 13.7 0.2 660
Pastoralist 29.2 34.2 18.7 3.2 9.4 5.3 9.5 10.1 18.0 35.9 22.1 3.1 10.3 0.4 162
Region
Tigray 10.6 73.2 9.1 5.3 0.4 1.3 7.6 27.6 0.0 31.8 20.3 3.2 15.5 1.7 85
Afar 78.7 16.7 4.6 0.0 0.0 0.0 4.3 11.0 17.0 0.0 6.3 65.7 0.0 0.0 24
Amhara 4.6 69.9 6.7 2.6 14.2 2.0 11.7 1.8 1.4 49.9 26.7 6.8 12.8 0.6 100
Oromia 10.4 39.6 32.0 1.5 16.3 0.2 4.0 6.7 11.9 21.2 41.9 2.6 15.7 0.0 168
Somali 34.8 39.4 3.0 6.2 6.9 9.72* 11.0 12.3 19.1 57.7 10.9 0.0 0.0 0.0 71
Benishangul- 33.6 49.8 3.8 0.0 7.2 5.5 24.6 9.0 8.8 43.8 18.0 15.6 2.2 2.5 107
Gumuz
SNNPR 15.7 43.0 16.1 3.3 16.8 5.2 7.3 4.5 3.0 63.4 15.0 7.6 6.5 0.1 103
Gambela 54.6 38.2 7.2 0.0 0.0 0.0 23.1 22.9 9.2 27.4 35.3 1.2 4.0 0.0 120
Harari 8.7 27.5 47.8 8.4 7.6 0.0 1.6 18.4 24.2 10.2 23.0 12.8 6.8 4.7 44
Wealth quintile
Lowest 7.7 67.0 9.3 0.0 14.4 1.6 5.4 3.1 3.3 51.2 31.7 5.0 5.8 0.1 143
Lower 8.6 65.6 12.1 0.1 11.4 2.2 9.5 3.7 8.1 38.4 26.2 7.3 16.3 0.0 134
Middle 22.7 48.4 9.9 2.5 16.0 0.6 10.6 7.4 9.5 36.6 16.8 3.9 24.9 0.9 166
Higher 5.4 36.1 37.6 1.9 16.9 2.1 5.8 6.2 9.8 31.2 42.6 4.3 5.6 0.3 173
Highest 11.2 46.8 23.0 4.3 13.4 1.1 4.1 9.4 6.2 28.1 35.5 2.9 18.0 0.0 206
* 6.5% of patients in Somali region received treatment from a pharmacy.
Abbreviations: HEW, Health Extension Worker; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page- 329

7.2.5 Non-communicable diseases: prevention and control
The burgeoning burden of NCDs in developing countries like Ethiopia has long
posed a multiple burden for the already constrained health system, along with
communicable diseases, maternal and perinatal conditions, and nutritional
problems. The burden of mental health has also increased over time.1 Risk
reduction through the creation of awareness and early detection and treatment
efforts are integral parts of the comprehensive NCD prevention and control
strategy developed by the GoE.6 The prevention and control of NCDs and
mental health were recently added to the HEP packages. One of the HEWs’
main tasks has been to raise awareness of the major NCDs and mental
health in the community and help the early detection and referral of patients
for treatment. This section presents the findings about the public awareness
of NCDs, the occurrence of NCDs and screening or diagnosis practices. In
addition, suicidality among women is presented to verify the relevance of
mental health services at HPs.
AWARENESS OF NCD AND COVERAGE OF SERVICES
Overall, about 5% of women (5% in agrarian vs. 10% in pastoralist areas)

and 3% of men (3% in agrarian vs. 4% in pastoralist areas) reported having
received information about all major NCDs, like cervical cancer, diabetes
mellitus, hypertension, heart disease, breast cancer, mental illness, and kidney
disease in the last 1 year from HEWs. More people residing in pastoralist areas
had heard about these NCDs from HEWs than had those in agrarian areas.
Among the NCDs, hypertension is the disease most commonly mentioned

by HEWs (as identified by 16% of women and 19% of men), while mental
illness is the least commonly mentioned disease across all population groups,
livelihoods, and regional classifications (Table 7-35, Table 7-36).
HEWs are responsible for referring women for cervical cancer screening. Less
than 1% of women had been screened for cervical cancer in the last 1 year, of
whom 27% were referred by an HP. Only 5.5% of women aged 30 to 49 had
been screened for the disease in the last 1 year (Table 7-35).
page-330
Table 7-35. Information received from HEWs on NCDs and cervical cancer
screening among women
Percentage of women who received information about non-communicable diseases, by type of disease
and background characteristics
Percentage of women who had heard
cervical cancer and got

who were screened for
Percentage of women
Percentage of women
from HEWs in the last 1 year about:
referral from health

ever screened for
cervical cancer
characteristics
Total number
All selected NCDs
Background
Diabetes mellitus
Cervical cancer
Heart disease
Mental illness
Breast cancer
Hypertension
post
Disease
Kidney
National 11.5 14.4 16.4 10.5 11.8 9.1 11.9 5.1 0.8 27.0 6 430
Age
15-19 6.9 10.3 13.4 6.8 9.3 10.0 6.9 3.6 0.0 0.1 281
21-24 11.5 12.5 14.0 7.0 12.5 8.6 13.3 4.5 0.4 2.5 763
25-29 14.2 16.6 19.4 10.0 12.5 8.4 11.9 5.1 0.7 20.6 1 209
31-34 11.2 15.0 17.7 10.9 12.8 10.3 10.0 5.7 0.3 10.1 943
35-39 14.1 16.5 18.8 12.7 14.5 10.8 14.7 6.9 1.0 20.7 923
41-44 11.1 15.4 16.9 12.7 13.0 9.2 11.6 5.1 1.4 1.0 519
45-49 13.8 23.1 26.1 18.0 17.9 13.2 21.6 7.2 2.9 2.4 353
50+ 8.1 10.0 10.9 8.6 7.2 6.8 8.6 3.5 0.5 4.3 1 439
Livelihood
Agrarian 11.4 14.1 16.0 10.1 11.6 9.7 11.4 4.9 0.8 26.5 4 421
Pastoralist 14.4 21.5 25.0 18.8 16.8 5.7 21.0 9.5 0.5 44.1 2 009
Region
Tigray 11.8 17.6 17.0 12.8 12.7 9.7 13.6 5.8 0.8 32.2 607
Afar 8.0 12.9 22.8 13.0 8.2 5.7 17.1 3.4 0.5 26.9 399
Amhara 14.0 18.2 21.9 16.8 16.3 16.0 17.0 10.0 1.2 2.5 1 060
Oromia 11.0 10.5 11.8 5.6 9.1 5.0 7.9 2.8 0.4 18.8 1 319
Somali 24.1 37.3 41.0 33.3 28.4 30.9 36.5 17.6 0.0 798
Benishangul-
7.3 9.6 13.3 7.5 8.7 4.1 9.7 2.7 2.4 31.8 406
Gumuz
SNNPR 8.1 15.2 16.7 10.0 10.1 6.8 10.9 2.4 1.1 68.7 1 009
Gambela 16.8 18.6 25.8 16.9 20.1 16.0 21.2 10.6 0.0 0.0 417
Harari 3.2 7.8 9.1 7.7 5.1 3.6 9.5 0.3 1.0 0.0 415
page- 331
Education
None 10.5 13.6 15.0 10.2 10.5 8.7 11.1 5.2 0.7 18.8 4 089
Grade 1-4 11.8 12.9 15.6 10.1 12.6 7.6 10.6 3.7 0.7 0.3 735
Grade 5-8 14.8 21.6 24.4 13.9 18.2 13.2 16.8 6.4 1.3 86.9 604
Grade 9+ 19.0 15.6 23.3 8.7 15.7 10.8 16.3 5.8 1.2 0.0 282
Marital status
Currently
11.5 14.3 16.1 10.0 11.8 8.6 11.6 4.9 0.7 33.2 5 162
married
Divorced 13.0 17.0 23.0 16.1 15.9 14.0 13.8 7.2 1.8 0.5 297
Widowed 11.4 15.2 17.0 12.8 11.3 11.3 12.9 6.2 1.0 2.7 721
Separated 8.5 11.2 11.1 9.0 8.6 7.3 10.1 6.1 0.1 100.0 201
Never
30.4 12.6 34.3 8.4 10.8 29.6 29.6 7.2 0.0 49
married
Wealth quintile
Lowest 7.8 11.7 14.5 10.8 9.2 8.8 11.1 5.2 0.9 4.0 1343
Lower 10.9 14.8 15.8 10.0 10.4 9.1 10.2 5.1 0.9 0.2 1291
Middle 11.2 14.0 16.5 10.2 12.9 9.5 11.2 5.3 0.7 68.5 1278
Higher 10.5 14.3 14.4 9.2 9.9 7.9 11.0 4.1 0.7 63.8 1274
Highest 16.6 16.8 20.8 12.5 16.4 10.1 15.7 6.1 0.8 0.5 1244
Abbreviations: NCD, non-communicable disease; SNNPR, Southern Nations, Nationalities, and

Peoples Region.
page-332
Table 7-36. Information received from HEWs on NCDs among men

Percentage of households who have heard from HEWs in the last 1 year
about:
Hypertension
All selected
Background Total
Diabetes
Cervical
mellitus
disease
disease
Mental
Kidney
characteristics number
cancer
cancer
Breast
NCDs
illness
Heart
National 12.7 14.3 18.6 13.5 14.0 11.7 15.1 3.1 4 416
Age
15-19 0.1 0.1 39.3 39.1 10.3 10.3 10.3 0.0 23
21-24 6.3 4.1 12.4 3.0 5.0 6.0 6.6 1.1 211
25-29 9.1 11.8 17.7 9.7 11.3 7.1 13.9 1.0 565
31-34 14.5 15.6 19.1 13.5 15.0 11.0 16.4 4.4 660
35-39 13.5 14.9 17.0 16.0 14.7 11.4 12.8 5.8 635
41-44 15.2 16.4 22.0 13.4 17.0 13.2 17.9 4.0 619
45-49 15.8 19.1 22.9 18.2 17.9 15.5 22.5 6.0 417
50+ 11.3 12.9 16.9 13.1 12.4 12.4 13.1 3.5 1 286
Livelihood
Agrarian 12.7 14.1 18.5 13.4 13.9 11.5 14.9 3.0 3 157
Pastoralist 13.3 21.5 23.1 18.5 16.1 16.8 22.0 4.3 1 259
Region
Tigray 13.4 17.6 18.5 16.3 15.5 14.6 19.8 2.3 407
Afar 3.9 12.9 28.2 17.3 7.1 10.8 26.0 1.0 275
Amhara 19.8 18.2 29.8 26.0 25.6 25.0 27.9 7.0 603
Oromia 11.8 10.5 12.0 7.0 10.6 6.7 9.2 1.7 1 139
Somali 24.9 37.3 47.5 38.8 33.8 37.9 46.2 8.2 376
Benishangul- 6.7 9.6 10.4 8.6 10.3 6.2 8.4 0.4 340
Gumuz
SNNPR 7.7 15.2 22.2 15.5 10.2 9.3 14.7 1.0 759
Gambela 8.6 18.6 14.9 11.1 9.6 8.5 11.2 3.1 157
Harari 4.6 7.8 7.7 7.2 5.5 5.8 6.9 0.3 360
Education
None 12.1 14.6 17.5 12.5 13.2 10.4 15.1 4.5 2 478
Grade 1-4 8.6 12.1 15.0 12.1 11.6 10.4 11.7 3.7 679
Grade 5-8 14.2 12.7 19.1 13.8 14.8 12.8 15.6 2.2 781
Grade 9+ 18.3 19.4 27.8 19.1 19.1 16.7 19.6 5.2 478
Wealth quintile
Lowest 7.9 11.7 14.4 10.2 8.1 8.2 13.1 2.2 708
Lower 9.0 14.8 14.6 10.6 9.9 10.1 12.3 2.3 843
Middle 14.3 14.0 20.9 15.6 16.3 13.2 17.9 3.6 881
Higher 11.7 14.3 19.7 13.4 15.1 11.7 13.9 2.7 1 028
Highest 18.3 16.8 21.0 16.1 17.4 13.6 17.3 4.3 956
Abbreviations: NCD, non-communicable disease; SNNPR, Southern Nations, Nationalities, and Peoples Region.
page- 333
In addition to the awareness level, information sources, and services received

with respect to NCDs, the number of diagnosed household members was
also assessed. The assessment showed that 9% of households had at least 1
member diagnosed with an NCD. Hypertension (4.5%) was the most frequently
diagnosed disease, while breast cancer (0.6%) was the least commonly
diagnosed. The proportion of households with at least 1 member diagnosed
with an NCD was higher in agrarian than in pastoralist regions (9.3% vs. 7.8%,
respectively). Afar had the highest proportion among all regions of households
(17%) with at least 1 NCD-diagnosed member, while Tigray had the lowest
proportion (3%). The proportion of households having at least 1 member
diagnosed with an NCD increases with wealth quintile (Table 7-37).
Overall, the largest group of patients (45%) diagnosed with an NCD were
diagnosed at HCs. Government hospitals, however, were the most common
places for NCD diagnosis in regions like Tigray (63%), Oromia (49%), Somalia
(35%), and Harari (49%). The main reasons for bypassing an HP for an NCD
diagnosis were service unavailability (44%) and a lack of confidence in the
capacity of HEWs (29%; Table 7-38).
page-334
Table 7-37. Occurrence of any NCD in households
Percentage of households who had a family member diagnosed with specified non-
communicable disease (NCD), by type of disease and background characteristics
Percentage of households who had a family
member diagnosed for a non-communicable
disease in the last 1 year about: Percentage
of households
Kidney disease
Cervical cancer
Heart disease
Background having at least Total
Mental illness
Breast cancer
Hypertension
Characteristics 1 member number
Diabetes
mellitus
diagnosed
with NCD
National 0.8 1.2 4.5 1.6 0.6 1.3 3.4 9.2 6 430
Livelihood
Agrarian 0.8 1.2 4.6 1.6 0.6 1.4 3.3 9.3 4 421
Pastoralist 0.4 0.9 2.6 2.8 0.2 0.4 4.9 7.8 2 009
Region
Tigray 0.0 0.2 0.3 1.5 0.0 0.9 1.2 3.3 607
Afar 0.1 4.0 8.4 4.1 0.4 0.4 12.7 17.4 399
Amhara 1.5 2.5 9.5 2.6 1.2 1.1 3.5 13.2 1 060
Oromia 0.5 0.6 1.9 1.0 0.0 1.3 3.1 7.2 1 319
Somali 0.0 0.7 1.7 4.6 0.2 0.5 7.3 9.5 798
Benishangul-
0.3 0.6 7.4 0.7 0.2 1.4 8.5 16.0 406
Gumuz
SNNPR 1.0 0.8 5.1 1.4 1.3 1.9 3.8 9.8 1 009
Gambela 0.6 2.2 7.0 0.0 2.2 0.5 4.2 9.5 417
Harari 0.6 0.8 8.0 2.0 0.4 0.4 5.9 13.7 415
Wealth quintile
Lowest 1.2 1.0 3.9 1.4 0.5 0.8 2.5 7.7 1 343
Lower 0.7 0.3 2.7 0.9 0.4 0.7 0.6 5.1 1 291
Middle 0.7 1.1 4.6 0.9 0.4 0.5 1.8 6.9 1 278
Higher 0.7 1.0 3.7 1.6 1.1 2.2 4.9 10.2 1 274
Highest 0.9 2.2 7.7 3.2 0.6 2.2 6.5 15.7 1 244
Abbreviations: NCD, non-communicable disease; SNNPR, Southern Nations, Nationalities, and

Peoples Region.
page- 335
Table 7-38. Place of diagnosis for NCD and reason for bypassing health posts
page-336
Place of diagnosis for an identified non-communicable disease (NCD), by background characteristics
Place diagnosis for an
identified non-communicable Reason for bypassing the health post without referral
disease in the last 1 year,
Background Total
Characteristic number
NCDs
from HP for
patients who
HC/
clinic
other
Percentage of
Other
center
service
Health
Private
private
received referral
anyway
hospital
close by
Hospital
to go, as
No need
available
HEW was
to provide
Hospital is
was closed
the specific
Service not
no capacity
Health post
HEWs refer
Unavailable
HEWs have
Government

National 45.0 33.1 5.9 14.8 1.2 11.2 6.3 3.3 43.9 28.7 6.8 10.5 0.6 668
Livelihood
Agrarian 44.9 33.2 5.7 15.1 1.1 11.1 5.8 2.5 44.4 29.1 6.9 10.7 0.6 464
Pastoralist 45.9 30.4 11.7 8.0 4.0 15.8 19.5 23.0 30.8 17.0 5.1 3.9 0.7 204
Region
Tigray 11.6 62.6 25.8 0.0 0.0 18.6 0.0 7.6 16.5 48.5 5.5 22.0 0.0 18
Afar 73.9 20.1 0.0 6.0 0.0 25.4 10.9 31.0 19.2 9.1 16.5 13.3 0.0 80
Amhara 68.4 18.6 4.8 8.0 0.2 15.2 0.3 2.4 56.1 18.2 9.7 13.4 0.0 134
Oromia 28.7 49.5 3.3 17.6 0.9 2.8 7.4 2.4 24.5 46.6 6.8 12.3 0.1 98
Somali 28.0 38.2 19.6 10.5 3.7 3.3 21.5 21.6 35.8 20.6 0.5 0.0 0.0 78
Benishangul-
60.4 28.4 5.5 5.1 0.8 35.1 16.1 0.0 31.0 32.4 12.3 0.0 8.2 63
Gumuz
SNNPR 32.6 29.8 9.4 24.9 3.3 17.1 13.6 3.1 63.4 13.4 1.9 2.2 2.6 99
Gambela 57.7 42.3 0.0 0.0 0.0 2.8 6.1 4.6 48.4 8.8 1.2 5.2 25.8 40
Harari 42.6 48.7 4.0 2.8 2.0 19.4 16.2 20.2 17.4 28.9 9.3 5.6 2.5 58
Wealth quintile
Lowest 60.2 24.1 3.3 12.4 0.0 15.8 8.0 2.8 49.0 26.6 1.9 11.7 0.1 106
Lower 60.4 26.8 4.5 8.1 0.2 16.1 0.7 6.6 52.9 11.7 21.1 7.0 0.0 104
Middle 57.8 26.9 2.1 11.4 1.8 20.8 6.2 5.3 45.9 23.5 5.5 10.8 2.8 125
Higher 34.2 41.0 10.0 14.6 0.1 10.1 9.0 2.0 42.7 32.3 5.3 8.6 0.1 132
Highest 36.1 35.8 6.2 19.5 2.5 4.4 5.3 2.7 39.8 33.4 6.2 12.2 0.5 201
Abbreviations: NCD, non-communicable disease; SNNPR, Southern Nations, Nationalities, and Peoples Region.
SUICIDE AMONG WOMEN
Suicidality among women, as a verification indicator for the burden of mental

health problems in rural communities, was assessed by asking 3 key questions
with respect to the last 12 months: whether they had thought of taking their
own lives, whether they planned to take their lives, and whether they had
attempted to end their lives. Following these 3 questions, they were also asked
whether they had sought medical help for their thoughts, plans, or attempts.
For those who did seek help, they were asked where they got medical help.
The overall prevalence of suicidal thoughts among women in the last 12 months
was 6.2%. This prevalence is highest among women aged 35 to 49 years
(8.0%), with no formal education (6.6%), in the lowest wealth quintile (10.5%),
in the SNNPR (15.9%), and among the agrarian population (6.4%). The
prevalence of making a plan to end one’s own life was 4.3%, and its relative
magnitude followed the same trend as having suicidal thoughts. Two out of
100 women had attempted suicide (2.1%) in the previous year, suggesting
its significance as a public health issue. The 1-year prevalence of a suicidal
attempt was highest in the age range of 35 to 49 (3.2%), with at least a 9th
grade education (2.7%), in agrarian areas (2.1%), in the SNNPR (6.0%), and
in the lowest wealth quintile (4.7%). The prevalence of seeking for suicidal
thoughts, plans, or attempts was low (15.3%; Table 7-39).
Those who looked for help sought it from health institutions (58.2%) family
members (8.1%), or HEWs (33.8%).
page- 337
Table 7-39. Suicidality among women in Ethiopia

Percentage Percentage Percentage of Total Percentage
of all women of all women all Total number
Background women who number of suicidal
who thought who planned of suicidal
characteristic attempted of who sought
about taking to take their women
suicide women help
their lives lives
National 6.2 4.3 2.1 6 430 15.3 284
Maternal age
15-19 1.1 0.1 0.0 281 0.0 4
21-24 4.6 3.5 1.4 763 0.0 29
25-34 6.1 3.8 1.7 2 152 8.7 94
35-49 8.0 6.1 3.2 1 795 20.4 97
50-59 5.2 3.7 1.7 903 11.8 42
60+ 6.0 2.6 2.2 536 40.1 18
Maternal education
No formal 6.6 4.5 2.2 4 809 16.0 208
education
Grade 1-4 5.5 3.4 1.8 735 9.6 46
Grade 5-8 5.4 4.2 1.9 604 16.2 23
Grade 9+ 4.8 3.3 2.7 282 18.3 7
Livelihood
Agrarian 6.4 4.4 2.1 4 421 15.1 214
Pastoralist 3.4 2.7 1.3 2 009 21.3 70
Region
Tigray 2.7 1.5 0.6 607 0.0 14
Afar 9.0 9.0 8.8 399 76.4 21
Amhara 3.4 2.0 0.9 1 060 19.6 31
Oromia 4.3 3.1 1.3 1 319 8.9 52
Somali 0.1 0.0 0.0 798 0.0 1
Benishangul- 5.8 3.7 1.0 406 13.5 17
Gumuz
SNNPR 15.9 11.1 6.0 1 009 17.9 125
Gambela 3.1 1.9 0.2 417 0.0 11
Harari 2.8 1.0 0.3 415 14.8 12
Wealth quintile
Lowest 10.5 6.7 4.7 1 343 22.3 68
Lower 7.7 5.6 1.9 1 291 5.1 56
Middle 5.4 3.3 1.8 1 278 9.6 55
Higher 4.9 4.1 1.9 1 274 19.8 55
Highest 3.5 2.3 0.7 1 244 20.3 50
page-338
7.3 Coverage of Family Health Services

One of the principal components of the HEP packages is the family health
service. This study assessed the coverage of FP, ANC, delivery, PNC, vaccination,
child health, and child nutrition.
Summary of findings
Overall, maternal and child health services are improving compared with the
2016 EDHS. This can be described by key indicators, including the contraceptive
prevalence rate (44.6%), percentage of pregnant mothers having at least 4
antenatal visits (ANC 4+; 48.3%), institutional delivery (54.9%), and PNC
(25.5%) among women. Child health services, like the full basic vaccination
coverage 35.7%, children under 5 who received treatment for acute respiratory
infection (ARI; 40.2%), diarrhea (53.3%), or fever (46.8%), were exclusively
breastfed (58.7%), and got a minimum acceptable diet (14.4%).
7.3.1 Family Planning
KNOWLEDGE OF CONTRACEPTIVE METHODS
At least 1 FP method is commonly known by women (96.5%), men (96.6%), and

youth girls (92.8%). The most commonly known methods across all respondents
are injectable contraception, implants, and pills. A higher portion of youth girls
had knowledge of male condoms (75.6%) and female condoms (28.3%) than
did the women and men respondents. Female sterilization was more known
by all respondents than male sterilization, with 37.1% of women, 31.9% of men,
and 31.8% of youth girls knowing about it, as compared to 13.5%, 15.5%, and
15.7%, respectively, who knew about female condoms (Figure 7-6).
page- 339
Figure 7-6. Method-based FP knowledge of women, men and youth girls

Abbreviations: steri., sterilization; IUD, intra-uterine device; FP, family planning; SDM, Standard Days
Method; LAM, Lactational Amenorrhea Method.
The mean number of methods known by all women, men, and youth girls were
5.7 (SD±2.9), 5.7 (SD±3.1), and 5.7 (SD±3.3), respectively.
page-340
DEMAND FOR, USE OF, AND UNMET NEED FOR CONTRACEPTIVE

METHODS
Definitions used to calculate unmet need

and demand for FP
Unmet need for family planning (FP): the proportion of women who (a) are
not pregnant and not postpartum amenorrhea, are considered fecund, and
want to postpone their next birth for 2 or more years or stop childbearing
altogether but are not using a contraceptive method, (b) have a mistimed or
unwanted current pregnancy, or (c) are postpartum amenorrhea, and their
last birth in the last 2 years was mistimed or unwanted (EDHS, 2016; DHS
Guide to Statistics, 2018).
Infecund women are excluded from the category of unmet need for FP. Women
are classified as infecund if they fall into any of the following categories
(EDHS, 2016; DHS Guide to Statistics, 2018):
1. Married 5+ years, have had no children in the past 5 years, and have
never used contraception;
2. Responded “menopausal/hysterectomy,” infertility, or old age for their
reason for not using contraception; or
3. Have had an absence of menses, do not use contraception, and their
last delivery was ≥25 months.
Demand for FP: the proportion of women with an unmet need for FP and
current contraceptive use (any method; EDHS, 2016; DHS Guide to Statistics,
2018).
Demand satisfied for FP: the proportion of women with needs met from the
total demand for FP.*
*EDHS, 2016; DHS Guide to Statistics, 2018.
page- 341
The total demand for contraceptive methods for women in the reproductive
age group was 61%; 60.5% of married and 62.8% of unmarried women.
Similarly, 61.7% from agrarian areas, 64% living in the SNNPR, 63.4% in the
highest wealth quintile, and those aged 35 to 39 had their demands met
for FP. Sixty-three percent of women in the reproductive age group had a
satisfied FP demand. Satisfied FP demand varies across region, ranging from
0% in Somali to 75.5% in the SNNPR. The unmet need for FP is very high
(22.5%) compared with the national target for 2020 (10%). The highest unmet
need was observed among unmarried women (50%), among women in Somali
(34.5%), and among women in the lowest wealth quintile (27.6%). Twenty-two
percent of women in agrarian areas and 29.1% in pastoralist areas had an
unmet need for any FP method (Table 7-40, Table 7-41).
The CPR is 44.6% for all women in the reproductive age group. Modern
contraceptive use is the most common method of FP; it accounts for 43.3% of
women and 35.7% of youth girls who had initiated sexual intercourse. Women
in agrarian setting use 3 times more contraception than those in pastoralist
settings. Women aged 45 to 49 are the highest non-users of FP methods.
The highest users of traditional FP methods (6.2%) are women who do not
have children. There is no variation by wealth quintile. There is high regional
variation in the use of FP, ranging from 0% in Somali to 55% in the SNNPR.
By contrast, the lowest percentage of users was observed in Afar (9.4%) and
Somali (0%; Table 7-40).
Table 7-40. Contraceptive prevalence rate among women and youth girls
Women Youth girls
Any Any Unweighted Any Any Unweighted
Background
modern traditional Total modern traditional total
characteristic
methods method methods method
National 43.3 1.3 4 088 35.7 0 199
Livelihood
Agrarian 44.5 1.4 2 813 36 0 172
Pastoralist 13.7 0.1 1 275 8.5 0 27
Region
Tigray 35.4 0.7 376 35.4 0 38
Afar 9.4 0 264 0 0 11
Amhara 51.4 0.4 621 45.9 0 46
page-342
Oromia 36.1 2.4 871 32.4 0 14

Somali 0 0 452 0 0 9
Benishangul-
47.7 1.2 275 46.6 3.2 19
Gumuz
SNNPR 55 0.3 697 15.9 0 19
Gambela 19.4 0.2 243 16.2 0 22
Harari 30.3 4.3 289 23.4 0 21
Age
15-19 41 0.7 127 40.3 0.1 106
20-24 53.3 1.4 548 33.1 0 93
25-29 49.9 0.9 979 N/A N/A N/A
30-34 48.3 1.8 809 N/A N/A N/A
35-39 43.7 2 834 N/A N/A N/A
40-44 30.6 0.3 470 N/A N/A N/A
45-49 15.4 0.7 321 N/A N/A N/A
Education
No formal
38.7 1 2 860 39.6 0 38
education
Grade 1-4 46.9 2.3 550 17.2 0 33
Grade 5-8 57.9 1.3 458 51.4 0.1 81
Grade 9+ 53.4 1.3 220 20.2 0 47
Marital Status
Currently
46.1 1.4 3553 52.2 0 71
married
Unmarried 14.6 0 535 28.4 0 128
Wealth quintile
Lowest 40.1 0 763 38 0 41
Lower 42 0.2 796 27.9 0 39
Middle 43.1 0.3 815 27.5 0 49
Higher 41.2 2.4 903 51.6 0.1 30
Highest 49.3 3.1 851 40 0 40
No of Children
No children 17.7 6.2 117 N/A N/A N/A
1-2 Children 49.3 0.8 1 069 N/A N/A N/A
3-4 Children 48.2 1.3 1 311 N/A N/A N/A
5-7 Children 38.3 1.7 1 303 N/A N/A N/A
8+ Children 27.8 0.3 288 N/A N/A N/A
page- 343
Table 7-41. Percentage of need and demand for FP among reproductive-age

(15-49) women
Demand
Unmet need for Family Planning Satisfied FP
Satisfied with
family planning Demand Demand
Modern FP
Unweighted
Unweighted
Unweighted
Unweighted
Background
Total
Total
Total
Total
%
%
National 22.5 4 991 61.0 4 991 63.1 4 991 61.2 2 643
Livelihood
Agrarian 22.2 3 346 62.0 3 346 64.2 1 991 62.3 1 991
Pastoralist 29.1 1 645 39.6 1 645 26.6 652 26.4 652
Region
Tigray 27.5 410 60.2 410 54.4 243 53.4 243
Afar 9.7 334 17.1 334 43.0 72 43.0 72
Amhara 17.7 729 61.6 729 71.2 458 70.7 458
Oromia 27.3 1 046 60.4 1 046 54.9 615 51.4 615
Somali 34.5 643 34.5 643 0.00 227 0.00 227
Benishangul-
67.1
Gumuz 18.7 325 60.3 325 68.9 193 193
SNNPR 15.9 788 65.1 788 75.5 487 75.2 487
Gambela 32.2 356 46.0 356 29.9 159 29.5 159
Harari 23.9 360 51.8 360 53.8 189 47.1 189
Age
15-19 16.4 281 36.0 281 54.5 89 53.5 89
20-24 18.3 763 59.1 763 69.1 406 67.3 406
25-29 23.4 1 209 66.5 1 209 64.8 672 63.6 672
30-34 26.1 943 70.8 943 63.2 571 61.0 571
35-39 24.1 923 66.6 923 63.8 552 62.0 552
40-44 26.1 519 54.8 519 52.4 246 51.8 246
45-49 12.5 353 27.6 353 54.6 107 52.2 107
No of children
No children 5.3 634 10.0 634 47.3 61 35.0 61
1-2 children 18.5 1 213 62.3 1 213 70.3 718 69.1 718
3-4 children 23.6 1 448 69.3 1 448 65.9 851 64.2 851
5-7 children 25.2 1 391 62.7 1 391 59.8 823 57.3 823
+8 children 41.0 305 67.8 305 39.5 190 39.1 190
page-344
Marital Status
Currently
20.0 4 334 60.8 4 334 67.1 2283 65.1 2283
Married
Unmarried 50.0 657 63.0 657 20.6 360 20.6 360
Wealth quintile
Lowest 27.6 933 62.7 933 56.0 486 55.9 486
Lower 25.8 999 62.1 999 58.4 486 58.2 486
Middle 21.1 996 58.4 996 63.8 498 63.4 498
Higher 21.4 1 037 59.1 1 037 63.7 571 60.2 571
Highest 18.4 1 026 63.4 1 026 71.0 602 66.7 602
Abbreviations: FP, family planning; SNNPR, Southern Nations, Nationalities,

and Peoples Region.
Among the women in the reproductive age group, the most common modern
contraceptive method used was short-acting (31.4%), followed by long-
acting (11.2%), permanent (0.7%), and traditional methods (1.3%). Of the
contraceptive methods used, short-acting contraceptives represented 70%, and
long-acting represented 25%. All of the youth girls assessed used exclusively
modern contraceptive methods, of which short-acting represented 75% of the
total contraceptives used (Figure 7-7).
Women Youth girls
Figure 7-7. Share of long-acting and short-acting contraceptive methods

among current contraceptive users
page- 345
REASONS FOR NOT USING CONTRACEPTION AMONG MARRIED

WOMEN
Among married women who did not want to get pregnant but were not using
any contraceptives during the survey, 20.6% stated that breastfeeding was
their reason for not using contraceptives, 16.4% stated their reason was that
using contraception was not allowed, and 8.8% mentioned their fear of side-
effects.
Figure 7-8. Reasons for not using any contraceptive methods among married
women who do not want to get pregnant soon
7.3.2 Antenatal care services
The results of this survey show that 85.7% of women had at least 1 ANC follow-
up during their most recent pregnancy. Forty-eight percent had 4 or more
ANC visits. There is huge variation in the use of ANC by livelihood. ANC-1 was
used by 87.4% of women in agrarian settings, compared to 48.9% in pastoralist
settings. ANC-4 was used by 49.6% of women in agrarian settings, compared
to 20.3% in pastoralist settings. Having at least 1 ANC visit increases with
wealth quintile and educational status (Table 7-42). Around 44% of pregnant
women had their first ANC visit during their second trimester (4-5 months).
Around 2% of women had their first ANC visit when their pregnancy had
reached 8 or more months. Only 22.8% of mothers had their first ANC visit on
the recommended schedule, during their first trimester (Table 7-43).
page-346
Table 7-42. Coverage of ANC visits
Background ANC visit

Unweighted Total
characteristics At least one At least 4
National 85.7 48.3 3031
Age % % Number of pregnant women
15-19 90.9 46.1 123
20-34 87.8 48.3 2 042
35-49 80.8 48.3 866
No live birth
≤2 92.7 53.2 911
3-4 83.7 46.5 942
5-7 80.0 48.9 925
≥8 86.9 34.0 251
Education
No formal education 80.1 43.0 2 054
Attended grade 1-4 92.3 56.7 423
attended grade 5-8 95.7 59.3 375
attended grade 9+ 98.3 50.1 179
Livelihood
Agrarian 87.4 49.6 2 018
Pastoralist 48.9 20.3 1 013
Region
Tigray 94.4 72.1 258
Afar 68.1 20.8 195
Amhara 87.0 57.1 397
Oromia 88.1 46.3 699
Somali 9.9 1.2 362
SNNPR 83.8 43.0 516
Gambela 52.8 30.1 165
Harari 68.5 25.5 237
Wealth quintile
Lowest 80.6 46.5 605
Lower 84.4 44.1 586
Middle 83.2 41.8 585
Higher 87.6 48.1 650
Highest 91.4 60.8 605
page- 347
Table 7-43. Timing of first ANC visit

Time for first ANC visit
Background
characteristics No ANC 4-5 6-7 ≥8 Do not
< 4 months
visit months months months remember
National 14.3 22.8 44.1 13.8 1.7 3.3
Age
15-19 9.1 41.0 28.6 17.6 0.0 4.1
20-34 12.2 21.9 44.5 15.1 2.0 4.3
35-49 19.2 23.3 44.2 10.8 1.3 1.0
Livelihood
Agrarian 12.6 23.4 45.0 14.0 1.7 3.3
Pastoralist 51.1 9.8 25.1 10.0 2.1 1.9
Wealth quintile
Lowest 19.4 25.7 34.8 16.3 2.7 1.1
Lower 15.6 20.0 46.5 13.0 1.3 3.5
Middle 16.8 16.9 42.8 17.5 4.2 1.8
Higher 12.4 25.3 48.9 9.1 0.5 3.8
Highest 8.6 26.4 44.7 14.4 0.1 5.8
Abbreviation: ANC, antenatal care.
REASON FOR LOW UPTAKE OF ANC SERVICE
The most common reason for the low uptake of ANC services was the long
distance to the HP (33.2%), followed by a lack of knowledge about ANC
(16.8%). Closed HPs and the absence of HEWs during the women’s visits for
ANC service were the other reasons offered. The long distance to the HP was
the most common reason in all settings (given by 24% of women in agrarian
settings and 32.1% in pastoralist settings). A lack of money was the second
most common reason in pastoralist settings (given by 22.5% of women; Table
7-44).
In Gambela, 50% of women gave the long distance from the HP as their reason
for not receiving ANC services during their last pregnancy. Thirty percent of
mothers in Somali identified a lack of money as a barrier to receiving ANC
services. The highest percentage of mothers in Oromia (29.5%) and Tigray
(29.4%) offered their lack of knowledge as a major reason for their low use of
ANC. Women in Amhara (26.4%) and Benishangul-Gumuz (17.3%) reported
closed HPs or the absence of HEWs at the time of their visit as a reason for
not having ANC visits (Table 7-44).
page-348
Table 7-44. Women’s reason for not having ANC visits for their most recent
birth
Reason for no ANC visit
Disapproval
by husband
accompany
HP closed/
knowledge
Background
desistance
No one to
No desire
HEW not
available
Lack of
Lack of
money
to HP
characteristics
Long
National 24.8 7.3 10.1 10.0 16.8 12.3 6.5

Livelihood
Agrarian 24.0 4.5 13.4 9.6 22.8 10.0 15.6
Pastoralist 32.1 22.5 5.9 11.1 14.9 2.3 11.2
Region
Tigray 23.3 0.0 0.0 16.9 29.4 16.4 14.0
Afar 21.3 11.1 16.9 38.1 1.8 1.1 9.8
Amhara 21.6 6.6 4.5 11.8 21.3 26.4 7.6
Oromia 20.4 6.5 18.7 14.8 29.5 0.1 9.9
Somali 33.0 30.1 2.2 7.45 14.4 2.1 10.7
Benishangul-
33.0 17.7 2.5 4.0 23.7 17.3 1.8
Gumuz
SNNPR 32.3 0.0 13.6 0.2 12.8 10.4 30.6
Gambela 50.3 4.8 20.5 14.3 2.1 0.0 7.9
Harari 30.7 7.0 3.4 8.8 21.9 5.6 22.7
Abbreviations: ANC, antenatal care; HP, health post; HEW, Health Extension Worker; SNNPR,
7.3.3 Delivery service
The place of delivery was assessed for each respondent’s youngest child
delivered in the last 5 years. Health facility delivery includes HPs, HCs, hospitals,
and clinics. The findings of this study show that in the last 5 years, 54.9% of
women delivered their last child at a health facility.
This study showed that 43.3% of women delivered their last child at home. In
Somali, the region with the highest rate of home delivery, more than 92% of
deliveries occurred at home, followed by Afar (78.7%) and Gambela (61.4%).
The lowest rate of home delivery was found in Tigray (16.9%), followed by
Amhara (27%; Figure 7-9, Table 7-45). Attendants of deliveries were health
professionals for 51.8% of the mothers, while HEWs attended 3.7%. Relatives
attended several deliveries (Table 7-46).
page- 349
Figure 7-9. Place of delivery for most recent birth during the last 5 years
UMBILICAL CORD CARE
Each participant with a delivery during the 5 years prior to the survey was
asked whether she had applied anything to her baby’s umbilical cord after
delivery; 57.7% of mothers stated that something was applied to their baby’s
umbilical cord. Among these babies, oil or butter was applied to 48.1% (Figure
7-10).
Figure 7-10. Treatments applied to the umbilical cord after delivery
page-350
Table 7-45. Place of delivery for the most recent birth in the last 5 years
Place of delivery
Background Other Unweighted
Characteristics Home HP HC Health Other places number of
Institution women
National 43.3 4.0 38.8 12.1 1.8 2 298
Livelihood
Agrarian 42.2 4.1 39.7 12.3 1.8 1 639
Pastoralist 69.9 1.9 18.0 9.3 0.8 659
Region
Tigray 16.9 3.8 59.4 17.0 2.9 283
Afar 78.7 4.6 7.6 9.1 0.0 71
Amhara 27.0 1.8 62.1 6.7 2.4 291
Oromia 51.5 5.4 27.8 14.2 1.2 662
Somali 92.0 0.1 1.4 5.8 0.7 122
Benishangul-
45.7 9.0 30.1 14.6 0.6 107
Gumuz
SNNPR 41.7 2.3 43.9 10.2 1.8 446
Gambela 61.4 2.8 28.5 7.4 0.0 42
Harari 55.9 0.1 23.6 19.8 0.5 274
Age
15-19 35.1 1.9 54.6 8.6 0.0 86
21-24 42.0 6.7 34.5 15.5 1.4 441
25-29 41.9 3.2 39.2 14.7 1.0 668
31-34 47.9 3.6 38.4 8.6 1.5 493
35-39 40.4 3.2 43.6 9.8 3.0 413
41-44 50.0 4.4 35.4 10.1 0.2 153
45-49 44.7 5.2 28.0 11.9 10.1 44
Wealth quintile
Lowest 47.5 4.6 40.9 5.3 1.6 372
Lower 43.2 8.9 38.0 8.5 1.3 420
Middle 45.6 2.7 40.6 9.5 1.4 455
Higher 44.6 3.4 33.5 16.2 2.3 548
Highest 36.9 1.6 43.0 17.1 1.4 503
Abbreviations: HC, health center; HP, health post; SNNPC, Southern Nations, Nationalities, and
Peoples Region.
page- 351
Table 7-46. Delivery attendance for the most recent live birth in the last 5 years
Delivery Attended by Unweighted
Background
Health No one number of
Characteristics HEWs TBA Relatives Other
professional attended women
National 3.7 51.8 5.5 37.1 0 .0 2.4 2298
Livelihood
Agrarian 3.2 52.8 4.9 36.6 0.0 2.4 1 639
Pastoralist 1.7 26.1 20.7 48.1 0.1 3.2 659
Region
Tigray 0.8 80.3 1.3 13.0 0.0 4.5 283
Afar 3.8 21.3 59.7 15.3 0.0 0.0 71
Amhara 1.8 69.1 7.8 18.8 0.0 2.6 291
Oromia 4.3 42.6 4.9 47.8 0.0 2.6 662
Somali 0.1 8.8 43.4 47.7 0.0 0.0 122
Benishangul-
2.4 51.5 3.4 34.0 0.0 8.9 107
Gumuz
SNNPR 2.1 54.3 3.9 31.1 0.0 8.5 446
Gambela 4.3 43.3 5.2 33.3 0.0 13.8 42
Harari 0.0 43.7 43.3 11.2 0.0 1.7 274
Age
15-19 1.9 62.9 16.3 18.8 0.0 0.0 86
21-24 2.9 52.5 9.3 33.4 0.0 2.1 441
25-29 2.1 54.9 4.7 36.7 0.0 1.5 668
31-34 4.2 47.3 2.9 41.5 0.0 4.0 493
35-39 3.6 53.8 2.6 37.5 0.0 2.5 413
41-44 4.4 45.4 11.3 38.5 0.0 3.4 153
45-49 3.9 45.2 8.1 42.6 0.0 0.2 44
Wealth quintile
Lowest 3.5 47.3 8.1 35.1 0.0 6.0 372
Lower 7.6 48.3 9.6 32.8 0.0 1.7 420
Middle 0.9 50.4 6.2 40.0 0.0 2.5 455
Higher 3.0 49.8 3.4 41.4 0.0 2.3 548
Highest 1.4 61.1 2.5 34.2 0.0 0.8 503
Abbreviations: HEW, Health Extension Worker; TBA, traditional birth attendant; SNNPC, Southern
page-352
7.3.4 Postnatal care services
A quarter (25.5%) of women who delivered during the 2 years prior to the
study had at least 1 PNC visit. Education level is associated with PNC follow-
up; women with no formal education received PNC at a rate of 21.4%, while
37.2% of those who had attended at least grade 9 did. In Somali, only 0.4% of
women had any PNC follow-up. Other regions with relatively low rates of any
PNC were Afar (12.3%), Oromia (15.2%), and Harari (16.1%). The highest rate
any PNC follow-up was found in Gambela (60.1%), followed by Benishangul-
Gumuz (48.9%) and the SNNPR (45.6%). Women in pastoralist areas had
lower rates of PNC follow-up (17.7%) than women in agrarian areas (25.9%;
Table 7-47)
CONTENT OF POSTNATAL CARE
Mothers who had had a PNC visit after their most recent delivery within the
past 2 years were asked about the content of the care they received. Half
(50.3%) of women received a physical examination during PNC visits, and
more than 40% were checked for heavy bleeding. Only 24.8% were counseled
on FP, and only 29% were counseled on nutrition (Figure 7-11).
Figure 7-11. Content of postnatal care for the mother
page- 353
For 42.4% of women, the service provider physically examined the babies.
More than 40% of women were counseled on breastfeeding. Only 18.9% and
14.5% of babies had their umbilical cords checked and were checked for signs
of illness, respectively (Figure 7-12).
Figure 7-12. Content of postnatal care for the newborn
Table 7-47. Postnatal visits for the youngest child in the last 2 years
Frequency of postnatal visits

Background Unweighted
characteristic No Total
1 Visit 2 Visits 3 Visits ≥4 Visits
Visit
National 74.5 11.8 8.7 2.3 2.7 1 545
Age
15-19 71 17.3 7.4 3.9 0.4 73
20-24 73.6 13.9 6.9 2.5 3 333
25-29 74.1 10 10.2 2 3.7 462
30-34 77.1 14.4 5.9 1.6 1 330
35-39 75.3 7.7 10.3 3.5 3.1 250
40-44 65.1 16.3 14.9 2.3 1,3 80
45-49 86.9 4.9 4.4 0 3.8 17
Education
No formal
78.6 8.9 8.6 1.7 2.1 1 027
education
Attended grade
70 17.1 4.9 3.8 4.2 221
1-4
Attended grade
69.1 14.4 10.1 3.1 3.3 206
5-8
page-354
Attended grade
62.8 17 15.8 1.6 2.8 91
9+
Marital Status
Currently
74.8 11.7 8.4 2.4 2.8 1453
married
Unmarried 65.9 14.7 16.8 1.9 0.7 92
Wealth quintile
Lowest 68.5 17.5 11.2 2.1 0.7 253
Lower 81.7 11.5 3.6 1.5 1.6 275
Middle 77.9 12.2 6.7 1.4 1.6 302
Higher 71.9 8.1 9.8 4.4 5.7 377
Highest 72.4 12.8 11.5 1.4 1.8 338
No of Children
1-2 Children 72.8 13.1 7.7 2.7 3.7 475
3-4 Children 73.1 12.9 9.8 1.6 2.5 496
5-7 Children 79.3 9.2 8.5 1 2.1 478
8+ Children 68.9 11.9 9.8 8.6 0.8 93
Region
Tigray 58.5 24.9 7.4 5.3 3.9 207
Afar 87.7 11.4 0 0 0.9 34
Amhara 59 19.5 18.8 1.7 1.1 167
Oromia 84.8 4.9 5.6 1.6 3.1 484
Somali 99.6 0.2 0 0.2 0 84
Benishangul-
51.1 10.2 23.8 12 2.7 80
Gumuz
SNNPR 54.4 27.3 11.4 4.7 2.1 278
Gambela 39.9 2.4 4.5 23.8 29.3 25
Harari 83.9 10.6 2.3 1.9 1.3 186
Livelihood
Agrarian 74.1 12 8.8 2.4 2.7 1 089
Pastoralist 82.3 8.1 6.9 1.5 1.1 456
page- 355
7.3.5 Childhood illness and treatment
The target of the assessment of childhood illness and its treatment was the
youngest child in the household below 5 years old. Mothers were asked whether
the child had had fever, symptoms of ARI, or diarrhea in the 2 weeks preceding
the survey. Those mothers confirming the presence of these conditions were
asked further questions about help-seeking for the child, the timeliness of help-
seeking (i.e., the same day or the next day), and the service provider.
PREVALENCE AND TREATMENT OF ARI
The 2-week prevalence of ARI among under-5 children was 17.9%. Children
whose families used fuel for cooking that caused indoor pollution had an
increased risk of showing symptoms. Across the regions, the prevalence of
ARI ranged from 8.3% in Afar to 23.9% in Tigray and 23.6% in Benishangul-
Gumuz. Among children with ARI symptoms, 40.2% received treatment, and
the prevalence of seeking help with a maximum delay of 1 day was 16.8%. Help-
seeking increases with maternal education and is relatively higher in agrarian
areas. Help-seeking and its timeliness was lowest in Somali and highest in
Gambela, followed by Benishangul-Gumuz (Table 7-48).
PREVALENCE AND TREATMENT OF FEVER
The prevalence of fever among the population of youngest children under

5 was 19.6% and was higher in the agrarian population than the pastoralist
population. Across the regions, it ranged from 10.6% in Afar to 39.7% in
Gambela.
Among children who had diarrhea, 46.8% received medical help, and the
timeliness of help-seeking (either the same day or delayed for a maximum of
1 day) was 17.6%. Timely help-seeking is associated with increased maternal
education and is higher in the agrarian population. There is significant regional
variability in the timeliness of help-seeking, which ranged from 3.7% in Somali
to 53.3% in Gambela (Table 7-49).
page-356
PREVALENCE AND TREATMENT OF DIARRHEA
Among the target children, 10.6% had had diarrhea within the 2 weeks prior
to the survey, without significant regional variability except in Benishangul-
Gumuz, in which the prevalence was 25.4%. Contrary to the expected result,
the source of drinking water and type of toilet facility were not significantly
associated with the occurrence of diarrhea. Among the children with diarrhea,
53.3% received treatment, with significant regional variability. Help-seeking
was lower in pastoralist areas than agrarian areas and in the lowest wealth
quintile (Table 7-50).
Among children with diarrhea, 45.3% were given ORS fluids, 13.3% were given
recommended homemade fluids (RHFs), and 51% were given either ORS fluids
or RHFs. Zinc was given to 26.5% of the children with diarrhea, and zinc with
ORS to 22.5%. (Table 7-51).
Table 7-48. Prevalence of ARI and treatment of ARI symptoms

Youngest child under 5: Among children under age 5 with symptoms of ARI:
Background Advice or Treatment was
characteristic ARI1 Unweighted treatment sought same or Unweighted
Number was sought2 next day number
National 17.9 2936 40.2 16.8 507

Cooking fuel
Electricity or 10.5 15 0 0 1
gas
Kerosene 0 5 0 0 0
Charcoal 26.1 61 25.5 13.4 18
Wood/straw 3
17.7 2809 40.2 17 479
Animal dung 27.5 44 59.1 16 9
Livelihood
Agrarian 18 1956 40.3 16.9 358
Pastoralist 14.6 980 36.9 14.6 149
Region
Tigray 23.9 265 38.8 18.5 63
Afar 8.3 187 58.6 23.4 22
Amhara 17.2 337 42.2 11.7 54
Oromia 19.6 743 40.6 19.7 122
Somali 22.5 333 30.3 6.7 75
page- 357
Benishangul- 23.6 209 86.9 40.2 51

Gumuz
SNNPR 12.2 497 35.9 11.5 56
Gambela 17.5 140 97.1 69.4 22
Harari 19.7 225 52.1 7 42
Mother’s education
No formal 18.5 1965 33.5 14.6 323
education
Attended 13.6 421 43.2 27.1 69
grade 1-4
Attended 20.5 372 56.7 13.4 78
grade 5-8
Attended 17.7 178 52 25.7 37
grade 9+
Wealth quintile
Lowest 15.4 556 32.1 5 108
Lower 13.6 556 47 20 74
Middle 19.4 569 46.6 29.2 113
Higher 17.3 648 26.9 7.1 94
Highest 22.5 607 47.3 19.5 118
Abbreviations: ARI, acute respiratory infection; SNNPR, Southern Nations, Nationalities, and Peoples
Region.
Table 7-49. Prevalence and treatment of fever for youngest children under 5
Percentage Percentage
Percentage Number
Number for whom for whom
Background Percentage who took of
of advice or treatment was
characteristic with fever antibiotic children
children treatment sought same
drugs with fever
was sought1 or next day
National 19.6 2 936 46.8 17.6 22.2 547
Livelihood
Agrarian 19.9 1 956 47.1 17.7 22.4 407
Pastoralist 13.5 980 38.3 14 14.4 140
Region
Tigray 22.2 265 47.5 17.9 26.7 59
Afar 10.6 187 37.2 6.6 9.4 20
Amhara 17 337 50.9 16 21.1 57
Oromia 20.4 743 45.1 20 21.1 114
Somali 20.7 333 28 3.7 15.4 64
Benishangul-
28.5 209 81.9 44.5 45.9 61
Gumuz
SNNPR 19.2 497 48.5 12.8 25.1 85
Gambela 39.7 140 68.9 53.3 20.4 53
page-358
Harari 17.2 225 58.3 4.8 34.4 34

No education 20.3 1 965 37.7 14.3 17.6 342
Primary 14.3 421 52 22.8 32.3 75
Secondary 23 372 71.3 23.1 31.2 84
More than
20.2 178 53.1 23 19.5 46
secondary
Wealth quintile
Lowest 19.2 556 29.5 4 7.7 116
Lower 13.1 556 52.7 5.5 33.8 72
Middle 21.4 569 38.1 22.5 17.7 127
Higher 22.1 648 47.6 19.8 24.4 118
Highest 20.8 607 63.2 25.7 27.5 114
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples Region.
Table 7-50. Prevalence and treatment of diarrhea for youngest children

under 5
Among children under age 5 with diarrhea:
Background Percentage Number of
with Percentage for whom Number of children
characteristic diarrhea children advice or treatment
was sought1 with diarrhea
National 10.6 2 936 53.3 330

Source of drinking water 2
Improved 10.2 1 953 63.2 227

Unimproved 11.5 983 32.2 103
Type of toilet facility3 11.4 423 63 53
Improved sanitation 10.7 1 063 50 117
facility
Unimproved facility 9.6 1 450 52.1 160
Open defecation
Livelihood
Agrarian 10.6 1 956 54 230
Pastoralist 9.5 980 37.3 100
Region
Tigray 9.1 265 54.4 24
page- 359
Afar 8.4 187 48.3 18

Amhara 10 337 60.2 34
Oromia 9.2 743 57 65
Somali 8.6 333 14.2 28
Benishangul-Gumuz 25.4 209 82.5 51
SNNPR 15 497 44 70
Gambela 11.3 140 90 17
Harari 12.1 225 57.6 23
No education 10.9 1 965 51.1 205
Primary 6.9 421 34.2 41
Secondary 14.8 372 65.1 54
More than secondary 7.5 178 77.9 30
Wealth quintile
Lowest 14.5 556 38 63
Lower 8.7 556 65.1 59
Middle 10.7 569 57.4 69
Higher 11.8 648 58.2 70
Highest 7.6 607 48.3 69
Abbreviations: SNNPR, Southern Nations, Nationalities, and Peoples Region.
Table 7-51. Oral rehydration therapy, zinc, and other treatments for diarrhea
Percentage of children with diarrhea who were given:

Percentage given
no treatment
children with
Number of
ORS packet or
Recommended
Either ORS or
pre-packaged
ORS and zinc

Fluid from
diarrhea
fluids (RHF)
homemade
ORS fluid
Background
characteristic
RHF
Zinc
National 45.3 13.3 51 26.5 22.5 46.7 330

Livelihood
Agrarian 45.2 13 51.1 26.4 22.6 46 230
Pastoralist 47 20.6 48.7 28.7 20.5 62.7 100
Region
Tigray 59 32.2 63.5 31.3 31.3 45.6 24
Afar 17.9 0.3 18.2 16.7 2.2 51.7 18
Amhara 32.1 15.1 46.4 18.7 12.8 39.8 34
Oromia 56.5 16 60.9 32.2 27.4 43 65
Somali 64.4 18.3 34.4 25.2 19.1 85.8 28
page-360
Benishangul- 65.3 9.8 69.6 32.1 20.1 17.5 51

Gumuz
SNNPR 32.8 5.2 35.6 21.8 20.1 56 70
Gambela 87.3 50.6 87.3 31.3 31.3 10.1 17
Harari 57.3 7.9 57.3 12.5 7.5 42.4 23
Mother’s
education
No formal 42.7 8.6 45.9 23.3 19 48.9 205
education
Attended 31.4 30.4 53.4 20.1 14.4 65.8 41
grade 1-4
Attended 57.7 14.5 61.1 39.2 38.6 34.9 54
grade 5-8
Attended 57 27.9 64.5 25.4 13.2 22.1 30
grade 9+
Wealth quintile
Lowest 24.4 12.1 34.3 17.2 13.6 62 63
Lower 54.2 19.7 66.3 28.8 28.5 34.9 59
Middle 30.8 11.4 36.9 39.5 28.8 42.6 69
Higher 62.4 9.8 62.6 24.2 24.1 41.8 70
Highest 54.8 18.1 56.6 23.4 17.1 51.7 69
Abbreviations: ORS, oral rehydration solution; RHF, recommended homemade fluids; SNNPR,
7.3.6 Child vaccination
UPTAKE OF BASIC AND NEWLY INTRODUCED VACCINATIONS
The overall percentage of children aged 12-23 months who had received the
Bacilli Calmette Guerin vaccine (BCG) was 79.3%, Oral Polio Vaccine (OPV)
1 was 82.4%, OPV 2 was 71.6%, OPV 3 was 51.4%, Penta 1 was 74.7%, Penta
2 was 65.7%, Penta 3 was 65.7%, and measles was 47.8%. Coverage of basic
vaccines was, in general, higher in agrarian areas than in pastoralist areas.
Tigray and Benishangul-Gumuz had relatively higher vaccination coverage,
while Somali had the lowest vaccination coverage for most antigens (Table
7-52). Seventy-five percent of children aged 12-23 months received their first
dose of the PCV vaccine, 62.8% received the second, and 49.5% the third. Of
the Rota vaccine, 76.3% received the first dose and 54.4% the second. (Table
7-53).
page- 361
COVERAGE OF ALL FULL BASIC VACCINES
All basic full vaccination coverage of children aged 12-23 months is the
composite indicator of: (a) 1 BCG, (b) 3 doses of DPT-HepB-HiB (pentavalent
vaccine), (c) 3 doses of polio vaccine excluding the dose at birth, and (d) 1
dose of the measles vaccine, according to the vaccination card or maternal
source.9 No vaccine at all is indicated if the child did not receive at least 1 of
the 14 vaccines, and incomplete vaccination is indicated if the child did not
receive all vaccines recommended by the Ministry of Health.
From this survey, it was found that 13.9% of children were not vaccinated
for any of the 14 vaccinations, 50.4% of children started vaccinations but
did not complete them, and 35.7% of children had received all full basic
vaccinations. Coverage of full basic vaccines was higher among children from
more educated mothers, wealthier households, and households with a larger
family size. Coverage of full basic vaccines was highest in Tigray (68.8%),
followed by Amhara (57.9%). In Amhara and Benishangul-Gumuz, there were
no children with a history of no vaccinations. In Somali, 63.8% of children
were not vaccinated at all, and only 0.1% of children had received full basic
vaccines. In pastoralist areas, only 8.2% of children were fully vaccinated for all
basic vaccines, 44.8% were not fully vaccinated, and 47% were not vaccinated
at all. Half of the children in agrarian areas (50.7%) failed to complete their
full dose of vaccinations (Table 7-54).
MOTHERS’ REASONS FOR NOT VACCINATING THEIR CHILDREN
Mothers of children with incomplete or no vaccinations were asked why they

had not gotten their children vaccinated. The mothers’ busy schedules, the
child’s being sick, and the absence of service on the date of the appointment
were common reasons for not vaccinating their children (Figure 7-13).
page-362
Figure 7-13. Mothers’ reasons for not getting their children vaccinated
Abbreviations: HP, health post; HEW, Health Extension Worker.
Mothers with an educational level of grades 1-4 gave the reason that their
child was sick (38.9%) for not getting their children vaccinated. Around 63% of
mothers with an educational level of grade 9 and above reported being busy
on the date of the appointment.
In the higher wealth quintile, 34.3% and 38.7% of women cited the cancellation
of the vaccine program and their own failure to remember the vaccine
appointment date, respectively, as their reasons for not getting their children
vaccinated (Table 7-55).
page- 363
Table 7-52. Percentage of children aged 12-23 months who received basic
vaccines
Background OPV Penta Unweighted

BCG Measles
characteristic 1 2 3 1 2 3 # Total
National 79.3 82.4 71.6 51.4 74.7 65.5 50.4 47.8 795
Mother’s age
15-19 73.6 75.2 59.1 46.4 74.7 72.7 69.7 66.9 43
20-24 76.1 76.3 65.6 44.6 70.7 62.2 45.2 43.6 180
25-29 79 80.2 71.6 47 71.2 59.1 44.2 42.4 241
30-34 73.5 85.1 75.7 55.6 78.5 64.3 47.5 43.6 158
35-39 84.8 87.3 72.5 55.8 75.2 72.1 57.7 54.9 127
40-44 90.3 90 79.9 66.6 90 79.8 62 61.6 37
45-49 97.8 97.8 97.8 86.9 97.8 97.8 97.8 97.8 9
Education
No formal
75.4 79.5 68.1 48.2 71.3 63.3 49.1 45.7 542
education
Attended
90.5 86.9 70.6 47.2 78 61.2 49.5 43.9 105
grade 1-4
Attended
82.8 89.8 84.5 67.5 85.3 75.9 58.8 59.3 106
grade 5-8
Attended
80.1 80.1 81.5 58.7 75.3 79.4 43.6 54.1 42
grade 9+
Wealth quintile
Lowest 75.9 80.3 70 59.1 69.7 62.8 57.8 48.8 188
Lower 74.2 76.8 70.5 46.7 70.8 62.1 48.1 46.1 155
Middle 74.7 82.4 68.5 47.7 75.1 67.3 48.2 48.7 150
Higher 88.1 87.3 73.9 51.3 78.4 71.3 58.6 57.1 147
Highest 81.6 84 74.3 52.8 78.1 63.2 40.1 38.6 155
No of Children
1-2 Children 80.5 80.9 73.9 46.2 70.1 62.5 47 53.1 228
3-4 Children 81.1 85.3 71.6 53.9 82.5 69.9 57.8 40.4 252
5-7 Children 77.6 81.5 69.3 52.6 71.5 62.8 44.9 47.2 225
8+ Children 79 87.4 77.6 67.5 85.8 80.2 63.4 65.6 42
Region
Tigray 95 97.4 92.4 81.2 93.5 88.7 82.4 79.6 64
Afar 58.3 56.6 51.3 38.9 56.6 51.7 39.4 28.5 59
Amhara 97.3 97.1 91 80.6 93.6 87.3 74.7 77.1 84
page-364
Oromia 75.3 78.4 65 41.1 70.9 58.9 42.7 37.1 167

Somali 28.1 29.3 22.9 10 8.1 2.6 0.8 1.5 158
Benishangul-
100 100 97.6 65.3 100 99 79.4 53.6 36
Gumuz
SNNPR 77.5 84.1 73.1 49.6 73.2 65.9 47.1 47.6 119
Gambela 75.7 70.1 70.1 65.5 70.2 67.7 38.2 41.3 41
Harari 75.6 81.3 73 54.6 78.7 75.3 55.7 33.8 67
Livelihood
Agrarian 81.7 84.6 73.7 53.3 77.6 68.1 52.6 50.3 473
Pastoralist 43 48 40.3 22.8 30.9 25.4 15.8 10.7 322
page- 365
Table 7-53. Percentage of children aged 12-23 months who received PCV and
Rota vaccines
Background PCV Rota Unweighted

characteristics 1 2 3 1 2 Total
National 75.1 62.8 49.5 76.3 54.4 795
Mother’s age at birth
15-19 73.9 72.7 45.2 74 54.6 43
20-24 69 58 41.2 72.8 51 180
25-29 71.9 56.5 47 74.6 55.5 241
30-34 78.1 62.3 45.9 78.3 53.3 158
35-39 78.9 71.2 59.6 79.8 60.9 127
40-44 90 79.4 70.3 79.9 39.2 37
45-49 86.8 55.8 55.8 86.8 55.8 9
Education
No formal education 71.5 59.9 48 72 50.5 542
Attended grade 1-4 78.1 62.6 47.5 81.3 53.6 105
Attended grade 5-8 86.8 75.3 58 89.4 68.3 106
Attended grade 9+ 76.4 62.3 49.6 76.5 65.9 42
Wealth quintile
Lowest 71.9 63.5 55.1 74.5 59 188
Lower 70.9 61.8 40.8 71.9 49.8 155
Middle 75.6 59.7 51 76.4 57.8 150
Higher 81.7 68.6 57.6 83.3 51.5 147
Highest 74.4 60 43.4 75.2 54.8 155
No of Children
1-2 Children 69.6 59.1 44.5 74.4 52.1 228
3-4 Children 83.1 67.5 57.4 83.3 60.8 252
5-7 Children 72.9 58.8 42.4 73.2 51.4 225
8+ Children 85.1 84.7 76 79.2 57.6 42
Region
Tigray 93.4 83.9 78.4 96.4 77.1 64
Afar 56.5 49.6 37.8 56.5 52.4 59
Amhara 93.6 91.2 77.1 94.6 80.1 84
Oromia 71.9 53.6 41.7 73.7 47.4 167
Somali 5.2 3.7 1.8 7.6 3.9 158
Benishangul-Gumuz 99 96.6 75.5 100 92.7 36
page-366
SNNPR 73.1 63.6 44.4 73.1 50.5 119

Gambela 70.1 58.6 33.8 72.4 39.1 41
Harari 79.8 69.7 50.4 79.8 59.6 67
Livelihood
Agrarian 78 65.2 51.8 79.3 56.7 473
Pastoralist 31.2 26 14.9 32.5 19.4 322
Table 7-54. Percentage of children aged 12-23 months who received complete,
partial, and no vaccinations
Background No Vaccine not All full basic

Unweighted Total
characteristics vaccination completed vaccinations
National 13.9 50.4 35.7 795
Age
15-19 23.2 31.3 45.4 43
20-24 19 48.6 32.4 180
25-29 14.4 56.5 29.1 241
30-34 13.4 52.9 33.7 158
35-39 9 51.5 39.5 127
40-44 9.7 28.7 61.6 37
45-49 2.2 11 86.8 9
Education
No formal education 16.9 47.9 35.1 542
Attended grade 1-4 4.5 65.8 29.7 105
Attended grade 5-8 10.1 44.6 45.3 106
Attended grade 9+ 18.7 46.6 34.7 42
Wealth quintile
Lowest 15.5 41 43.4 188
Lower 21 45.5 33.5 155
Middle 13.2 53.4 33.3 150
Higher 10 48.3 41.7 147
Highest 10.8 61.5 27.7 155
No of Children
1-2 Children 14.8 49.1 36.1 228
3-4 Children 12.1 55.4 32.5 252
page- 367
5-7 Children 15.1 49.7 35.2 225

8+ Children 5.3 38.8 55.9 42
Region
Tigray 1.6 29.6 68.8 64
Afar 37.8 35.1 27.1 59
Amhara 0 42.1 57.9 84
Oromia 17.3 54.8 27.9 167
Somali 63.8 36.1 0.1 158
Benishangul-Gumuz 0 66.1 33.9 36
SNNPR 12.6 54.6 32.8 119
Gambela 24 47.7 28.2 41
Harari 14.7 54.2 31 67
Livelihood
Agrarian 11.7 50.7 37.6 473
Pastoralist 47 44.8 8.2 322
page-368
Table 7-55. Mothers’ reasons for not vaccinating their children
No service /HP
Child Was Sick
on the date of
appointment
appointment
Not wanting
I don’t Know
closed/HEW
Fear of side
Unweighted
Vaccination
session was
Reason not
mentioned
cancelled
I forgot
date of
absent
effect
Background
Total
Busy
characteristics
National 14.9 16.6 12.7 10.9 13.4 10.2 7.4 7.9 6.1 167
Education
No formal
16.3 16.5 16 13.9 14.5 12.5 6.6 2.7 1 132
education
Attended
38.9 0 4 2.2 0.2 8.2 6.3 34.7 5.6 16
grade 1-4
Attended
1.3 0.2 0.2 0 8.6 0.2 8.4 34.9 46.1 11
grade 5-8
Attended
0 62.7 0 0 20 0.4 16.5 0.4 0 8
grade 9+
Wealth quintile
Lowest 1.3 6.4 12.5 13 33.1 12.1 16.5 4.6 0.6 52
Lower 2.8 29.2 20.7 8.9 18.7 2.5 2.9 14 0.2 44
Middle 53.9 12.4 22.5 1.7 1.8 1.2 6.4 0 0 25
Higher 7.1 3.4 0.1 34.3 6.8 38.7 3.2 6.6 0 26
Highest 17 23.3 0 0.2 0 4.7 9.5 10.6 34.7 20
Livelihood
Agrarian 16.8 16.6 11.9 13 14.3 6.6 4 9.6 7.1 54
Pastoralist 7.8 16.5 15.4 2.8 10 23.6 20.2 1.4 2.2 113
Abbreviations: HP, health post; HEW, Health Extension Worker; SNNPR, Southern Nations,
7.3.7 Infant and young child feeding
BREASTFEEDING
As a national public health recommendation, infants should be exclusively

breastfed for the first 6 months of life. To meet their evolving nutritional
requirements, they should receive adequate and safe complementary foods
while breastfeeding continues up to 2 years of age and beyond.
page- 369
Exclusive breastfeeding is defined as no other food or drink, including water,

than breastmilk (including milk expressed or from a wet nurse) for the first 6
months of life, but does allow the infant to receive ORS, drops, and syrups
(vitamins, minerals, and medicines).55
Continued breastfeeding at 1 year is the proportion of children aged 12 to 15

months who are fed breastmilk.
Continued breastfeeding at 2 years is the proportion of children aged 20 to 23

months who are fed breastmilk.
Introduction of solid, semi-solid or soft foods is the proportion of infants aged

6 to 8 months who receive solid, semi-solid, or soft foods.
Age-appropriate breastfeeding is the proportion of children aged 0 to 23

months who are appropriately breastfed, as calculated from the following:
• Infants 0-5 months of age who received only breastmilk the previous
day; and
• Children 6-23 months of age who received breastmilk, as well as solid,
semi-solid, or soft foods.
More than half (58.7%) of children aged 0-23 months were exclusively
breastfed until the age of 6 months. More children in agrarian areas (60%)
were exclusively breastfed than children in pastoralist areas (31%). Somali had
the lowest percentage of exclusive breastfeeding (3%), while Tigray had the
highest (84%). The proportion of children who continued breastfeeding until 1
year was 94%, and 86% continued until 2 years. Age-appropriate breastfeeding
for children aged 6-23 months was 84% (Table 7-56).
FREQUENCY OF BREASTFEEDING AND KNOWLEDGE ABOUT TIME

FOR COMPLEMENTARY FEEDING INITIATION
Twenty percent of children aged 0-5 months were breastfed fewer than 7 times
per day, while 55% were breastfed 8 -12 times per day. Twenty-three percent
of children less than 2 years old were fed using a bottle. The percentage of
children being bottle-fed decreased with an increase in the age of the mother
and increased with an increase in the mother’s wealth index (Table 7-57).
page-370
MINIMUM MEAL FREQUENCY
According to IYCF guidelines, the minimum meal frequency (MMF) is defined

as the proportion of breastfed and non-breastfed children aged 6 -23 months
who received solid, semi-solid, or soft foods (including milk for non-breastfed
children) at least the minimum number of times daily.55 Minimum is defined as:
• 2 times daily for breastfed infants aged 6 to 8 months;
• 3 times daily for breastfed children aged 9 to 23 months; and
• 4 times daily for non-breastfed children aged 6 to 23 months.
The proportion of breastfed children aged 6-23 months fed with MMF was
77.6%, ranging from 60.1% to 82.8% for different age categories. MMF varies
across regions, ranging from 49.5% in Gambela to 81.6% in Oromia and 85% in
Afar. Children aged 6-23 months born to mothers aged 15-19 had lower MMF
(55%) than those children born to mothers aged 35-49 (80%) (Table 7-58).
MINIMUM DIETARY DIVERSITY (MDD)
Minimum dietary diversity (MDD) is defined as the proportion of children

aged 6-23 months who received foods from 4 or more food groups.55 The 7 food
groups used to calculate this indicator are:
page- 371
• food group 1: grains, roots, and tubers

• food group 2: legumes and nuts
• food group 3: dairy products (milk, yogurt, cheese)
• food group 4: flesh foods (meat, fish, poultry, and liver/organ meats)
• food group 5: eggs
• food group 6: vitamin A-rich fruits and vegetables
• food group 7: other fruits and vegetables
MINIMUM ACCEPTABLE DIET (MAD)
Minimum acceptable diet (MAD) is defined as the proportion of children

aged 6-23 months who received a minimum acceptable diet (apart from
breast milk). It is the composite indicator of:
• The proportion of breastfed children aged 6-23 months who had at least
the minimum dietary diversity and MMF the previous day; and
• The proportion of non-breastfed children aged 6-23 months who received
at least 2 milk feedings and had at least the MDD, not including milk
feeds and the MMF, during the previous day.
From the 7 food groups, the majority (70.6%) of children aged 6-23 months
received food from the grains, roots, and fibers group (food group 1) followed
by legumes and nuts (foodgroup 2), which were consumed by 27% (Table
7-59).
Only 15% of children were fed according to the minimum standards with respect
to food diversity (4 or more food groups). Children in agrarian areas were more
likely to be fed according to IYCF-recommended feeding practices than were
pastoralist children. In addition, feeding practices improved with an increase in
the wealth quintile and educational level of the mother (Table 7-60).
Overall, 14.4% of children received MAD. The percentage of children receiving

MAD differs between regions. Gambela and Harari had the highest achievement
of MAD (29% and 25%, respectively), compared to the national average of
14% (Table 7-60, Figure 7-9).
page-372
Figure 7-14: Percentage distribution of children aged 6- to 23 months with
minimum acceptable diet, by region
page- 373
CONSUMPTION OF IRON-RICH FOODS AND KNOWLEDGE OF

INITIATION OF COMPLEMENTARY FEEDING
Of all children aged 6-23 months, 45% consumed iron-rich or iron-fortified

food. Children’s iron intake increases with an increase in the mother’s age and
household wealth index. Iron consumption is high in Afar and Benishangul-
Gumuz (45% each) and lower in Tigray (16%) and the SNNPR (17%).
Women with children aged 0-23 months were asked about the initiation
of complementary feeding. Responses showed that 88.8% responded that
complementary feeding should be started between 6 and 8 months. Nineteen
percent of mothers in pastoralist areas did not know the exact time to start
complementary feeding for their children (Table 7-61).
Table 7-56. Breastfeeding status of youngest children under age 2

characteristics
Continued Continued
Background
6-23 All age 0-23

EBF breastfeeding breastfeeding
appropriate bf appropriate bf
at 1 year at 2 years
% Number % Number % Number % Number % Number

National 58.7 466 94.05 366 85.6 114 83.6 1057 76.8 1 523
Maternal Age
15-19 64.3 36 98.6 12 100 5 69.3 48 66.5 84
20-34 58.9 341 96.4 277 83.8 78 83.5 780 76.8 1 121
35-49 56.2 89 87.6 77 87.8 31 85.3 229 78.3 318
Marital status
Currently
59.09 435 94.3 341 85.5 111 83.3 1 007 76.6 1 442
married
Divorced 50.1 9 100 10 - - 98.8 19 84.8 28
Widowed 11.3 7 22.8 7 - - 28.8 11 20.8 18
Separated 55.01 14 99.7 7 22.9 2 99.5 17 94.6 31
Never
100 1 100 1 100 1 100 3 100 4
married
Educational status
page-374
No formal
59.9 306 93.03 251 89.1 70 83.9 689 77.8 995
education
Attended
54.4 60 94.5 52 77.08 23 80.5 153 73.6 213
grade 1-4
Attended
45.3 66 95.7 43 80.08 17 82.2 157 71.9 223
grade 5-8
Attended
78.1 34 100 20 98.9 4 96.3 58 87.6 92
grade 9+
Livelihood
Agrarian 60.4 302 95.04 219 86.7 84 84.3 706 77.7 1 008
Pastoralist 31.2 164 75.8 147 50.9 30 70.4 351 57.5 515
Region
Tigray 83.6 47 100 37 89.4 11 79.6 105 80.9 152
Afar 56.3 28 74.4 33 61.9 7 79.1 63 71.7 91
Amhara 47.2 38 84.06 34 84.9 19 85.7 130 77.3 168
Oromia 60.4 130 98.3 71 91.3 25 85.1 248 77.6 378
Somali 2.9 56 62.3 67 4.9 7 54.7 123 39.6 179
Benishangul-
61.3 41 98.3 24 29.7 5 92.2 66 81.3 107
Gumuz
SNNPR 59.5 64 94.8 62 71.5 18 81.7 184 76.4 248
Gambela 27.2 13 100 12 88.5 6 91.4 43 77.2 56
Harari 71.1 49 91.3 26 54 16 79.1 95 76.5 144
Wealth quintile
Lowest 65.9 91 87.7 78 87.7 21 78.3 198 75.01 289
Lower 60.7 81 97.5 79 96.6 22 87.2 201 79.4 282
Middle 49.9 87 99.6 70 92.1 17 89.02 211 81.1 298
Higher 54.4 117 91.4 69 81.5 19 84.8 215 74.7 332
Highest 64.05 90 90.5 70 75.7 35 78.01 232 74.2 322
National 58.7 466 94.05 366 85.6 114 83.6 1 057 76.8 1 523
Abbreviations: bf, breastfeeding; EBF, exclusively breastfed; SNNPR, Southern Nations, Nationalities,
and Peoples Region.
page- 375
Table 7-57. Breastfeeding frequency, initiation of complementary feeding, and

bottle feeding among children aged 6-23 months
Breastfeeding frequency Complementary initiation Bottle feeding
Background
characteristics ≤7 8-12 >12 Unweighted % Unweighted Total % Total
times times times Total
National 20.5 54.6 24.9 1 424 66.6 228 23.1 1 057
Mother's Age
15-19 19.1 32.3 48.6 83 48.8 20 40.51 48
20-34 21.6 55.0 23.4 1 038 66.5 167 24.88 780
35-49 17.4 56.8 25.7 303 72.1 41 16.7 229
Live birth order
≤2 23.3 50.3 26.4 461 63.3 72 25.1 332
3-4 22.5 52.5 24.9 430 73.9 79 23.4 344
5-7 16.5 57.7 25.7 419 69.8 61 20.8 307
≥8 15.3 66.5 18.1 114 52.3 16 21.6 74
Educational status
No formal 20.4 53.4 26.2 918 69.9 143 18.99 689
education
Attended 23.3 58.4 18.2 204 58.5 34 27.8 153
grade 1-4
Attended 23.2 56.2 20.5 211 54.6 38 32.4 157
grade 5-8
Attended 7.2 51.5 41.2 91 99.03 13 25.74 58
grade 9+
Livelihood
Agrarian 19.8 55.9 24.8 976 66.4 143 22.71 706
Pastoralist 34.5 38.8 26.5 448 70.4 85 31.21 351
Region
Tigray 22.7 48.1 29.2 151 53.4 22 2.05 105
Afar 51.2 48.6 0.1 78 90.3 17 15.51 63
Amhara 29.2 48.9 21.9 160 78.2 23 10.97 130
Oromia 18.2 53.4 28.4 371 68.4 64 33.78 248
Somali 47.7 28.09 24.1 135 76.8 25 21.33 123
Benishangul- 43.2 51.3 5.5 104 89.7 11 28.04 66
Gumuz
SNNPR 15.4 65.8 19.05 238 53.8 39 16.49 184
Gambela 7.4 76.8 15.7 54 74.4 9 20.93 43
Harari 13.9 55.5 30.4 133 66.7 18 43.17 95
Wealth quintile
Lowest 24.2 49.6 26.1 264 59.6 38 7.4 198
Lower 15.8 57.7 26.4 263 79.6 38 18.7 201
Middle 33.5 54.1 12.3 285 59.8 49 20.4 211
Higher 15.4 52.2 32.4 315 79.8 45 24.08 215
Highest 14.4 58.3 27.3 297 61.7 58 38.6 232
page-376
Table 7-58. Minimum meal frequency

Percentage of youngest children aged 6-23 months who are fed a minimum acceptable diet
based on breastfeeding status, and times they were fed during the day or night before the
survey, by background characteristics
2 times for
3 times for 4 times for non-
breastfed MMF 6-23
breastfed children breastfed children
Background infants 6-8 months
9-23 months 6-23 months
Characteristics months
Total Total Total Total
% % % %
number number number number
National 60.1 226 82.8 729 77.1 102 77.6 1 057
Mother’s age
15-19 47.9 20 65.4 26 86.4 2 55.4 48
20-34 62.4 165 81.8 534 75.9 81 77.4 780
35-49 55.6 41 85.8 169 79.5 19 80.2 229
Livelihood
Agrarian 59.69 143 82.87 523 76.7 40 77.7 706
Pastoralist 67.2 83 77.7 206 79.1 62 75.5 351
Region
Tigray 42.7 22 62.2 82 100 1 58.4 105
Afar 86.6 17 79.2 33 99.5 13 85.1 63
Amhara 69.4 23 78.3 99 82.9 8 76.9 130
Oromia 65.1 64 87.9 172 56.7 12 81.6 248
Somali 73.1 24 62.4 59 70.7 40 67.4 123
Benishangul-
89.7 11 82.2 51 21.9 4 79.5 66
Gumuz
SNNPR 41.5 38 81.8 134 99.5 12 74.9 184
Gambela 73.2 9 43.9 33 0 1 49.5 43
Harari 66.7 18 84.9 66 77.3 11 80.1 95
Wealth quintile
Lowest 53.8 38 80.9 138 89.02 22 73.8 198
Lower 64.5 38 86.8 141 77.7 22 83.8 201
Middle 48.7 48 80.5 153 86.7 10 74.5 211
Higher 79.8 45 86 150 63.4 20 83.3 215
Highest 55.5 57 79.6 147 79.7 28 73.3 232
Abbreviation: MMF, minimum meal frequency; SNNPR, Southern Nations, Nationalities, and Peoples
Region.
page- 377
Table 7-59. Food groups consumed by children under age 2 in the day or night
preceding the interview
Background Group Group Group Group Group Group Group Unweighted

characteristic 1 2 3 4 5 6 7 Total number
National 70.6 27 16.3 16 17.3 23.9 23.4 1 057
Mother’s age
15-19 28.6 24.6 34 11.9 12.7 24.1 22.4 48
20-34 29.1 30.1 15.8 15.1 17.6 25.4 24.4 780
35-49 30 19 16.2 18.6 16.8 19.8 20.7 229
Marital Status
Currently
70.5 27 16.6 15.9 17.1 23.7 23.2 1 007
Married
Divorced 90.6 38.8 1.4 10.1 15.7 38.3 47.5 19
Widowed 94.2 8.5 81.3 3.4 71.2 89.1 80.6 11
Separated 63.3 14.6 0.4 33 24.8 2.6 2.9 17
Never married 25.7 74.3 0 0 0 99.9 25.5 3
Wealth quintile
Lowest 64.7 28.9 2.5 4.9 5.4 9 4.7 198
Lower 71.5 32.4 14.9 8.9 4.4 22.8 19.2 201
Middle 71.4 13.8 9.8 14.9 10.6 14.7 16.2 211
Higher 75.8 22.8 18.8 21.7 29.4 26.1 30.1 215
Highest 68.3 39.4 31.3 24.1 29.8 41.8 39.8 232
Region
Tigray 77.2 37.2 12 8.2 19.4 9 6.7 105
Afar 55 31.9 8.4 36 17.7 15.3 21.2 63
Amhara 62.5 22.8 7.5 13.2 14.6 21.9 13.8 130
Oromia 76.1 35.2 23.1 22.3 23.2 29.6 28.5 248
Somali 46 14.3 13 10.1 0 8.6 6.3 123
Benishangul-
79.4 22 23.4 50 41.4 66
Gumuz 2.9 40.6
SNNPR 67.1 11.9 11.4 6.4 7.8 18 26.2 184
Gambela 52.6 25.1 31 47.1 13.8 46.1 39 43
Harari 74 33.4 14.4 25 24.9 30.2 42.2 95
Livelihood
Agrarian 71.4 27.4 16.4 16 17.7 24.3 24.1 706
Pastoralist 54.4 19 14.2 14.7 9.1 15.9 8.9 351
page-378
Table 7-60. Percentage of youngest children aged 6-23 months who were fed a
minimum dietary diversity during the day or night before the survey
MDD Breastfed Non-breastfed Non-breastfed MAD

Background MAD MDD MAD
characteristics
% No. % No. % No. % No. % No.
National 15.1 1 057 13.9 955 22.5 102 21.7 102 14.4 1057
Mother’s age
15-19 17.8 48 17.9 46 0 2 0 2 17.8 48
20-34 16.8 780 15.2 699 21.2 81 20.09 81 15.4 780
35-49 11.2 229 10.4 210 25.2 19 25.2 19 11.2 229
Marital status
Currently 15.4 1 007 14.1 910 19.9 97 19.15 97 14.4 1 007
married
Divorced 18.2 19 18.4 18 0 1 0 1 18.2 19
Widowed 80.6 11 32.7 10 100 1 100 1 80.6 11
Separated 0.5 17 0.3 14 50.1 3 50.1 3 0.5 17
Never 0 3 0 3 - - - - 0 3
married
Educational status
No formal 10.8 689 9.9 610 9.4 79 8.5 79 9.9 689
education
Attended 11.5 153 8.9 142 25.4 11 25.4 11 9.8 153
grade 1-4
Attended 32.5 157 29.5 147 62.6 10 61.8 10 31.9 157
grade 5-8
Attended 28.6 58 29.2 56 1.6 2 1.6 2 28.5 58
grade 9+
Livelihood
Agrarian 15.7 706 14.1 666 26.7 9 26.4 40 14.6 706
Pastoralist 9.4 351 10.1 289 2.5 3 0 62 8.2 351
Region
Tigray 9.7 105 8.9 104 0 1 0 1 8.8 105
Afar 23.06 63 21.5 50 30.2 13 0 13 17.6 63
Amhara 8.8 130 6.7 122 29.7 8 29.7 8 8.2 130
Oromia 22.5 248 20.2 236 41.3 12 41.3 12 20.8 248
Somali 3.3 123 2.3 83 0 40 0 40 1.5 123
BenishangulGumuz 30.2 66 25.3 62 82.7 4 13.3 4 24.5 66
SNNPR 8.07 184 8.5 172 0 12 0 12 8.07 184
Gambela 35.4 43 29.7 42 0 1 0 1 29.3 43
Harari 27.01 95 28.7 84 15.4 11 0 11 25.06 95
Wealth quintile
Lowest 5.6 198 5.9 176 0.2 22 0 22 5.6 198
Lower 11.9 201 12.2 179 2.7 22 0 22 11.8 201
Middle 6.7 211 6.02 201 0.8 10 0.8 10 5.9 211
Higher 15.9 215 16.2 195 8.22 20 6.8 20 15.6 215
Highest 32.8 232 27.4 204 52.06 28 52.06 28 29.4 232
Abbreviations: MAD, minimum acceptable diet; MDD, minimum dietary diversity; SNNPR, page- 379
Southern Nations, Nationalities, and Peoples Region. National Assessment of
Table 7-61. Percentage of youngest children aged 6-23 months who are
fed iron-rich or iron-fortified food, HEW visits for IYCF, and knowledge of
complementary feeding
Consumption HEW visit Knowledge on complementary feeding initiation

Background of iron-rich or for IYCF
characteristics iron-fortified I do
≤5 6-8 ≥9 Unweighted
foods not
months months months total
know
National 28.6 1057 28.8 1549 3.5 88.8 1.7 5.9 1565
Mother’s age
15-19 20.02 48 26.4 87 13.7 76.6 1.06 8.5 87
20-34 26.9 780 28.1 1 136 3.1 88.3 1.9 6.5 1 148
35-49 33.7 229 31.3 326 3.05 91.9 1.09 3.8 330
Live birth order
≤2 26.2 332 25.9 491 3.7 88.04 0.7 7.5 498
3-4 32.7 344 29.2 476 2.5 88.3 2.7 6.6 480
5-7 26.5 307 32.1 463 4.02 92.09 1.4 2.4 467
≥8 29.5 74 28. 119 4.6 83.6 3.3 8.5 119
Educational status
No formal
26.1 689 26.1 1 013 3.6 89.1 2.5 4.7 1 024
education
Attended
32.7 13 32.5 217 1.4 86.9 0.9 10.7 218
grade 1-4
Attended
34.7 157 33.6 225 4.3 88.5 0.4 6.7 229
grade 5-8
Attended
24.05 58 32.2 94 5.7 91.7 0 2.5 94
grade 9+
Livelihood
Agrarian 28.7 706 29.6 1 024 3.3 89.7 1.7 5.2 1 032
Pastoralist 24.7 351 13.1 525 7.07 71.1 2.4 19.4 533
Region
Tigray 15.6 105 43.6 152 4.09 85.02 4.9 6.6 152
Afar 44.5 63 16.7 95 5.7 70.7 0.1 23.5 95
Amhara 19.9 130 39.09 169 1.8 89.01 4.4 4.7 170
Oromia 39.04 248 22.5 386 4.7 87.09 0.6 7.5 391
Somali 24.6 123 3.6 184 11.3 53.4 3.8 31.5 188
Benishangul-
44.8 66 41.5 109 8.4 88.4 0 3.18 110
Gumuz
SNNPR 16.9 184 32.9 251 1.3 97.05 1.07 0.5 253
page-380
Gambela 73.4 43 37.3 59 10.4 69.4 3.2 16.9 62

Harari 33.4 95 12.9 144 6.7 90.3 1.04 1.4 144
Wealth quintile
Lowest 11.9 198 26.3 293 3.8 83.8 4.9 7.4 296
Lower 22.3 201 31.7 290 1.2 89.3 1.7 7.8 291
Middle 24.5 211 35.4 301 1.5 92.2 1.2 5.3 306
Higher 38.08 215 24.09 336 4.4 90.2 0.2 5.1 338
Highest 39.4 232 27.4 329 3.8 83.8 4.9 7.4 334
Abbreviation: HEW, Health Extension Worker; IYCF, infant and young children feeding, SNNPR,
7.3.8 The role of the HEP in the delivery of family health services
The HEP has contributed significantly to the delivery of essential health services
to the community. The contribution is highest in FP and TT vaccination and
lowest in delivery service, followed by diarrhea treatment for children. Although
the 1st and 4th ANC series are expected to be provided in HCs, 32.2% of
pregnant women received all their ANC services at an HP. This is not without
potential risk to the mothers. This needs programmatic attention to minimize
the potential risk of providing all ANC at HPs (Figure 7-15).
Figure 7-15. Proportion of selected services provided by HEWs at health posts

or in home visits among those who got these services from any healthcare
provider
Abbreviations: ANC, antenatal care; TT, tetanus-toxoid; HEW, Health Extension Worker; PNC, Post-natal
care; IYFS, Infant and Young Child Feeding
page- 381
FAMILY PLANNING
A majority of women (56.6%) who currently use a contraceptive method

received the service from their catchment HP. The second most common source
of service provision for contraception for women was the HC (33.7%). For youth
girls, however, HCs (42.3%) served as the main place for FP, followed by HPs
(40.7%) and private clinics (13.4%; Figure 7-16).
Figure 7-16. Place of FP service delivery for women and youth girls
In addition to serving as sources of services, the HEP has also been the most
common source of FP information for communities, with HEWs being mentioned
as a primary source of information by 61.8% of women and 37.7% of men
(Figure 7-17)
page-382
Figure 7-17: Primary sources of information on family planning
Abbreviations: HEW, Health Extension Worker; WDA, Women’s Development Army; SMC, Social
Mobilization Committee.
HEWs are the most common source of FP information for both agrarian women
(62.3%) and men (37.8%) and pastoralist women (49.6%) and men (34.1%).
School teachers, however, are the primary source of information for youth girls
in agrarian settings (45.6%; Table 7-62).
page- 383
Table 7-62. Primary source of information about family planning, by livelihood
professional
Unweighted
Friends or
teachers
relatives
WDA or
number
Others
Health
Rradio
School
HEWs
TV or
Background
SMC
total
characteristics
Women 61.8 1.3 15.6 2.6 15.7 1.5 1.6 4 483

Men 37.7 0.8 4.2 26 4.2 21.5 5.5 4 416
National
Youth
15.2 0.9 4.6 4 21.9 45.2 8.1 793
girls
Women 62.3 1.2 15.6 2.6 15.2 1.5 1.6 3 193
Men 37.9 0.8 4.2 26.3 4.2 21.2 5.4 3 157
Agrarian
Youth
14.9 0.9 4.6 4.1 21.9 45.6 8.1 615
girls
men 49.7 4 14.1 2 27.9 1.1 1.3 1 290
Women 34.1 2.1 4.3 17.7 4.7 29.2 7.9 1 259
Pastoralist
Youth
32.7 2.3 7.9 1.6 22 28 4.8 178
girls
Abbreviations: HEW, Health Extension Worker; WDA, Women’s Development Army; SMC, Social
Mobilization Committee.
MATERNAL AND NEWBORN HEALTH (ANC, DELIVERY, PNC)
Ethiopia’s current ANC guidelines recommend ANC visits at HPs and HCs. As
the guidelines stipulate, the first and last ANC visits should be conducted at
the HC, while the second and third visits can be conducted at the HP. Despite
this guidance, this assessment revealed that HPs were the primary place for
ANC-1, while HCs were the primary place for ANC-4 for women in their last
pregnancy.
Of the women who had had at least 1 ANC visit, 49% of them had their first
visit at the HP. The percentage of women who attended their first ANC visit at
the HP ranged from 26% in Amhara to 90% in Somali. Nearly 69% of women
in pastoralist areas received at least their first ANC visit at the HP, and 49% of
women in agrarian regions attended their ANC visits at the HP (Table 7-63).
page-384
Among women who had at least 4 visits of ANC, the HC was the main (59.4%)
place for ANC-4. Thirty-two percent (56.2% in pastoralist and 31.8% in agrarian
areas) of women who had at least 4 ANC visits attended ANC-4 at the HP
(Table 7-64).
Among women who reported being told about danger signs of pregnancy
during their most recent pregnancy, 23% mentioned HEWs as their primary
sources of information (Table 7-65). Similarly, HEWs also served as important
sources of information on a birth preparedness plan (Figure 7-18; Table 7-66).
Figure 7-18: Primary source of information about danger signs of pregnancy

among mothers who delivered during the last 5 years
More than half of the women (51.8%) in this study delivered their youngest child
with the assistance of health service providers, followed by relatives (37.1%).
HEWs attended 3.7% of deliveries.
Among women who received PNC for their most recent delivery during the
last 2 years, the HC was the most common place of service provision (44.3%),
followed by home (23.4%). Only 12.3% of PNC services took place in HPs.
HEWs provided 43.2% of PNC follow-up visits (Table 7-67, Table 7-68).
page- 385
Table 7-63. Place of first antenatal care visit among mothers who had at least
1 visit during their most recent pregnancy
Background Health Health Government Private Unweighted number of

Other
characteristic Post Center Hospital Hospital women
National 49.1 45.1 5.1 0.5 0.1 2 186
Livelihood
Agrarian 48.6 45.7 5.0 0.5 0.1 1 669
Pastoralist 68.6 22.0 8.2 1.1 0.1 517
Region
Tigray 28.5 63.0 7.3 0.9 0.3 244
Afar 55.2 26.1 18.3 0.4 0.0 134
Amhara 26.0 72.0 2.0 0.0 0.0 343
Oromia 50.5 41.4 7.2 0.8 0.0 578
Somali 90.1 6.4 3.5 0.0 0.0 59
Benishangul- 52.9 43.0 2.4 0.5 1.1 178
Gumuz
SNNPR 74.3 22.4 2.8 0.3 0.2 406
Gambela 40.5 54.8 5.0 0.0 0.0 87
Harari 31.4 47.8 18.1 2.7 0.0 157
15-19 70.3 23.1 2.5 3.9 0.1 86
20-34 50.0 43.2 6.2 0.6 0.1 1 502
35-49 45.4 2.8 2.8 0.2 0.0 598
Live birth
≤2 46.8 44.4 7.5 1.1 0.1 723
3-4 48.7 45.9 5.1 0.2 0.2 657
5-7 52.3 43.8 3.7 0.1 0.0 629
≥8 49.4 49.4 0.2 0.9 0.0 177
Education
No formal 48.3 48.4 3.0 0.2 0.0 1 329
education
Attended grade 1-4 56.8 36.8 5.6 0.5 0.3 354
Attended grade 5-8 48.2 41.0 9.5 1.2 0.1 342
Attended grade 9+ 36.9 51.1 10.4 1.6 0.0 161
Wealth quintile
Lowest 56.3 39.0 4.0 0.6 0.0 353
Lower 55.4 41.7 2.5 0.1 0.3 383
Middle 54.9 42.1 2.7 0.1 0.1 420
Higher 45.5 47.3 6.1 1.0 0.0 525
Highest 36.9 52.9 9.4 0.8 0.0 505
Abbreviation: SNNP, Southern Nations, Nationalities, and Peoples Region.
page-386
Table 7-64. Place of fourth antenatal care visit among mothers who had at
least 4 visits during their most recent pregnancy
Background characteristic Health Post Health Government Private Number of

Center Hospital Hospital women
National 32.2 59.4 9.5 4.7 1 197
Livelihood
Agrarian 31.8 59.8 7.8 0.6 974
Pastoralist 56.2 37.6 5.7 0.8 223
Region
Tigray 9.5 83.2 6.1 1.0 180
Afar 32.4 45.6 22.0 0.0 42
Amhara 9.9 85.8 4.2 0.0 216
Oromia 38.8 50.8 9.8 0.6 299
Somali 94.4 0.0 5.6 0.0 14
Benishangul-Gumuz 21.5 66.6 8.5 0.8 113
SNNPR 54.2 36.6 8.0 1.1 225
Gambela 36.0 58.5 5.4 0.0 48
Harari 20.5 43.9 32.2 3.4 60
15-19 28.6 57.2 9.5 4.7 44
20-34 31.5 59.2 8.7 0.5 814
35-49 33.9 60.1 5.5 0.5 339
Live birth
≤2 38.9 60.1 7.8 1.2 408
3-4 25.3 66.8 7.5 0.4 362
5-7 38.7 52.6 8.5 0.2 347
≥8 43.4 51.5 5.0 0.0 80
Education
No formal education 31.0 62.5 6.4 0.0 685
Attended grade 1-4 41.5 50.7 5.7 2.0 205
Attended grade 5-8 31.4 59.5 8.2 0.8 210
Attended grade 9+ 14.3 60.5 25.2 0.0 97
Wealth quintile
Lowest 38.6 57.1 3.6 0.7 197
Lower 38.3 54.5 6.8 0.4 180
Middle 30.8 63.3 5.9 0.0 219
Higher 31.0 62.3 6.7 0.0 285
Highest 26.3 58.7 13.4 1.6 316
page- 387
Table 7-65. Information about danger signs, by source of information and background information
page-388
Told Danger sign
Vaginal gush of Severe
Vaginal bleeding Edema Fever Abdominal Pain Convulsion
fluid Headache
Background
characteristics
HEWs
HEWs
HEWs
HEWs
HEWs
HEWs
HEWs
Health
workers
Not told
Not told
Not told
Not told
Not told
Not told
Not told

# of pregnant women
Health workers
Health workers
Health workers
Health workers
Health workers
Health workers
Livelihood
Agrarian 19.0 17.4 63.6 14.8 15.0 70.2 12.2 10.7 76.9 12.4 13.7 73.9 11.8 13.1 75.1 12.6 13.4 74.0 9.2 9.9 80.9 2018
Pastoralist 9.5 4.1 86.4 7.4 4.7 87.9 6.0 4.4 89.6 6.9 3.8 89.3 7.6 3.4 89.0 6.7 3.7 89.5 4.9 3.3 91.8 1013
Region
Tigray 21.9 31.9 46.2 16.5 27.1 56.3 10.2 13.2 76.6 11.1 17.7 71.2 9.3 20.9 69.8 13.7 21.8 64.5 9.0 15.8 75.2 258
Afar 24.8 14.4 53.5 17.7 20.6 61.7 12.2 21.2 66.6 16.5 21.5 62.0 17.9 16.4 65.6 12.9 18.1 68.9 6.0 17.7 76.2 195
Amhara 16.9 15.1 68.0 16.8 20.7 62.3 10.0 9.3 80.7 14.4 18.6 66.9 10.4 19.1 70.2 12.3 18.0 69.7 8.6 14.2 77.2 397
Oromia 0.0 0.2 99.7 13.6 13.8 72.5 0.0 0.5 99.5 12.9 12.9 74.2 10.4 11.0 78.7 12.5 11.9 75.6 10.2 9.1 80.7 699
Somali 16.9 25.3 57.8 0.1 0.2 99.7 15.3 13.2 71.4 0.2 0.3 99.5 17.2 24.3 58.5 1.4 0.1 98.4 0.0 0.4 99.5 362
Benishangul-
21.8 8.6 69.6 11.4 22.3 66.2 17.0 6.0 77.0 15.7 20.5 63.7 16.1 8.9 75.0 20.8 23.5 55.7 16.3 14.3 69.3 202
Gumuz
SNNPR 21.8 8.6 69.6 14.8 7.9 77.3 1.6 25.8 72.6 9.3 8.6 82.1 16.1 8.9 75.0 12.4 9.0 78.5 7.6 5.1 87.2 516
Gambela 3.4 24.6 71.9 2.3 26.9 70.8 1.6 25.8 72.6 2.8 26.9 70.3 1.6 25.7 72.7 2.8 24.8 72.4 1.6 24.8 73.6 165
Harari 8.3 20.5 71.1 8.2 20.5 71.3 4.9 18.9 76.2 4.3 19.3 76.4 5.1 18.8 76.1 6.3 17.7 76.0 4.3 18.0 77.7 237
Table 7-66. Information about birth preparedness plan, by source of information
Percentage of mothers who delivered their last child during the last 5
Background years and who were informed about a birth preparedness plan by:
characteristics Other Health Unweighted
HEW Not told
workers Total
National 33.0 24.2 43.3 3 031
Livelihood
Agrarian 33.7 24.4 41.9 2 018
Pastoralist 18.2 7.7 74.0 1 013
Region
Tigray 35.9 37.8 26.2 258
Afar 27.7 21.6 50.6 195
Amhara 26.9 30.2 42.8 397
Oromia 30.6 24.7 44.7 699
Somali 6.0 0.3 93.7 362
Benishangul-
28.6 25.4 46.0 202
Gumuz
SNNPR 46.3 13.6 40.2 516
Gambela 8.4 26.1 65.5 165
Harari 9.5 32.0 58.5 237
Age
15-19 32.7 13.6 53.6 123
20-34 35.0 23.7 41.3 2 042
35-49 28.7 24.2 47.0 866
Wealth quintile
Lowest 32.4 16.3 51.2 605
Lower 35.1 18.1 46.8 586
Middle 33.9 20.5 45.6 585
Higher 34.6 25.4 40.0 650
Highest 28.7 36.1 35.2 605
Region.
page- 389
Table 7-67. Service providers for postnatal visits for most recent delivery during
the last 2 years
PNC given by:

HEWs HC Staff Unweighted total number
National 43.2 47.2 453
Livelihood
Agrarian 43.1 47.2 358
Region
Tigray 73.6 35.1 92
Afar 7.2 0 4
Amhara 36.2 66.9 70
Oromia 40.8 32.5 85
Somali 50 50 2
SNNPR 39.9 53.9 120
Gambela 39.5 56.5 16
Harari 26.4 53.2 27
Age
15-19 33.4 72.4 22
21-24 29.7 51.7 93
25-29 42.1 45.7 148
31-34 53.1 45.1 88
35-39 56.2 43.8 73
41-44 33.8 41.2 23
45-49 71.3 62.6 6
Education
No formal education 46.9 48.8 251
Attended grade 1-4 35.1 37 80
Attended grade 5-8 41 54.7 83
Attended grade 9+ 45.8 45.5 39
Marital Status
Currently married 43.3 47.2 421
Unmarried 41.4 46.3 32
Wealth quintile
Lowest 43.7 61.5 61
page-390
Lower 37.3 62 69
Middle 47.1 51.2 90
Higher 45.8 29.9 119
Highest 40.3 48.4 114
No of Children
1-2 Children 26.5 54.7 151
3-4 Children 60.1 40.5 143
5-7 Children 43.9 45.7 132
8+ Children 46.6 45.3 26
Abbreviations: HEW, Health Extension Worker; HC, health center; PNC, postnatal care; SNNPR,
Table 7-68. Place of PNC visit among mothers who had a PNC visit for their
most recent delivery during the last 2 years
Background Health
Home Health post Hospital Other
characteristic center
National 23.4 12.3 44.3 19.2 0.7
Livelihood
Agrarian 23.4 12.5 44 19.4 0.7
Pastoralist 24.0 8.2 53.3 14.5 0.0
Education
No formal education 25.1 13.8 40.1 19.8 1.2
Attended grade 1-4 10.0 16.1 49.6 24.2 0.0
Attended grade 5-8 41.6 7.6 41.6 9.1 0.1
Attended grade 9+ 0.1 0.0 74.8 25.1 0.0
Wealth quintile
Lowest 18.2 15.9 58.1 7.6 0.1
Lower 8.2 8 62.3 16.1 5.7
Middle 34.7 3.8 40.6 20.9 0.1
Higher 19.3 19 41.6 20.1 0
Highest 34.1 7.8 31.6 26.4 0.1
Abbreviation: PNC, postnatal care.
page- 391
CHILD HEALTH AND VACCINATION
According to evidence from documents and interviews at health facilities and

communities, HPs are the primary sources of vaccination services for rural
communities. In addition, HPs also play a significant role as sources of advice
and treatment for common childhood illnesses. With respect to treatment and
advice for children under 5 with symptoms of ARI, medical help was primarily
received at HCs (48.3%), followed by HPs or HEWs, which were the source of
services 24.4% of the time (Figure 7-19).
Figure 7-19: Source of advice or treatment for children with symptoms of ARI
Abbreviations: ARI, acute respiratory infection; HEW, Health Extension Worker.
Advice and treatment for children with diarrhea were mainly provided at HPs
(21.8%), followed by HCs (17.9%) (Figure 7-20).
page-392
Figure 7-20: Source of advice or treatment for children with diarrhea
7.3.9 Quality of HEP services: evidence from family health quality

of care indicators
The quality of care provided through the HEP was assessed for selected family
health interventions provided through the HEP. Whenever data were available,
findings were compared between HPs and HCs to make interpretation easier.
This sub-section presents findings on the quality of ANC and the continuity of
maternal health services and vaccination services.
COMPONENTS OF ANC SERVICE
Standard guidelines for ANC in Ethiopia emphasize that every pregnant

mother should receive, from a skilled provider, ANC that includes a thorough
physical examination, blood tests for infection screening and anemia, a urine
test, TT injections, iron and folate supplements, deworming medications, blood
pressure measurement, and education about danger signs in pregnancy.
In general, important components of ANC were provided to a sub-optimal

proportion of women who had ANC visits during their most recent pregnancy.
The provision of the essential components of ANC ranged from 15.3% for
deworming to 77.2% for iron tablet supplementation.
The national guidelines on ANC recommend that the first and fourth visits of
ANC be handled at an HC, while the second and third visits can be handled
at the HP level. Compliance with this recommendation was sub-optimal. Some
page- 393
women had all their ANC visits at the HP, others had all visits at the HC, and
some had a mixture of visits at both the HP and HC. Having all ANC visits
at the HP or some mixed with visits at the HC results in a relatively lower
quality of service than having all ANC visits at the HC. When the 3 groups are
compared, women who had all visits at the HC were more likely to receive most
of the ANC components. Mixing visits also failed to address the quality gap
between the ANC services provided by HPs and HCs (Figure 7-21).
Figure 7-21: Components of ANC received for the last pregnancy, by provider
category
Abbreviations: ANC, antenatal care; HC, health center; HP, health post; TT, tetanus-toxoid.
page-394
CONTINUITY OF MATERNAL AND NEWBORN CARE
The first ANC visit is considered an entry point to a continuum of care

involving multiple visits of antenatal care, health facility delivery, PNC, and
child vaccination. This study found that, in general, there is a high dropout
across the continuum of care. Overall, coverage of ANC-1 is 85.7%, while that
of ANC-4 is only 48.3%, and only some women who attended ANC-4 delivered
in a health facility.
A comparison of the continuum of care between mothers who received ANC

from the 3 categories of service providers (only HP, only HC, or mixed) was
performed using a sub-set of the data for which both ANC and data related
to place of delivery were available. Findings showed that the quality of service
delivery at HPs is more compromised than that received at HCs. Health facility
delivery was lowest among women with no ANC at all. Mothers who attended
all their ANC visits at an HP were less likely to deliver in a health facility
compared to those who attended all their ANC visits at HCs (Figure 7-22).
Figure 7-22: Place of delivery, by place of ANC visit
Abbreviations: ANC, antenatal care; HP, health post; HC, health center.
page- 395
QUALITY OF VACCINATION SERVICES
The rate of Penta 3 and measles vaccination is lower than that of Penta 1 or
BCG. The vaccine coverage discrepancy is higher than the acceptable level
of 10%. Penta 1 coverage is 74.7%, and Penta 3 coverage is 50.4%, so the
resulting discrepancy is 24.3%. BCG coverage is 79.3%, and measles vaccination
coverage is 47.8%, so its discrepancy is 31.5%. Of the children who received
BCG, 11.6% had no scar.
page-396
The Health Extension Program
The Health Extension Program has been a
major component of the Ethiopian health
system and is commonly claimed to be a major
contributor to gains in health outcomes in the
country.
page- 397
Chapter 7
CHAPTER 8
HEP Service
Delivery
Outcomes
page-398
8. Health Extension Program Service
CHAPTER 8
Delivery Outcomes
The HEP has been a major component of the Ethiopian health system and is
commonly claimed to be a major contributor to gains in health outcomes in the
country. The contribution of the HEP to behavioral change at the household
level was assessed by examining the validity of assumptions involved in the
program logic model presented in Figure 1-4, (in part 1 section I). A theory-
based approach to contribution analyses was attempted. This section synthesizes
the evidence from the existing literature and cross-sectional household survey
data from the HEP National Assessment, with the purpose of generating
plausible evidence regarding the contribution of the HEP. This section includes
an assessment of trends in morbidity, mortality, and health behavior indicators,
followed by a systematic review with meta-analyses on the effectiveness of the
HEP in improving health outcomes. Finally, an association is presented between
the intensity of implementation of the HEP and household-level outcomes,
based on data from the household survey of the HEP National Assessment.
8.1 Trends in health indicators in Ethiopia

To describe the trends in health and health-related indicators across time and
their relationship with key policy initiatives relevant to health development in
Ethiopia, we extracted indicator data from 4 rounds of Demographic Health
Surveys (DHS) conducted since 2000. We compared the values and trends of
indicators using these data. The results are presented for the broader domains
of mortality, morbidity, and health service use.
8.1.1 Trends in mortality
MATERNAL MORTALITY
MMR in Ethiopia was estimated at 897 per 100 000 live births in 2000. This
decreased to 743 in 2005, 523 in 2010, and an estimated 353 in 2015. The
average annual rate of reduction in MMR between 2000 and 2015 was
page- 399
HEP Service Delivery Outcomes
between 7% and 8%. The time of this reduction included the period of the roll-
out of the national HEP. The trend in MMR over the past 30 years is shown in
Figure 8-1.56
Figure 8-1: Trends in Maternal Mortality, 1990-2015, Ethiopia
CHILDHOOD MORTALITY
As seen in Figure 8-2, during the time horizon of analysis, there was a declining
trend of all childhood mortality indicators. The rate of decline however, differed
for the different mortality indicators. The highest overall decline was in child
mortality rate (74% decline), followed by under-5 mortality (60% decline) and
infant mortality (50% decline). While there was at least a 20% decline between
the DHS years for all other childhood mortality indicators, there was only a 5%
decline in neonatal mortality between 2005 and 2011.
page-400
Figure 8-2: Trends in childhood mortality indicators, 2000-2016, Ethiopia
MALARIA DEATHS
As shown in Figure 8-3, there has been a declining trend in the number of
malaria deaths since 2004. Apart from irregular surges, the overall decline,
despite an increase in the population size, could be considered a success in
malaria prevention and control. There could be, however, some data quality
limitations associated with the overall HIS in Ethiopia, which may affect the
trends.
page- 401
Figure 8-3: Malaria deaths, 2002-2015, Ethiopia
8.1.2 Trends in morbidity
MATERNAL MORBIDITY
One proxy indicator of maternal morbidity is maternal nutritional status.

Nearly one third (30%) of mothers were underweight in 2000. This declined to
27% by 2005 and 2011. Later, it declined again to 22% by 2016. Although the
differences could be due to sampling variation, the 5-percentage point decline
between 2011 and 2016 may reflect the contribution of the HEP.
About a fourth (26%) of the women surveyed had anemia in 2005. Based on the
DHS report, this had declined to 16% by 2011. The 10-percentage point decline
occurred during the period of the HEP’s roll-out. The 2016 DHS, however,
showed a higher prevalence of anemia among women (23%).
page-402
CHILD MORBIDITY
One key indicator of childhood mortality is the prevalence of malnutrition in

children. There has been a remarkable decline in the prevalence of childhood
malnutrition during the study period. As seen in Figure 8-3, there was a
20-percentage point decline in the prevalence of stunting and underweight.
This decline likely reflects a significant contribution from the HEP.
Figure 8-4: Nutritional status, children under 5, 2000-2016, Ethiopia
Some of the other indicators of childhood morbidity show mixed results. For
instance, the prevalence of severe anemia in children dropped from about
54% to 44% between 2005 and 2011, but then increased to 56% by 2016. The
prevalence of low birth weight remained stable, at about 13%, during the study
period.
MALARIA MORBIDITY
As shown in Figure 8-4, the number of confirmed malaria cases increased from
2008 to 2013. This likely reflects improvements in the availability and accessibility
of diagnostic services and improvements in the HIS instead of a true increase
in risk of malaria. After 2013, the number of confirmed cases began to decline,
most likely reflecting a true decrease in the number of malaria cases.
page- 403
Figure 8-5: Trends in number of confirmed malaria cases, 2002-2015, Ethiopia.
8.1.3 Trends in Health Service Use
MATERNAL HEALTH SERVICE
The following maternal health services are covered under the HEP: ANC, skilled
birth attendance (SBA) or assisted and clean delivery, and PNC. As seen in
the figure below, ANC coverage increased from 27% in 2000 to 62% in 2016.
Almost all of this increase has been since 2005, suggesting that this may reflect
a major contribution of the HEP at the national level. It should also be noted
that the ANC has remained above the annual target of 90% since 2011, and
this is likely to be linked, in part, to increased access to ANC as a result of the
HEP.
Similarly, there was a 17% increase in SBA coverage between 2011 and 2016.
This is 3 times the increase in SBA coverage between 2005 and 2011. This
probably reflects the full-scale implementation of the HEP since 2010. About 3
out of every 4 pregnant women, however, are still not receiving SBA services;
the next phase of the health sector’s strategic plan should pay special attention
to this large gap. Note that the proportion of mothers who delivered in a
health facility shows a similar pattern (Figure 8-5).
page-404
ANC coverage remained very low throughout the study period. Until 2011,
about 1 out of every 10 women who delivered received ANC services. The
increase in ANC coverage between 2011 and 2016 was two-fold, and this likely
reflects the contribution of the expanded HEP. There remains, however, an 80%
gap in ANC coverage that should be addressed in the coming years (Figure
8-5).
Figure 8-6: Coverage of key maternal health services, Ethiopia
FAMILY PLANNING SERVICES
For countries like Ethiopia, where economic growth and population growth are in
great disparity, improving the coverage of FP services is critical. During the study
period, the coverage of FP demand satisfied with a modern method increased
from 14% in 2000 to 61% in 2016. The highest increase, about 22 percentage
points, was between 2005 and 2011, the period of the HEP’s incremental roll-
out. This has contributed to a decrease in the Total Fertility Rate (TFR) from
5.5 in 2000 to 4.6 children per women in 2016 (2.3 in urban and 5.2 in rural
areas). The rural-urban disparity is evident in all fertility indicators.
page- 405
Table 8-1. Fertility status indicators, 2000-2016
2000 2005 2011 2016

Married women currently using any method of
8.1 14.7 28.6 35.9
contraception
Desire for more children: want to have another soon 22.3 16.1 16.9 17.5
Unmet need for family planning 36.6 36.1 26.3 22.3
Demand for family planning satisfied by modern methods 14.2 27.4 49.8 60.6
Total fertility rate, age 15-49 5.5 5.4 4.8 4.6
Mean ideal number of children for all women 5.3 4.5 4.3 4.5
CHILD HEALTH SERVICES
Another key intended outcome of the HEP is improvement in coverage of child

health services. In this review, we tracked 4 key indicators of maternal health
services: immunization coverage, coverage of diarrhea treatment, coverage of
ARI treatment, and exclusive breastfeeding. The most recent DHS data show
coverage levels increasing to 44% and 31% for the treatment of diarrhea and
ARI by health providers, respectively. The proportion of children who were
exclusively breastfed for the first 4 months of life has remained stable at about
60% during the study period. By contrast, the proportion of children who
received all the recommended vaccinations by their first birthday increased
from 14% in 2000 to 39% in 2011. A significant proportion of this change
occurred between 2011 and 2016. Details of the coverage of key child health
services is shown in Table 8-2.
Table 8-2. Coverage of key child health services (in percent)
2000 2005 2011 2016
Diarrhea (taken to provider) 13.3 22.2 31.8 44
ARI (taken to provider) 15.8 18.7 27 31.3

Exclusive breastfeeding for 4
62.3 56.8 61.8 58
months
Received all 8 basic
14.3 20.4 24.3 38.5
vaccinations
Abbreviation: ARI, acute respiratory infection.
page-406
There was a remarkable improvement in the coverage of water, sanitation, and

hygiene (WaSH) indicators between 2000 and 2016. As shown in Figure 8-7,
the proportion of households using an improved water source increased by 40
percentage points, from 25% to 65%. During the same period, the proportion of
households using toilet facilities increased by 50 percentage points, from 18% to
68%. The increment was linear and could be related to variety of factors, both
economic and social, as well as health interventions.
Figure 8-7: Coverage of WaSH Facilities
TUBERCULOSIS TREATMENT COVERAGE
Analysis of the trends in the coverage of TB treatment shows a steady increase

between 2002 and 2011. The coverage levels declined between 2011 and 2014
before increasing again.The trend of this indicator is shown in Figure 8-8.
page- 407
Figure 8-8: Trends in Tuberculosis treatment coverage

Source: WHO database
8.2 Effectiveness of the Health Extension Program:

Evidence from a systematic review of previous
studies
8.2.1 Impact of the HEP on use of family health services
A total of 29 articles have reported the effect of the HEP on maternal

healthcare services, although they measured different health outcomes and
different versions of “health extension package” exposure. A dose-response
study reported that, for every unit increase in the program intensity score, the
odds of receiving ANC increased by 1.13 times (95% CI, 1.03-1.23). Similarly,
the odds of birth preparedness increased by 1.31 times (1.19-1.44), the odds of
receiving PNC increased by 1.60 times (1.34-1.91), and the odds of initiating
breastfeeding immediately after birth increased by 1.10 times (1.02-1.20).42
Medhanyie A et al. also reported that, on average, within 6 months, an HEW
assisted in 5.8 births.57 Afewrok et al. reported that a visit by an HEW during
pregnancy predicted a PNC visit during the first 3 days after delivery and
page-408
at least 4 ANC visits.58 A study from West Gojjam of Ethiopia revealed that
HEWs visited 52.7% (95% CI, 50.0-55.4%) of households, and 78.5% (95% CI,
76.2-80.7%) of mothers also visited HPs during the 1 year preceding the survey.
Mothers who had frequent household visits by HEWs were 1.289 more likely
to visit HPs (AOR=1.289, 95% CI, 1.028-1.826) than were mothers who did not
get frequent visits.59 Three years after their graduation, mothers from model
households were 2.4 times more likely to visit an HP than were mothers from
non-model households.60 Medhanyie et al. also reported that women who had
been working toward graduation or had graduated as a model family (OR,
2.13) were more likely to demonstrate good use of maternal health services.61
Over the 10 years since the introduction of the HEP (between July 2002 and
June 2012), an increase in the use of ANC, delivery and PNC was observed in
Tigray.62
A study evaluating the effect of Community Maternal and Newborn Health

family meetings by HEWs—to build skills and care-seeking behaviors among
pregnant women and family caregivers―resulted in a 151% increase in care
completeness between baseline and end line. In addition, they found that
women who participated in 2 or more meetings had more complete care than
did women who participated in fewer than 2 meetings (89% vs. 76% of care
elements; P<.001). Furthermore, women who had received any ANC were
nearly 3 times more likely to have used a skilled provider or HEW as a birth
attendant. Women who had attended 2 or more meetings with family members
were over 5 times more likely to have used these providers as birth attendants
than were women who had not received ANC or who had attended fewer than
2 meetings (odds ratio, 5.19; 95% CI, 2.88-9.36; P<.001).
Regarding contraceptive use, mothers from households that fully benefitted from
the HEP (model households) were 3.97 times more likely to use contraceptives
compared with mothers from non-model households (adjusted odds ratio, 3.97;
95% CI, 3.01-5.23). Model household status contributed to 29.3% of the increase
in current contraceptive use.63 In one study, about 84.7% of study participants
agreed that HEWs were the primary source of information about Implanon.63,64
A study from the SNNPR showed that mothers who had received information
about MCH services from HEWs were significantly associated (AOR 2.09; 95%
CI 1.06-4.14) with good MCH service use status.65 A very recent article, from
2019, reported that focused antenatal care was used by 24.36% of women and
HEWs accounted for use by 46.47% of these mothers, which suggests a major
contribution by HEW to the use of these services. Furthermore, mothers who
page- 409
received a PNC home visit by an HEW were 1.35 times more likely to have
good PNC practices than unvisited mothers (adjusted odds ratio [AOR] 1.35,
95% CI [1, 1.71]).66
Some studies have shown that the contribution of the HEP is lower than
the expectations established by the MoH. Karim et al. found no association
between the program and SBA, nor with some of the other newborn healthcare
indicators.42 Afewrok et al. reported that the effect of a visit by an HEW on
institutional delivery was unclear.58 Another study found no association between
the contribution of HEWs and the level of health facility delivery or PNC.61 One
study published in 2017 reported that the coverage of ANC (4 ANC services
in one district, Dale, in southern Ethiopia) was lower than national levels, that
input from the HEWs was unsatisfactory, and that the number of home visits
was inadequate.65
A few studies have examined the role of the WDAs. With the support of
WDAs, HEWs have contributed to an increase in the rate of SBA by calling
ambulances to transfer women to HCs either before their estimated date of
delivery or when labor starts at home.67 WDA density was also associated with
better service. Kebeles with WDAs serving no more than 40 households had a
7% (95% CI 2-13) higher contraceptive prevalence rate, an 11% (5-17) higher
coverage of 4 or more ANC visits, and a 9% higher coverage of institutional
deliveries compared with kebeles with 1 active WDA team leader for 60 or more
households; this suggests that the WDA strategy has contributed to improving
coverage of healthcare services.60
Because studies have reported multiple exposures and outcomes (i.e., they are
heterogeneous), and assuming these outcomes are correlated and defined as
“maternal health services use behaviors,” we did further meta-analysis (Figure
8-6). A random effects model predicts that exposure to the HEP improves the use
of maternal healthcare services (2.46, 95% CI [1.93, 3.14]), although significantly
larger heterogeneity was observed among the studies. Visual inspection of the
funnel plots provides little evidence for publication bias; Begg’s and Eggers
tests did not provide any statistical evidence for publication bias.
page-410
Figure 8-9. A random effect model-analysis of the effects of the HEP on

maternal healthcare service use
“Maternal healthcare services use” had multiple outcomes and was defined if
the study reported the use of any of the maternal healthcare services (FP, ANC,
SBA, PNC), which we considered “maternal healthcare services use.” There are
also multiple exposures to the definition of the HEP. For instance, if the study
reported exposure as “visited by HEWs,” “exposed to model family training,”
“visited the HP,” and “received counseling by HEWs,” they were considered an
exposure to the Health Extension Program.
page- 411
8.2.2 Disease prevention and control
Admassie et al. reported that, although the number of regular monthly HEW
visits to each assigned household was high, study participants’ overall knowledge
regarding malaria did not improve.This finding is in contrast to study participants’
high levels of knowledge regarding HIV/AIDS.68 The proportion of children and
women using ITNs for malaria protection was significantly higher in program
villages than in non-program villages.69 A randomized controlled community
trial also reported that the involvement of HEWs in sputum collection and
treatment improved smear-positive case detection and treatment success rate,
possibly because of improved service access.70 Proactive contact with HEWs
was associated with HIV testing; being visited at home by an HEW, however,
did not have a noteworthy effect on HIV testing.71
Ashenafi et al. reported that not being a model family in the HEP is an
independent predictor of childhood diarrhea (OR: 4.50, 95% CI [2.52, 8.03]).72
Higher-intensity implementation of the HEP and other accessibility factors were
associated with higher levels of care-seeking for childhood illnesses.73
In 2017, Mathewos et al. concluded that HEWs’ provision of management of

possible serious bacterial infection (PSBI) for newborns at HPs was associated
with improved implementation of the existing policy for community-based
newborn care. Adding PSBI management at the community level was estimated
to reduce neonatal mortality after day 1 by 17%, translating to a cost averted
per DALY of $223. At 47% of per-capita GDP, this is a highly cost-effective
intervention by WHO standards. In this context, strengthening the existing
HEW package by adding PSBI management would be cost-effective.74
A 2017 cross-sectional study reported that 66.1% and 53.4% of the study’s
respondents, respectively, had not received a community TB health education
message, nor a home visit from an HEW. Even though referral from HEWs
was significantly associated with being a model household (AOR=21.2, 95%
CI=9.5-47.3) and a home visit from an HEW (AOR=2.8, 95% CI 1.2-9.6), the
qualitative data from the same study reported that HEWs’ involvement in
referring patients with presumptive TB was limited. Communities reported
having low confidence in HEWs, that TB services were inaccessible, and that
HEWs lacked in-service training. The authors concluded that the contribution
page-412
of HEWs in identifying and referring presumptive TB cases was limited in the

Somali pastoralist region.75
8.2.3 Hygiene and environmental sanitation
HEP exposure is associated with the improved regular use of pit latrines, but
the actual difference in use rates with comparison groups remains low (1.1%).
Respondents who were advised by an HEW or development agent were more
likely to have built a latrine.76 The decrease in open defecation associated with
teacher-facilitated Community-Led Total Sanitation (CLTS) was 8.2% lower
than conventional CLTS (including HEWs; p=0.048). Follow-up after 1 year,
however, found that the effectiveness was unclear and/or that sustainability
remained poor.77
A pre-post analysis of Performance Review and Clinical Mentoring Meeting

(PRCMM) interventions - an approach to improving and sustaining HEWs’
skills and performance - demonstrated that the improvement of pneumonia
management was consistently successful: 54.1% (95% CI, 47.7%-60.5%)
6 months prior to intervention and 78.2% (73.9%-82.5%) 6 months post-
intervention. Similar findings were observed for malaria (50.8% [42.9%-58.7%]
vs. 78.9% [73.4%-84.4%]) and diarrhea (33.7% [27.9%-39.5%] vs. 70.0% [64.7%-
75.3%]). Furthermore, the caseload increased significantly after PRCMM: 6.6
(95% CI, 5.9-7.3) cases/HP/month before intervention and 9.2 (8.5-9.9) after
intervention.78 The integrated case management of childhood illnesses (iCCM)
- designed to build the capacity of Ethiopia’s existing cadre of HEWs to assess,
classify, and treat pneumonia‐helped HEWs engage in curative health services,
gain trust from community members, and increase the acceptance of HEWs’
messages related to preventive healthcare.79
8.2.4 The role of HEWs in the provision of health information and

services
In a 2013, study in Jimma zone, 169 (51.7%) of respondents reported having

an interaction with an HEW during the 1 year prior to the survey, while 271
(71.5%) reported having received a visit from an HEW during the same period,
and 298 (78.6%) reported receiving information from the HEP.80 In the rural
page- 413
Sidama zone, 12.4% of mothers and their newborns were visited by an HEW
during the first month after birth. Of all households who had a history of HEW
visits, 55.2% had a single visit. Of these, only 26.2% of the first visits were within
the first 24 hours after birth.66
Throughout the country, 37% of FP messages were disseminated by HEWs. In

addition, 35% of outpatient services, 24% of contraceptives, 17% of ANC, and
17% of fever treatment services were provided by HEWs.24 The training and
graduation of model families began in 2006/07 and peaked between 2008/09
and 2010/11, by which point more than 12 million households had graduated
as model families, representing 70% of eligible households. The percentage
of model families varies greatly by region, however, ranging between 85% in
Oromia to 0% in Gambela; coverage of model families is much higher in rural
regions than in urban areas or pastoral regions. Because of the HEP, coverage
of PHC, the EPI, contraceptive acceptance, and ANC all increased, from 76.9%
to 90.0%, 76.8% to 81.6%, 37.9 to 56.2, and 50.4 to 67.7, respectively, over a
5-year period, during which the prevalence of HIV fell from 3.2% to 2.1%.81 In
2009, the HEP was associated with significant increases in the proportion of
children fully and individually vaccinated against tuberculosis, polio, diphtheria–
pertussis–tetanus, and measles.69
8.3 Association between intensity of Health

Extension Program implementation and
household-level adoption of health behaviors
As a short-term result, the adequate implementation of the HEP is expected to
bring about changes in household behavior in areas targeted by the program.
These include family health, hygiene and sanitation, and disease prevention
andcontrol practices. Progress toward full implementation of these packages
wasdetermined for each HP by averaging the proportion of target households
thatwere found to meet the criteria for which they were eligible. The intensity of
HEPimplementation was measured for each kebele as the proportion of
households that interacted with HEWs through home visits, HP visits, and
outreach.
The association between the average progress of households and the intensity of
implementation of the HEP was then assessed through multiple linear regressions.
National
page-414Assessment of
Separate regression models were fit for agrarian and pastoralist settings. The
findings showed that progress in the household-level implementation of the HEP
was significantly associated with home visits in agrarian settings and HP visits
in pastoralist settings. A 10% increase in the proportion of households reached
through home visits was associated with a 19% increase in household progress
toward full implementation of the HEP at the household level. Similarly, a
10% increase in the proportion of pastoralist households who had interactions
with HEWs through HP visits was associated with a 16% increase in household
progress toward full implementation of the HEP at the household level.
Table 8-3. Association between intensity of HEP implementation and household-

level behavior
Agrarian Settings Pastoralist Settings

95% CI of B 95% CI of B
B P B P
LL UL LL UL
Proportion of households
reached through home 0.186 0.109 0.263 0.000 0.017 -0.190 0.223 0.872
visit
-0.056 -0.146 0.033 0.215 0.156 0.001 0.311 0.048
reached through HP visit
0.018 -0.088 0.123 0.738 -0.042 -0.304 0.221 0.751
reached through outreach
Median age of women 0.131 -0.139 0.400 0.340 -0.066 -0.887 0.756 0.873
with a woman having -0.067 -0.174 0.041 0.222 0.361 0.096 0.625 0.008
some formal education
in the medium, higher, or 0.134 0.069 0.199 0.000 -0.106 -0.205 -0.006 0.038
highest wealth quintiles
Constant 37.0 25.2 48.8 0.000 35.9 7.4 64.4 0.015
Abbreviations: HP, health post; LL, Lower Limit; UL, Upper Limit
Bivariate analyses examining the association between receiving supervision

from different sources and the average progress of households toward the
adoption of HEP-related behaviors showed that receiving supervision from
any source was associated with higher progress in household-level behavioral
change. Variability in progress varied, albeit only to a limited extent, by the
type of supervisor (Figure 8-9).
page- 415
Figure 8-10. Mean progress toward full implementation of the HEP at the
household level
Abbreviation: HEP, Health Extension Program; HH, household.
page-416
page- 417
CHAPTER 9
Governance,
Leadership, and
Management of
the HEP
page-418
9. Governance, Leadership,
CHAPTER 9
and Management of the Health
Extension Program
INTRODUCTION
This chapter briefly describes the general governance structure of the health
system and provides the results of the study from the perspectives of policy
environment, the planning process, the decentralized autonomy of HEWs, the
commitment of managers, coordination of HEP activities, and collaboration
with other sectors. The other 2 critical aspects of governance—community
engagement and M&E—are detailed in other chapters.
The leadership and governance of health systems, also called stewardship, is

arguably the most complex but most critical building block of any health system.
It concerns the role of the government in health and its relation to other actors
whose activities affect health. This involves overseeing and guiding the whole
health system, both private and public, in order to protect the public interest.
It requires both political and technical action because it involves reconciling
competing demands for limited resources, in changing circumstances: for
example, with rising expectations, more pluralistic societies, decentralization,
and/or a growing private sector.
Good governance is imperative for the effective planning, implementation,

monitoring, and evaluation of the HEP. Governance embraces the entire
framework of decision-making process at the central, regional, woreda, and
community/kebele levels. Hence, details of evidence from the literature and
the results of the quantitative and qualitative assessments are described in the
following sections:
• The existence of an enabling policy environment;
• The engagement of leadership in the planning, monitoring, and
evaluation of the HEP;
• The decentralization, collaboration, and coordination of HEP; and
• The availability and functionality of standards and guidelines.
page- 419
Governance, Leadership, and Management of the
KEY FINDINGS
The HEP structure that extends from the federal to the community level is well
designed and accepted. Governance and leadership roles, however, particularly
at the woreda and kebele levels, are not functioning well.
Achievements:
• The rapid and massive construction of HPs (over 17 000) that now cover
almost every agrarian and pastoralist village and HCs (over 3500) are
well linked and operate as a PHCU health delivery system in the country;
• Full-time, salaried HEWs (around 40 000) who are now the key drivers
for the HEP within the public system, have been trained and deployed,
and 6 million community-led Health Development Armies have been
mobilized;
• The government’s commitment and leadership has resulted in the
successful mobilization of resources within and outside the country to
finance such a huge community-based program in the country;
• The HEP has a well-structured 4-level governance system comprising
federal, regional, woreda, and kebele levels in which multi-sectorial
partners are all supporting the HEP;
• The HEP is seamlessly integrated into the PHCU, with enabling policies
and guidelines available;
• The HEP is being used as a platform for recent gains in the maternal,
neonatal, and child morbidities and mortalities.
CHALLENGES:
• Woreda and kebele administrations pay minimal attention to the HEP;

they lack the technical and administrative capacity to lead and follow
up on HEP activities. Due to this difficulty, some suggest assigning a
salaried focal person at the kebele level;
• HEP coordinators at HCs and in some woredas are non-technical, and
this has profound implications for the depth and breadth of technical
support HPs and HEWs are receiving;
• The allocation of a limited or no budget for HEP M&E activities. In
addition, some HEP services (i.e., vaccination, FP, or commodities) are
donor- or project-dependent, with a limited life span, creating doubt on
the future sustainability of the program;
• There is less integration and less of a multi-sectorial approach to HEP
page-420
activities at the kebele and woreda levels, which remains a great challenge
to the HEP’s implementation. HEWs parallel reporting and overlapping
responsibilities with local actors.
• The findings from this study suggest that assigning HEWs to non-health-
related activities (i.e., tax collection) was seen as an indication of neglect
of or less commitment to the HEWs by kebele leaders;
• The roles, responsibilities, and links between WDAs and HEWs were
properly designed, but their actual functioning is in many places unclear:
• The WDA approach is found to be less feasible in pastoralist
setting, and its members found it difficult to work seamlessly
due to the temporary settlement of the communities who move
seasonally, as well as other socio-cultural incompatibilities.;
• Community acceptance of WDAs and HDAs is very low for a
number of reasons, including community perceptions about the
organization, which is seen as having a political agenda.
9.1 Overall governance and evolution of the health

sector
This section describes the overall health governance structure and how it
has evolved over time, as described in various policy documents. Ethiopia’s
healthcare system has had an established and well-structured governance
system since 1998; it was put in place for the implementation of HSDP I, under
which its governance structure had 4 layers:
• The Central Joint Steering Committee (CJSC), a consultative forum
between the MoH and health development partners (HDPs). The
Ministry of Health chairs the CJSC, which is technically supported
by the group formed by the MoH and partners called the Joint
Core Coordinating Committee. The Joint Consultative Meeting
and Annual Review Meeting (ARM) are the tasks assigned at the
federal level;
• The Regional Joint Steering Committee at the regional level;
• The woreda joint steering committee (WJSC) at the woreda level;
and
• The kebele HIV and health committees at the community level.5
page- 421
These committees at all levels of the health system form the governance
system; they are connected to each other through regular reporting, and all are
tasked with serving as coordination mechanisms for planning, implementing,
monitoring, and evaluating health programs and projects in the health sector.
The evaluation of HSDP I and II, however, found that the subnational steering
committees were not functioning properly; in many regions, committees were
not established at all. Where committees existed, they met only irregularly and
were not functioning as anticipated. While trying to prop up the subnational
governance committees without changing their structure, the federal-level
governance structure established by HSDP I was later modified at the federal
level. This resulted in the establishment of the MoH-RHB Joint Steering
Committee, which has essentially ended the CJSC and reorganized the JCM
into the Joint Consultative Forum (JCF). It has been claimed that the new
restructuring of the governance system enabled the MoH, RHB, and HDPs to
enjoy a progressively improved dialogue and deliberation in the health sector.
As part of the MoH’s process of continuous system development, huge steps

were taken in transferring some responsibilities, authority, power, and resources
to the local level, which has created opportunities for effective governance
there. To improve the governance structure at the points of service delivery, the
government has introduced facility governance boards (HCs and hospitals)
comprising various relevant bodies in a given local community.6
The HEP has multiple stakeholders at the federal (i.e., MoH, RHBs, HDPs,
Ministry of Education [MoE], Ministry of Agriculture [MoA]), regional, woreda,
and kebele levels whose interest and concerns are to be addressed through
well-established and functional governance mechanisms. These are detailed in
the context of the Ethiopian Health System section (Part – 1, Section – 1, under
1.4.3).
9.2 Existence of an enabling policy environment

This section examines the availability of both broader and HEP-specific policies
and guidelines. The HEP has enjoyed strong policy support throughout its
evolution, as stipulated in the National Health Policy and subsequent health
strategies. This support has also been rooted in the broader national and global
policy contexts. As stipulated in the GTPII (2015/16 to 2019/20),82 Ethiopia
page-422
is advancing to becoming a middle-income country. Within this broader

national goal, achieving universal PHC is a priority agenda item of the country
between 2016 and 2035. As the country transitions, the health sector intends
to continue to invest in primary care to advance the overall health and well-
being of the population and serve the priority health needs of the majority
of its people. This healthcare goal within the context of differences in disease
profiles, population growth, technological advancements for inputs, and socio-
economic considerations, has continued since 2003 to form an enabling policy
environment in the implementation of the HEP (HSDP II, HSDP III, HSDP IV,
HSTP, EHSP).
Apart from the availability of the broader country-wide health and development
policies and strategies that facilitate implementation of HEP services, a number
of other healthcare strategy documents support these broader goals and policy
documents on reproductive health, infant and newborn care, maternal and
child health, nutrition, TB and HIV, NCDs (including mental health, health
education, and communication), and other quality-of-care strategic documents
that foster an enabling policy environment for the implementation and scale-
up of the HEP (HSDP I–IV, HSTP, EHSP).2-6,83
The major evidence for the enabling broad policies and strategies that have
greatly helped the accelerated expansion of the HEP is:
1) The rapid construction of HPs, which have reached over 17 000
patients since the establishment of the HEP, was the key indicator
of the government’s commitment. HPs are the operational level of
the HEP and are linked to HCs as part of PHCUs, which themselves
have increased significantly, now reaching over 3764;
2) The massive human resource development in the form of HEW
training and deployment. This is essentially the institutionalization
of the community-based health system, thus phasing out the
voluntary-based CHA and TBA system. Ethiopia has trained and
deployed over 40 000 HEWs (2 per HP) for the management
of the HEP. HEWs are the key drivers of the HEP and form the
foundation of the pyramid of the public system25,29
3) The government’s success story in the mobilization of resources
from both within and outside national borders. The government
has spurred the resource mobilization initiative through its well-
page- 423
intended strategic motto of “all roads lead to the Health Extension

Program!”81 Significant resources have been channeled from
partners to pay for medical equipment, drugs, supplies, and pre-
and in-service training and teaching materials in support of the
HEP.22 For example, the HSTP 5-year healthcare financing plan
stipulated that, of the total healthcare financing requirement in the
base scenario, as much as 55% of the total has been allocated for
primary-level healthcare, and of this total, as much as 32% was to
be used exclusively for the financing of the HP-based HEP;6 and
4) The HEP’s well-structured 4-level governance from the federal
down to the kebele level, which clearly shows its multi-sectorial
nature involving various stakeholders, including the community’s
genuine participation in the planning, implementation, monitoring,
and evaluation of the HEP.30
Overall, the HEP in Ethiopia has become the foundation of the PHC system,
and PHC in turn is the foundation of Ethiopia’s health system.21 The HEP in
Ethiopia is owned and led by the government, which increases commitment
from the Ethiopian government in all aspects, including its policy initiatives for
supporting the HEP. This support has yielded results that could be attributable
to the implementation of the HEP at the community level, and perhaps this will
help ensure its future sustainability.
The importance of HEP-specific guidelines cannot be overstated for a nationwide

program like the HEP, which has well-organized sets of comprehensive
implementation guidelines that adequately describe the packages, service
delivery modalities, and roles and responsibilities of HEWs, HDAs, facilities
(HPs and HCs), and the different levels of health management and governance
structures of the healthcare system of the country (from the kebele to the
federal level). The HEP’s services are also well defined and articulated in each
package. The health sector strategic documents that have been developed
and implemented, including the current ones (HSTP and EHSP), are other
operational guiding documents available for use by implementers at all levels of
the HEP. For each package, HEWs are trained and supplied with the different
job resources that facilitate their static and outreach health education and
communication activities.
In line with this, the findings from the quantitative assessment show that iCCM
(75%) was the most frequently available guideline, followed by the family health
page-424
(65%) and CNBC (64%) guidelines, while CHIS (27%), PNC (30%), and ANC
(33%) were the least available guidelines in HPs. The availability of guidelines
did not vary much between agrarian and pastoralist HPs. Among the regions,
iCCM was least available in Gambela (36%), FP (10%) and CBNC (12%) in
Afar, and ANC (8%), PNC (9%), and CHIS (3%) in the SNNPR (Table 9-1).
Correspondingly, respondents for the qualitative assessment were asked about

the availability and clarity of the HEP implementation strategies and guidelines
and the challenges they could think of with respect to them. Accordingly, most
respondents indicated that the HEP guidelines and strategies were available at
all levels. They also indicated that these guidelines and standards were helpful
in facilitating the HEP’s implementation. The availability of these operational
guidelines for the implementation of the HEP were reportedly available in
all regions, which has been confirmed. It is best explained by the statements
expressed by 2 respondents, 1 from the MoH and 1 from the Gambela RHB:
“
In the HEP, what I know is three things…the agrarian, pastoralist and
UHEP…the guideline is clear in the recruitment of the professionals
[HEWs] and allows them to be different. For example, for the
Urban Health Extension, they are nurses, and the time needed for
pre-service training is 3 months, whereas it is 1 year for agrarian…
But for the pastoralist areas, since it is difficult to find those who
have attained this level of education and difficult to make all of
them female, the guidelines also allow them to include males and
recruit from those who have only completed grade 8 as HEWs.
“
KII, MoH
Guidelines and packages, even HDA manuals, were provided to us

and to the HEWs. When new modules come for implementation,
we distribute them to HEWs and other implementers.
KII, Gambela RHB
page- 425
While acknowledging that the guidelines were helpful to their work, some
study participants also explained that they didn’t have the required guidelines
regarding the operation of model families or for the M&E of HEP activities.
Statement from respondent in the Yallo (Afar) woreda is exemplary in this
respect:
“
They [HEWs] teach the community based on the given guidelines
and strategies. It is very good. It is additional knowledge for HEWs.
It makes their work easy. They will not forget what they have learned.
They will gain more experience…but they don’t have guidelines they
can use in their dealings with select model families.
KII, WorHO HEP Coordinator, Afar
Table 9-1. Percentage of HPs with HEP guidelines, by region and livelihood
Percent of health posts with the HEP guidelines listed
Information System
Expanded Program
Community Health
Community-Based
Community Case
on Immunization
Family Planning
Number of HPs
Antenatal Care
Postnatal Care
Newborn Care
Family Health
Management
Integrated
Guideline
(CBNC)
(iCCM)
(ANC )
(CHIS)
(PNC)
(EPI)
% % % % % % % %
National 75.3 64.6 63.5 52.8 49.6 33.4 30.1 27.2 343
Region
Tigray 74.4 78.2 58.3 67.6 65.9 33.4 39.6 60.2 32
Afar 70.1 22.5 11.9 43.7 10.3 31.3 22.3 10.3 118
Amhara 86.5 78.8 82.3 62.3 65.4 44.9 44.8 28.1 60
Oromia 84.7 59.1 65.4 44.5 44.1 39.4 30.7 36.4 74
Somali 70 54.9 52.2 66.8 56.4 58.4 59.4 50.2 43
Benishangul-
76.5 48.4 54.1 46.6 53.4 34.5 32.2 7.5 24
Gumuz
SNNPR 51.6 65.2 49.6 54.8 42.9 7.5 8.8 3.4 59
page-426
Gambela 35.7 30.6 40.9 45.9 40.8 35.8 30.6 20.4 17

Harari 63.5 55.6 39.7 77.8 86.5 54.7 54.8 9.5 16
Livelihood
Agrarian 75.5 65.2 65.4 53.4 50.9 32.9 29.5 27 235
Pastoralist 72.6 53.7 44.9 47.1 36.9 37.8 35.7 29.2 108
Abbreviations: HEP, Health Extension Program; HP, health post; SNNPR, Southern Nations,
The strategies for HEP implementation in pastoralist settings do not differ

from those in agrarian ones. Creating community ownership by improving
community participation using 1-to-5 groups and the WDA is similar in both
settings. Forming WDAs in pastoralist areas, however, cannot be done on a
permanent basis, as the communities’ settlement and movement patterns do
not allow such a formation to last very long, nor for members to work together
seamlessly. Therefore, some participants for this study strongly believed that the
strategies should be redesigned to tackle such challenges. A statement from a
WorHO head in Oromia (a pastoralist community) confirms this example:
“
The strategy for the formation and working modalities of the
WDA should be redesigned to fit the movement and lifestyle of
the pastoralist communities….People are scattered and usually
move from place to place and from time to time with their herds
and cattle. If you establish a WDA, they don’t stay together for
long enough to plan, implement, and monitor the HEP activities
together as a team or as a group.
WorHO Head, Oromia
The strategies and guidelines were also found to be clear enough to specify the
roles and responsibilities of different actors sufficiently and to take immediate and
prompt actions, especially during disease epidemics. The HEP’s standards and
guidelines consider the context of population dynamics during the recruitment
of HEWs in terms of education level, gender, and socio-cultural sensitivities. For
example, in the pastoralist case, community members of both genders who had
reached grade 8 could be recruited as HEWs to serve their communities. The
recruitment guidelines for HEWs in agrarian settings require candidates to have
completed grade 10, and in urban settings, nurses or midwives are recruited
page- 427
to deliver HEP services. Job resources, on the other hand, are very helpful and
accommodate the information need of illiterate community members because
they are prepared and presented in pictorial form. Statements from a process
owner in a pastoralist woreda in Oromia, a program advisor from the MoH,
and a senior staff member from the Amhara RHB have all described the
usefulness and clarity of the HEP strategies and guidelines:
“
For example, if we see the family health guideline, it is meant to
be useful to women in the community, and there are also posters,
which are very appropriate for those community members who are
unable to read but can well understand the message in the pictures
of the posters. The communities have been using these guidelines
and posters both at the HC and HP levels.
Process Owner, Pastoralist Woreda, Oromia
“
It is not about the packages. Rather, we have focused on availing
guidelines to help them use them as reference materials for their
activities. For example, when we talk about child health, we need
clarity on what the focus areas of child health are. We should ask
questions, such as, “Do HEWs have sufficient knowledge and skill?”
If we find out that they don’t have the required skills, then we may
have to train them or provide them with operational guidelines
and do follow-up to ensure their adherence to the guidelines. The
packages are not changed, but they are changed to be practical
beyond education.
Program Advisor, MoH
“
HEP guidelines are useful. Basically, we use the 18 packages
for implementation. All packages have their own booklets. They
are written in an easily understandable way. Even mothers can
understand them easily. They are inclusive and user-friendly. There
are also monitoring tools for HEWs. They are translated into
different languages. They are well suited to our policy and strategy.
But they also require timely revisions.
Senior Staff Member, RHB, Amhara
page-428
Some study participants complained about the lack of availability of the

guidelines in local languages other than the major languages (Amharic, Affan
Oromo, and Tigray).
The guidelines and strategies for the implementation of the HEP are written very
well, distributed widely, and available in all regions. HEWs and implementers
are also trained on their uses. The greatest challenge faced by many, however,
is the lack of adherence to the strategies and the different stakeholders’ lack
of common understanding of and familiarity with the guidelines. Additional
challenges complained about by many include the inadequacy of the training
regarding, the insufficient distribution of, and the lack of regular updating of
the HEP guidelines. The following statements from WorHO staff in Amhara
“
and Oromia refer to these challenges:
We aren’t updating the manuals/guidelines, and we are not

implementing them well. We received them a long time ago. There
is also a huge gap in the knowhow and understanding of the HEP
guidelines among professionals.
WorHO Staff Member, Amhara
“
The HEP has its own guidelines….The guideline is clear that, when
1 of the HEWs stays at the HP, the other should be out to conduct
house-to-house visits in the community…. But now you will find both
of them at the HP.…they prefer to sit at the HP all day long. That
has affected household visits.
“
WorHO Staff Member, Oromia
In our case, we have 13 HPs as catchment areas to 1 HC, which is

beyond the standard of 5 HPs per HC...so some of the standards
written in the guidelines are not implemented accordingly.
page- 429
According to the respondents, the shortage of staff at the HC level was a

major reason for not adhering to the HEP guidelines. A statement from a staff
“
member from Surupha woreda (Oromia, pastoralist) is an example:
As per the HEP guideline, at least 1 health worker from the HC is

expected to provide supportive supervision for HEWs on a weekly
basis; however, we could not do it, due to the shortage of staff at
the HC.
WorHO Staff Member, Surupha Woreda, Oromia
9.3 Planning in the Health Extension Program

Planning is an important step in the HEP’s process of implementation. The
expected standard and process is as follows.
In collaboration with the kebele council, HEWs conduct a baseline assessment to

identify health problems and count the kebele-level households and population
to be targeted through HEP packages. Following the baseline assessment,
plans of action are prepared. The kebele- and woreda-level councils approve
the HEW’s plans before they embark on implementing the HEP activities. Once
the plans are approved, higher-level offices (woreda, regional, and federal
health offices) receive these plans for follow-up and supportive supervision.
In line with the national health planning framework, the quantitative findings
from the HPs indicated that there was some sort of community participation
in the processes of planning, implementation, and performance review (Table
9-2). Accordingly, more than three fourths (79.9%) of HPs have an annual
plan for 2011 EFY, around 38% of which was prepared jointly with the HC or
WorHO. Regarding community participation in the planning processes, around
one third (35.5%), three fourths (73.5%), and half (49.7%) of the HPs included
community members in the planning, implementation, and performance review
processes, respectively.
page-430

community participation in the planning processes, by livelihood
Livelihood
(n=235) (n=108) (n=344)
% % %
Availability of annual plan 83.4 45.9 79.9
Involvement in planning preparation n=220 n=65 n=285
HP prepares plan by itself 21.5 43.8 22.7
HP prepares plan with HC/WorHO 39.4 23.4 38.2
HC/WorHO prepares for HPs 36.9 32.8 36.7
Others (Steering committee, Kebele council,
2.6 0.0 2.4
school members)
Community involvement in: n=235 n=108 n=343
Planning process 36.4 20.4 35.5
Implementation 76.3 45.1 73.5
Performance review 52.7 20.0 49.7
Abbreviations: HP, health post; HC, health center; WorHO, woreda health office.
As shown in Table 9-3, the availability of an annual plan at the HPs varied
from region to region; for instance, it was 100% in Tigray and Harari, but only
12% in Somali. On the other hand, community participation during planning
was high (47%) in Tigray and low (7%) in Afar. During implementation of
the HEP, community participation ranged from high (88%) in Tigray to low
(19%) in Somali. A similar percentage distribution was observed for community
participation in performance review, with the highest rate in Tigray and the
lowest in Somali.
page- 431

community participation in the planning and review process, by region
Community participation during:

Annual plan
Region Performance Number of
availability Planning Implementation
review health posts
National 79.9 35.5 73.5 49.7 343
Tigray 100 46.7 87.9 87.9 32
Afar 77.2 7.4 59.8 29.9 18
Amhara 89.9 18.0 85.8 66.6 60
Oromia 84.8 40.6 74.3 41.2 74
Somali 11.8 8.5 18.6 5.2 43
Benishangul-
93.2 10.6 82.2 51.6 24
Gumuz
SNNPR 75.0 44.8 71.8 55.6 59
Gambela 38.4 46.0 48.6 25.5 17
Harari 100 45.1 68.3 9.5 16
As part of this study, both agrarian and pastoralist communities and program
staff were asked about the status and practices of planning for the HEP and
their engagement in this critical activity. In the qualitative findings, a reflection
“
from a WorHO Head in Benishangul-Gumuz affirmed this critical step:
Well, if we don’t have a plan, you may consider that we will be

blinded in our implementation of the HEP. The first thing to do
before the budget year is to plan. We all participate in the planning,
which involves the region, zones, woredas, and the kebele levels.
The planning also goes down to the household and family level
concerning the HEP packages. If there is a reduction in the budget,
we re-plan to fit the reduction.
WorHO Head, Benishangul-Gumuz
From the qualitative findings, most program participants (including HEWs)

in both agrarian and pastoralist communities agreed that HEWs receive
general directions and schedules from higher officials, particularly from HCs
page-432
and WorHOs on how to prepare the monthly, weekly, and daily plans for
HEP implementation at the HP level. The annual HEP planning process and
activity planning, however, involve woredas guiding the HPs (i.e., a top-down
approach). An example of this process is given by an HEW in Kurmuk woreda
“
(Benishangul-Gumuz):
Our plan directives for the planning comes from the woreda, I
mean our yearly planning. We will have a meeting to share and
assess this plan, including reviewing the activities to be done at
the HP. We then use this as a basis to develop detailed weekly and
monthly pans. That is how we do the planning.
HEW, Kurmuk woreda, Benishangul-Gumuz
In planning HEP activities, HEWs also allocate activities based on the 3 service
delivery modalities: (a) house-to-house visits, (b) static services at the HP, and
(c) outreach service. An HEW from Harari (Dire Tiya woreda) described this
“
process well:
First of all, the three of us have a schedule for a meeting to discuss

on our plan. After discussion, we will develop action plans identifying
activities that have to be delivered through the outreach, HP-based,
and home visits. We are not working without having a plan.
HEW, Dire Tiya woreda, Harari
HEWs facilitate their planning and engage their communities in the planning
process using socio-culturally adaptable times. For example, they usually select
the slack period in the agricultural seasons and holidays, when many would
stay at home. They use these opportune moments to involve the community in
the planning and to educate and deliver HEP services at the household level.
The communities are meant to participate actively, through the local level
structure, in the human-resource development (by directly selecting the HEW
candidates for training) and activity planning of the HEP. Both agrarian and
pastoralist participants for this assessment, however, agreed that HEW and
HP plans were not implemented well for several reasons. The inadequacy of
human resources and lack of budget for HEP activities were cited as the major
reasons that plans were not translated into actionable results.
page- 433
9.4 Decentralization of the Health Extension

Program
At present, the implementation of economic policies and development programs
has shifted to a large extent from federal to local structures. To deepen the
decentralization process, the government’s strategy is to empower woredas as
the center of socio-economic development. To ensure their autonomy, woredas
are given a block grant to cover recurrent and some capital expenditures.
As indicated by the planning section, once the HEWs produce their plans,
the woreda council, in collaboration with the kebele council, approves them
and communicates them to the WorHO, the regional council, and the RHB.25
Some of these decentralized roles that were noted in both the quantitative and
qualitative findings at the lower levels of the HEP structure are reflected below.
9.4.1 HCs’ decentralized roles in HEP implementation
In the quantitative assessment, HCs were asked about their different roles in
implementing the HEP. As shown in Table 9-4, 92% of HCs play roles in both
administrative and technical oversight with their catchment HPs. This role was
much more likely to be reported in agrarian (96%) than pastoralist (80%) HCs.
The most frequently cited roles and responsibilities of HCs in their catchment
HPs were providing reporting formats (97%), providing supportive supervision
(96%), supplying program drugs (96%), supplying drugs and supplies from the
revolving fund (92%), reviewing the performance of HPs and HEWs (92%),
and providing consumables and supplies (91%). Paying HEWs’ salaries and
training HEWs were also reported as HC roles that had been decentralized to
the local level. A similar percentage distribution was observed among agrarian
and pastoralist HCs.
In performing these tasks at the HC level, most HCs (78%) had a dedicated
HEP coordinator, while 16% of HCs assigned other technical staff to coordinate
the HEP at the HC level. Six percent, however, of the total HCs interviewed
(10% in pastoralist and 4% in agrarian communities) did not have anyone to
page-434
coordinate HEP activities. On the other hand, over half of the HCs (56%) had
trained staff on the HEP, but only 29% of them provided HEP training to their
catchment HEWs.
Table 9-4. Percentage of HCs reporting their decentralized roles in the HEP
implementation
Livelihood
HC roles Agrarian Pastoralist Total
n=139 n=40 n=179
Working relationship with health post
Administrative oversight
1.4 7.5 2.8
only
Technical oversight only 2.9 12.5 5.0
Administrative and
95.7 80.0 92.2
technical oversight
Role of health center on directly linked health posts
Pay salary of HEWs 48.2 40.0 46.4
Supply program drugs 99.3 85.0 96.1
Supply drugs from revolving
95.7 77.5 91.6
drug fund
Provide consumables
94.9 77.5 91.1
supplies
Provide reporting formats 99.3 87.5 96.6
Provide supervision 100 82.5 96.1
Provide training to HEWs 50.4 22.5 44.1
Renovate health posts 33.1 30.0 32.4
Performance review of
94.9 82.5 92.2
HPs/HEWs
Availability of dedicated responsible body for HEP services in the Health center
HEP coordinator 79.1 75.0 78.2
Other technical staff 16.6 15.0 16.2
No 4.3 10.0 5.6
Availability of trained staff on HEP 61.2 40.0 56.4
Training provision for HEWs in the
31.6 20.0 29.1
last 2 years
Abbreviations: HC, health center: HEW, Health Extension Worker; HEP, Health Extension Program.
page- 435
9.4.2 Woreda health offices’ decentralized roles in HEP implementation
According to the findings of this assessment, some HPs were directly supervised
by WorHOs. As shown in Table 9-5, about 21% of WorHOs had at least 1 HP
linked to their catchment that received direct supervision from the WorHO.
Regarding the WorHO role, about 44% assigned HEWs to the HPs as well as
provided supportive supervision. Over half of WorHOs (57%) had assigned 1
dedicated expert, and around 61% of the woreda had trained staff on the HEP.
There is, however, a remarkable variation between agrarian and pastoralist
areas in the status of these issues.
Table 9-5. Percentage of WorHOs reporting their decentralized roles in the

HEP implementation
Livelihood
n=42 n=20 n=62
HPs directly linked to WorHO 16.7 30.0 21.0
Role of WorHO in catchment HPs that
are not linked to catchment HCs
Assign HEWs to HPs 47.8 36.4 44.1
Pay salary of HEWs 47.8 20.0 39.4
Supply program drugs 38.1 30.0 35.5
Supply drugs from revolving drug
36.4 0.0 25.0
fund
Provide consumables supplies 31.8 30.0 31.3
Provide technical supervision 45.5 40.0 43.8
Availability of a 1 expert 61.9 61.9 57.4

dedicated staff for HEP Team 38.1 47.4 40.9
coordination No 0.0 5.3 1.6
Availability of trained staff on HEP 50.0 83.3 61.1
Abbreviations: WorHO, woreda health office; HC, health center; HEW, Health Extension Worker; HP,
health post; HEP, Health Extension Program.
page-436
9.4.3 HEW and kebele administration decentralized roles in HEP

implementation
From the qualitative assessments, most participants in both agrarian

and pastoralist areas described HEWs as having autonomy in planning,
implementation, and monitoring their activities once the plan was approved.
This is a sign of decentralization. It is also reported, however, that they are
engaged in non-health activities in most of their working hours: for instance,
tax collection, the social mobilization of communities for political purposes,
and community health insurance activities, which had implications for their
autonomy and the decentralization of the HEP structure. A pastoralist HEW
“
and a WDA in the SNNPR described this scenario:
When you try to achieve your goals, the different activities and
orders coming from different offices become obstacles. If you say,
“I am going to work on my plan today,” they will order you to do
something else, and there are a lot of disagreements.
“
HEW, Pastoralist
They follow directions from higher officials about drainage system

construction, saving for fertilizers, and other non-health related
activities. They mobilize mothers to stand by their husbands by
saving money for fertilizers. This is an example in which they follow
directions given by higher officials.
WDA Leader, SNNPR
Contrary to such claims, some kebele administrators do not accept any

suggestions that the HEWs are engaged in non-health activities. Instead,
kebele leaders do try to motivate HEWs and share duties and responsibilities;
they work together in some activities and make sure that HEWs represent
the health sector in the kebele administration. This shows that HEWs have
autonomy in relation to health activities. A statement from an agrarian kebele
leader in the SNNPR corroborates this autonomy and decentralization of roles
and responsibilities.
page- 437
“
We don’t order them[HEWs]. It is a professional job; we don’t

provide other jobs to HEWs. If there is a disease outbreak, we may
send them [HEWs] to the sites to vaccinate children. They could file
a report if they are engaged in a non-health activity, but we don’t
order them to do such a thing.
Kebele Leader, SNNPR
Expert participants from the program admitted that the kebele administration
staff and staff from other sectors paid less attention to HEP activities despite
HEWs’ having a decision-making mandate on health-related activities at the
kebele level. A program staff member from other process owners in Abaya
“
Woreda, Oromia confirms this challenge:
The challenges with other sectors include a lack of attention to

the HEP at the kebele level. Even though HEWs are the decision-
makers at kebele as per the structure, they are not influential. This
is their [other sectors’] weakness.
Despite the HEP’s being fairly decentralized to the woreda- and kebele-
level structures, including HPs and HEWs, the support these local-level
structures gave to HPs and HEWs was found to be very low. Findings from
the quantitative results (HEW’s interviews regarding their perceived support
from the HC, WorHO, and kebele administration) indicate that such support
from the kebele, HC, and WorHO was perceived to be as low as 63%, 51%,
and 44%, respectively. Relatively, the perceived support from these local-level
structures to HPs and HEWs was higher in pastoralist areas than agrarian
ones. Similarly, the pastoralist regions had better support than the agrarian
regions (Table 9-6).
As shown in (Table 9-6,) less than half of HEWs agreed that HCs’ and WorHOs’
commitment to the HEP had increased over time. A similar distribution was
observed among the regions in the agrarian setting. Such commitment was
observed more in pastoralist areas than agrarian ones.
page-438
Table 9-6. Percentage of HEWs reporting their perceived agreement with the
support they receive from kebele administration, HC, and WorHO, by region
and livelihood
HEP is HEP is HCs/Woreda

HEP is
adequately adequately health offices’
adequately
supported supported commitment Number of
supported
by health by Woreda towards HEP is HEWs
by Kebele
center health office increasing over
administration
supervisors supervisors time
National 51.4 43.5 62.8 49.2 583
Region
Tigray 59.5 52.0 60.4 40.4 63
Afar 69.8 72.2 87.8 86.8 19
Amhara 63.3 51.8 56.5 49.0 95
Oromia 47.1 35.3 58.0 42.9 123
Somali 91.4 90.8 89.3 92.2 75
Benishangul-
57.4 67.0 69.9 71.8 37
Gumuz
SNNPR 37.0 35.9 69.4 49.1 96
Gambela 57.7 59.1 84.4 66.2 41
Harari 63.6 65.4 76.6 71.4 34
Livelihood
Agrarian 48.4 39.7 60.3 45.8 413
Pastoralist 80.7 79.5 86.6 81.3 170
Abbreviations: HEW, Health Extension Worker; HC, health center; WorHO, woreda health office; HEP,
Health Extension Program.
Some experts and officials at the regional and MoH levels argued that the
HEP was declining due to changes in leadership and political dynamics, a
lack of commitment from the leadership and governance bodies, HEWs’ low
motivation, HEWs’ workload (i.e., in implementing multiple packages), HEWs’
engagement in non-health activities, HEWs’ poor working conditions (i.e., denial
of transfer, lack of annual leave, lack of promotion or career development
opportunities, unfavorable living conditions compared to their income level).
Another reported reason for the decline of the HEP was the structural shifting
of HEP’s supervisory body from the woreda to HCs without proper coaching or
an adequate budget. An FGD participant from the Amhara RHB confirmed
the declining status of the HEP:
page- 439
“
When we compared the current situation of the HEP with the
previous one, the attention given to the program was declining. At
the beginning, the program got attention from the federal/regional
government up to political leaders and the health sector manager.
However, the current status of the HEP is the opposite of what it
was previously. This created a question of whether our managers
at different levels can manage, be aware of, and provide attention
to the HEP. When we compare the current performance status of
the program with the beginning, it has been declining. There is even
sometimes a situation in which you conclude that the program has
already died.
FGD, RHB, Amhara
9.5 Coordination of Health Extension Program

activities
Links between the WDAs and HEWs, HEWs and HPs, and HPs and HCs
or WorHOs were also properly designed so that the HEP would be well
integrated into the PHC system. To strengthen the performance and links, these
structures have a regular monthly meeting that helps them review the planned
performances and achievements, identify bottlenecks, and make concerted
efforts to resolve impediments in order to bring about meaningful changes
in the health conditions of the communities. Such links are working well, as
“
reflected in the statement of a participant from Harari:
In the HEP’s implementation, all health workers at the health center

have schedules for supporting the HPs. When they [the HC staff]
visit the HP for supportive supervision, they make assessments
about whether all 18 packages have been implemented or not. In
collaboration with HEWs, they conduct a meeting with pregnant
mothers, do a syphilis testing for pregnant mothers, and also
provide curative services.
FGD Participant, Harari
page-440
Some participants in this study, however, reported that the HEP’s coordination
structure was not identical across HCs: in some, the vice manager is the
HEP coordinator, while in others there is an assigned process owner as the
HEP coordinator. WorHOs usually have a dedicated HEP coordinator. Such
reflections were echoed by an Oromia and an SNNPR WorHO process owner
“
(both pastoralists):
The HEP structure at the woreda and HC levels are not the same.
In the health center, the vice-manager is the HEP coordinator. In
the office, the program is getting attention and the office has an
HEP coordinator. Accordingly, the HEP coordinator at the health
center has been providing support to HPs at least once a week.
Similarly, at the kebele level, one of the HEWs is assigned to be
an HEP coordinator. This is how we coordinate the HEP activities
in our localities.
“
Process Owner, Oromia
There is an HEP focal person assigned in this HC to support the

HEP activities in the catchment HPs. Sometimes a request may
come from the woreda health office to assign additional staff to do
field-level support to the HEWs.
Process Owner, SNNPR
9.6 Collaboration with other sectors

Participants in FGDs and key informants reported that different sectors
collaborated on and facilitated HEP activities in various ways. For example,
the agriculture sector teaches farmers the importance of latrine construction
and the education sector jointly works with HEWs to convey health messages
through school programs (like mini-media) on HIV, reproductive health, and
related issues. In addition, The water and energy sector supplies electricity
to the HPs, and the environmental sanitation representative office works
collaboratively in the area of malaria prevention and control and some other
programs, such as water and sanitation, including training community members
page- 441
on water treatment and sanitation. A statement from an HEW in the Bahir Dar
Zuria woreda in Amhara) describes this sectorial collaboration to facilitate the
“
HEP:
One of the HEP’s goals is to have a clean and healthy environment.

So, while the agricultural professionals are working on their duties,
they also teach farmers to build their own latrines, integrating the
HEP activities so that it facilitates inter-sectorial collaboration.
HEW, Bahir Dar Zuria Woreda, Amhara
Apart from the involvement of other government sectors, different NGOs and
the private sector are collaborating with the HEP, depending on their program
focus and/or objectives. They support the implementation of some HEP
packages. Among the major activities in which they participate or collaborate,
as indicated by the participants of this study, are: training, creating awareness,
immunization, training of model families, forming WDA groups, preparing
manuals, providing material support to the HEP, and monitoring the HEP. In
addition, some partners provide support to the construction of HPs and youth
centers. Key informants from Gambela and Amhara asserted that there had
“
been collaboration with other sectors and partners:
Some NGOs that have roles in supporting the HEP provide us

budgets for trainings. When a new health project is signed with the
MoH, the project channels funds for the trainings in line with their
scope of the project. They help us in the printing of manuals, assist
us in implementation of the packages, and provide immunization,
nutrition supplements, and job-aids.
“
KII, RHB HEP Coordinator, Gambela
They [NGOs] provide trainings on child feeding. They also provide

feeding bowls and shoes to the development groups to motivate
them.
KII, HEW, Amhara
page-442
9.6.1 Adequacy of stakeholder collaboration
Resources given by collaborators to health programs are always limited. Mostly,

partner organizations implement projects with a limited duration and only
provide specific support for activities in line with their own program objectives
and major focus areas. Due to these problems, gaps will be created when the
partners’ projects phase out, raising serious sustainability issues with regard to
implementation. This requires a concerted effort by the MoH, RHBs, and
WorHOs to streamline partners’ assistance towards implementing the
existing program to minimize loopholes and redundancy.
Although most respondents stated that Ethiopia has been receiving huge
support (both technical and financial) from international and local collaborators
(mostly NGOs), the support has been declining over time. Participants perceived
several possible reasons for this: the political situation of the country in the last
few years, as well as a global decrease in the flow of development assistance
from developed to developing countries, including Ethiopia. Such circumstances
create uncertainty in the much-sought sustainability and continuity of health
projects that are supported by external sources, including those with a short
duration. A few reflections from the MoH and Amhara RHB suggest the general
“
uncertainty about and inadequacy of funding from external sources:
While the government budget for the HEP has increased

significantly over the years, support from external sources has
declined substantially...so there is no question that it will decline
further….This is now going to become inevitable; we should prepare
ourselves to fund our health programs by ourselves….
“
KII, Policy Advisor, MoH
It [donor assistance] has declined dramatically, both in terms of

financial and material support….This is one of the reasons behind
the decline in the HEP’s performance, because some program areas
are donor-dependent. The reason behind this may be the effect of
global politics and the recent instability in our country. The decline
is not only in the HEP. There is a huge shortage of support in most
other sectors as well.
KII, RHB Head
page- 443
9.6.2 Benefits and challenges of collaboration
According to the study participants, there are undeniable benefits of

collaborations with international and national partner organizations that help
fill budget gaps and provide technical support to various service providers
with current technologies and techniques. Improvements in the health status of
the people as a result of substantial reductions in mortalities and morbidities
due to the eradication of vaccine-preventable diseases, such as polio, and
other communicable diseases are a few of the many outcomes in which the
country benefited from the collaboration with and assistance of donors. Two
statements from respondents in Amhara and Benishangul-Gumuz are examples
of the benefits of collaborations. Despite such benefits, however, there is also
the downside of support from external sources, which has become increasingly
inconsistent and unpredictable, creating uncertainties about the future
sustainability of health programs that began as a specific project for a limited
“
duration with relatively narrow objectives and specific areas of interventions.
When NGO projects phase out, we have no choice but to keep

working on our own to sustain the changes. We learn from what
they have been doing and use that lesson to safeguard the observed
changes by the time the project is phased out.
KII, WorHO Process Owners, Benishangul-Gumuz
9.6.3 Commitment of managers
We assessed the commitment of HEP managers and partner organizations

to implementing HEP activities through the qualitative method (FGDs and
KIIs). Most participants in both agrarian and pastoralist settings agreed
that the HEP was a priority health delivery government program. In line with
this priority, the government has hired a large number of HEWs, allocated
budget for HEP implementation, provided pre-service and in-service training
for HEWs, conducted continuous supportive supervision to HPs, and undertook
monitoring, evaluation, and review of HEP’s performance. A reflection from a
partner organization argued that the HEP is a priority for the GoE:
page-444
“
I think in terms of priority, the HEP is one of the highest priorities

for the government…. The evidence shows that the government
recruits and pays salary for more than 40 000 HEWs in rural,
pastoralist and urban…and also the infrastructures, there are more
than 17 000 HPs…in terms of the structures from federal to regional
and regional down to the primary healthcare level, there are good
linkages between HCs and HPs.
FGD Participant, Federal-Level Partners
Key informants (kebele administrators) stated that they were committed

to implementing HEP in their localities. As evidence, they mentioned their
facilitation of and engagement in the construction of HPs, mobilization of
WDAs, and creation of 1-to-5 networks, as well as their support to HEWs in
the implementation of HEP activities through home visits and review meetings,
These are a few visible examples showing the government’s commitment. Such
views were reflected in statements from a participant in Demba Gofa (the
“
SNNPR) and a process owner from the Abaya woreda (Oromia):
When the HP was constructed, the kebele administration facilitated

and supported it by providing materials and also helped in getting
electricity for the HP. They wanted the HP to function well so that
the communities could use it.
“
KII, Kebele Leader, SNNPR
We appreciate the work of HEWs. In our effort to maintain the

HEWs’ morale, we engage ourselves in supportive supervision,
monitoring and evaluation, including in the maintenance work of
the HPs when needed. We also closely work and collaborate with
the community to maintain the linkages of the kebele with the
HC as a means of strengthening the commitment of the Woreda
Health Office.
KII, WorHO, Other Process Owners, Oromia
Most respondents for this assessment, however, described the level of the HEP
managers’ commitment at the kebele, woreda, regional, and federal levels as
having declined over time. The following statements from the HEP coordinator
page- 445
in the SNNPR and a process owner in Benishangul-Gumuz are examples in this
“
regard:
In general, the effort to achieve PHC through the HEP is not

satisfactory. The top and the grassroots-level health professionals
do not support each other….When things are good at the grassroots
level, the same will happen at the higher level. But in our case, that
is not the fact we see. The capacity-building efforts at the lower
level have not been enough.
“
KII, WorHO HEP Coordinator, SNNPR
Priority in this regard is only given at the policy level. Down at

the operational level, the HEP does not have its own budget. The
woreda does not give attention to the HEP since the officials are
changed frequently. Their [officials’] awareness about the HEP is
not improving over time.
KII, WorHO Process Owners, Benishangul-Gumuz
Some participants, however, reported that kebele leaders were not responsive
to the HEWs’ concerns. The kebele leaders are not considering the HEP as their
major agenda and therefore don’t pay due attention to or provide support
to the HEP. There have also been false reports about the performance of
HPs. Assigning HEWs to non-health-related activities was also reported as an
indication of the lesser commitment by kebele leaders to the HEP. Allocation
of an inadequate budget for HPs’ and HEWs’ activities, the non-participation
in the HEP’s M&E activities, and the lack of attention paid to the construction
and maintenance of HPs were among the other complaints from community
members and partner organizations. One statement from a man from the
“
Wama Hagalo woreda (Oromia) is an example of such complaints:
We complained many times to the Woreda Health Office about

the delay in the construction of the HP. We want them to take
action to finish it soon. Their response was not good….I myself went
to the office, together with the HEW. They promised to do it next
year with an excuse of the lack of construction budget. This has
happened many times.
FGD, Male Community Members, Oromia
page-446
Another idea for defining attention was to have a strategic plan for the HEP
alone instead of integrating it into the PHC system. This was reflected in a
“
statement from a partner organization to the MoH.
The other way to boost attention is work on soft activities, such as

guidelines, training materials, SOP, job-aids, and so on. That is also
mixed because I have not really seen like a 10-year strategic plan
for the HEP. There are different support materials and SOPs….We
need a binding document showing the vision and targets…for the
HEP… different materials that are part of the HSTP are included
in those documents…but we need to see a dedicated strategic plan
for the HEP.
FGD, Federal-Level Partners
page- 447
SECTION III
Conclusions and Recommendations
1. Health Extension Program service packages
CONCLUSIONS
• Despite substantial improvement in health indicators since the beginning

of the HEP, communicable, maternal, neonatal, and nutritional disorders
(CMNNDs) still constitute 60% of the total DALYs lost in Ethiopia.
Moreover, the increasing burden of NCDs calls for enhanced access to
prevention and control.
• All packages of the HEP are relevant in addressing the major causes of
morbidity and mortality among rural communities. The recent addition
of non-communicable diseases and mental health packages also create
opportunities to address the increasing burden of non-communicable
diseases. HEWs are providing clinical services, including ANC, other
maternal health services, and the treatment of sick children; providing
clinical services through the current HPs, however, given the existing
capacities of Health Extension Workers and other material resources,
has led to compromises in the quality of care.
• Exposure to HEWs is positively associated with improved maternal and
child health outcomes. It has also played a role in improving access to
basic sanitation and personal hygiene and contributed to the prevention
and control of communicable diseases.
• The availability of more comprehensive services at HPs has a positive
influence on the acceptance of HEWs and the links of community-based
health-promotion activities with service uptake. Communities have high
demand for more comprehensive services at the HP level. This demand
has not been adequately addressed for different reasons, including 1)
the decision to keep clinical and curative services separated from health-
promotion and disease-prevention activities, 2) the limited trust of the
community in HEWs’ ability to provide clinical services, despite a generally
high level of trust and acceptance and the availability of selected clinical
services at HPs, and 3) the low level of community awareness regarding
services that are already available at HPs.
• None of the health-promotion or disease-prevention activities are
adequately implemented to a level where the community can sustain the
adopted behaviors.
page-448
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• The current packages should be continued by addressing their
implementation challenges.
• HEP packages should evolve with clear milestones to graduate

(bring to an end) packages or specific components of packages
upon achievement of sustainable behavioral change at house-
hold level. This also applies to activities targeting behavioral
change through home visits and outreach sessions.
Modify • Health and health system literacy has to get adequate attention
either as part of each package or as a separate package.
• Provide clear standard operating procedures for health post
operations (health post-based activities, home visits, outreach
sessions) in different contexts to guide implementation, monitor-
ing and evaluation, and controlling of health posts.
• Packages should incrementally expand with the goal of meet-
ing communities’ expectations for more comprehensive services
at health posts. This includes comprehensive maternal health
services, treatment and referral for common childhood illnesses,
and treatment of common communicable diseases among adults,
and screening and referral for common other illnesses including
NCDs. Allow packages to vary across health posts depending on
local realities including proximity to health centers and hospitals.
Add Versions of HEP packages should be developed for:
o HEP unit in a HC: for rural kebeles with HCs or pri-
mary hospitals
o HPs that should implement basic set of packages
o HPs that should implement comprehensive set of
packages
• Conduct an in-depth study on birth outcomes among deliveries
attended by HEWs to generate additional evidence on capacity
of HEWs in handling labor and delivery.
Drop
• Avoid delivery attendance by HEWs until adequate evidences
are generated regarding the skill levels of HEWs in managing
normal delivery and detecting complications.
page- 449
Conclusion and Recommendations
2. Service delivery modalities
CONCLUSIONS
• The health-seeking behavior of rural communities is sub-optimal. Static

services alone won’t allow the health system to adequately expand its
coverage of essential health services.
• There is high community acceptance and approval of HEP service
delivery through home visits, HP visits, and outreach sessions. Female
HEWs are highly recommended for home visits; limiting HEWs to
women only, however, was criticized for: 1) the difficulty of reaching
all segments of the population within a kebele because of distance,
geographical barriers, and security concerns and 2) the difficulty of
achieving behavioral change at the household level without involving
men in a patriarchal society.
• Involving males as HEWs has been approved as appropriate by large
portions of the community.
• Campaign-based approaches and strategies involving coercion or
punishment to increase implementation of the HEP at the household
level did not achieve sustainable behavior change in the construction
and use of latrines or use of maternal health services.
• Implementation of the HEP has a high level of deviation from theories
that are thought to have informed its design (i.e., diffusion of innovation).
• In pastoralist communities, the HEP requires a different arrangement
of service delivery that addresses challenges related to the sparse
population settlement and mobile lifestyle.
page-450
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain • Static, home visit, and outreach service delivery modalities

• Female HEWs responsible for contacting women during home
visits
• Enhance the use of health post visits as entry point for provision
of comprehensive health promotion and disease prevention
services. The increasing role of health post visit as an entry point
and means of exposure to HEP has to be promoted to bring a
comprehensive behavioral change at household level.
• Revise behavior change theories and strategies based on
Modify
variations in the needs of specific behavioral outcomes and
cultural contexts.
• Increase involvement of men and youth as targets of HEP
• The strategy for outreach modality should be designed in a way
that includes social capital or indigenous social institutions like
the idir, equb and others.
• Include male health workers in the cadre of HEWs.
• Redesign pastoralist HEP by conducting more detailed analyses
of experiences in addressing health and other social needs of
pastoralist communities including villagization/settlement of
mobile communities, mobile health team, mapping movements of
Add
pastoralist communities, and other program specific experiences.
• Strengthen inter-sectoral collaboration to ensure that strategies
to implement HEP in pastoralist communities are integrated/
coordinated with other community-based services including
villagization and animal health services.
• Avoid the use of campaign-based approach to influence
Drop
behaviors that need continuous communication with household
members such as latrine construction and use.
• Avoid punishment or coercion measures for not having facilities
or using services
page- 451
3. Implementation of the Health Extension Program
CONCLUSIONS
• HPs are almost universally available at the kebele level and physically
accessible for the vast majority of the community. In some cases, however,
the availability and physical accessibility of HPs has not translated into
actual access to services.
• Exposure to the HEP among the agrarian and pastoralist communities
is low. Despite the relatively high level of importance of home visits and
outreach sessions to bring about household-level behavioral changes,
exposure to the HEP is shifting toward HP-based services over household
and community-based health-promotion and disease-prevention
activities.
• Implementation of the HEP has been very slow in pastoralist communities
compared to agrarian communities.
• Human-resource-related factors are likely to be the primary drivers of
the intensity of HEP implementation. The professional mix and levels of
education, rather than the number of HEWs, in an HP, are associated
with better implementation of the HEP through home and HP visits. HPs
with midwives/nurses or level IV HEWs had better implementation of
the HEP in terms of both home visits and HP visits.
• Progress toward full implementation of the HEP at the household level is
sub-optimal. HEP-related factors explained only a small amount of the
variation in the household-level implementation of the HEP, signifying a
lack of effectiveness of current behavioral-change strategies to achieve
household-level behavioral changes.
page-452
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Universal availability of HPs at the kebele level
• Model family training as a strategy for HEP implementation
• Behavior change strategies should be adapted to behavioral

outcome and context specific approaches/models. HEP is a complex
program targeting multiple outcomes in very diverse contexts; no
single behavior change model will allow transformation of all HEP
related behaviors.
• Strengthen linkage between demand creation and service provision
activities by increasing availability of services at health post level
and further enhancing health center – health post linkage.
Modify
• Home visits and most of the outreach sessions of HEWs should
focus on demand creation through health and health system literacy
instead of attempting to take facility-based services to the home of
potential users.
• Intensify focused outreach services to selected areas where
men and youth can be targeted (markets, schools, periodic
community gatherings, religious institutions, and community-based
organizations) depending on local context.
• Expand workforce at health posts by number and professional
mix to ensure that HEWs have adequate time for home visits and
outreach sessions while maintaining fulltime operation of health
posts.
• Arrange flexible but regulated working days and working hours to
allow HEWs to plan and reach target populations including women,
men, and youth in different public gatherings such as market days,
Add religious gatherings, and other social events.
• Consider a phased approach to implementation of HEP packages
through which each package that requires behavior change at
community level will be a focus area of intervention at different
time periods during which intensified social and behavior change
strategies will be implemented until a sustainable change is
observed. This will allow avoiding unnecessary spread of HEP
resources over too many activities at a time.
Drop • Forcing households to adopt desired behaviors
• Coercion/punishment as a strategy to influence household behavior
• Campaign based approaches for behaviors that require time to
change
page- 453
4. Human resources for the Health Extension Program
CONCLUSIONS
• Most HPs are staffed with at least 2 HEWs, but the current workforce
has challenges related to competence and motivation.
• Gaps in the competence of HEWs are primarily linked to sub-optimal pre-
service training in the 1) recruitment of trainees, 2) medium of instruction
in colleges, 3) training capacity of institutions as opposed to large class
sizes, and 4) limited compliance of trainings with training curricula.
• Several HPs have more than 2 HEWs and at least 1 level IV HEW; some
have nurses or midwives. The availability of at least 1 level IV HEW,
midwife, or nurse is associated with better implementation of the HEP,
but an increase in the number of HEWs within an HP was not associated
with better performance.
• HCs are attempting to fill the skill gap at the HP level by assigning their
staff to rotate at HPs. This approach has been criticized for its logistical
challenges and the inefficiency associated with travel time.
• The introduction of additional interventions over time markedly
increased the workload of already strained HEWs. Full implementation
of the current HEP packages requires more health workers in each HP.
The current HEP packages require skill sets in diverse areas of health
disciplines that can be broadly categorized as midwifery, clinical, and
environmental-health-related skills. A single category of health worker is
unlikely to have mastered all the required skills.
• The attrition rate is fairly low despite the high intention to leave
among HEWs, implying that there is a high level of work dissatisfaction,
the retention of only less competent staff over time, and the high risk of
losing a large number of HEWs if alternative job opportunities emerge.
Increasing the satisfaction level of HEWs requires a comprehensive
package of incentives that respects their rights as civil servants and
allows them to grow professionally to more diverse fields of related
specialties.
page-454
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Upgrade level III HEWs to level IV
• Revise entrance criteria for HEW training to consider

opportunities created by large numbers of students completing
high school and university preparatory schools. Introduce
entrance exams for HEW training institutions.
• Build the capacity of HEW training institutions in the areas
of involvement in student recruitment, instructors’ capacity,
management of practical attachment programs, and skill labs.
• Strengthen regulation of HEW training institutions.
• Review and balance duration of training for HEWs with content
of curriculum
• Match practical attachment sites with learning outcomes.
Consistently assign trainees at health post level as part of
practical attachment.
• Strengthen the provision of IRT on regular basis using training
Modify
materials translated into local languages whenever possible.
• Respect the rights of HEWs as civil servants in the areas of
transfers, leave, and career structure.
• Transform workflow and information system of health posts in a
way that guarantees continuum of care that is resilient to staff
turnover. This will require making CHIS a more dependable
source of information about households than the memory of
individual HEWs.
• Revise human resource standards of health post to allow
assignment of more health workers in each health post.
• Mobilize underused staff of health centers to work in health posts
until adequate health post capacity is built. Develop incentive
packages to motivate HC staff to work temporarily at health
posts.
page- 455
• Address the language barrier in training HEWs by introducing

English-language competency tests for entrance.
• Open career development for HEWs to allow them grow in
more diversified areas of specialties allowing competent HEWs
to compete and occupy positions in other levels of health
institutions.
• Consider staffing health posts with a team of health workers
composed of HEWs and other health professionals with expertise
allowing the provision of more comprehensive services at health
post level. Assigning more than 2 HEWs in a health post
may not be an effective way of using limited public resources.
Consider adding health workers with additional set of expertise
Add
in response to the need to increase number of health workers
in a health post. The willingness of Federal and Regional
Governments to assign more HEWs in each health post is an
opportunity that can be redirected to diversifying skills and
gender at health post while at the same time alleviating the
workload of existing HEWs.
• Initiating virtual learning modalities for HEWs as a continuous
professional development strategy.
• Provide simple technology applications serving as job aid and
decision support tools.
• Consider performance-based incentives to health posts and
HEWs based on auditable performance data.
5. Physical facilities, infrastructure, and basic utilities
CONCLUSIONS
• HPs are almost universally available.

• Most available HPs do not meet the standards for infrastructure, physical
facilities, or basic utilities.
page-456
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Health center to health post linkage to overcome challenges
related to lack of electric power at health posts.
• Responses to the increasing population size within a kebele

should focus on expanding capacity within a health post instead
of constructing additional health posts.
• A phased approach to renovation/reconstruction of health posts
should be introduced with due consideration to: 1) the need to
expand services 2) the importance of avoiding any more sub-
standard construction, 3) the limited capacity of the country, and
4) the availability and accessibility of infrastructure and utilities
within the kebele
• Coordinate efforts to renovate or reconstruct health posts in line
with plans for expansion of services within each PHCU.
Modify • Initiate an innovative approach to mobilize resources for
renovation of health posts from government, community, and
other funding sources.
• Multi-sectoral approach: Negotiate at a higher level to ensure
that health posts are prioritized in infrastructure development
projects (road, electricity, water, and telecommunication)
targeting rural communities.
• Consider long term plans to solve lack of residential houses
for health post staff. Actions should include multiple options
including constructing residential houses in health post
compounds, incentivizing private leasers, and facilitating access
to land for HP staff intending to construct their own houses.
Add
• Introduce enforcement of regulatory standards on future
health post construction and/or renovation activities to prevent
investment on sub-standard constructions.
page- 457
6. Equipment, drugs, and other medical supplies
CONCLUSIONS
• The essential equipment required for the provision of services under the
current packages is very often unavailable or non-functional at HPs.
• The availability of tracer drugs varied by item. Both the shortage of
supplies and the inadequacy of the supply management system were
related to stockouts of tracer drugs and other medical supplies.
• The lack of availability of functional medical equipment possibly explains
the compromised quality of HP-based services, including the low effective
coverage of ANC.
page-458
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• The supply of program-specific drugs and other medical supplies,
like family planning commodities and vaccine supplies.
• Strengthen IPLS implementation through regular supportive

supervision and introduction of simple electronic technologies.
• Revise drug list of health posts to match revision in scope of
services provided at health posts including possible expansion of
clinical services.
• Build the capacity of health post staff on supply chain
Modify management system for drugs and other medical supplies.
• Ensure appropriate storage and usage of drugs and other
medical supplies.
• Strengthen quality assurance of imported medical equipment
including BP apparatus.
• Avail durable and quality assured equipment at health posts and
strengthen continuous maintenance
• Explore and introduce alternative sources of funding the supply

of drugs and other medical supplies for consumption at health
post level.
• Assess the feasibility and effectiveness of alternative for
financing HP based services through mechanisms including
Add
community-based health insurance and incentivizing private
sector involvement at the village level.
• Explore options for expansion of investigations involving dip-
stick technologies to address the gap in laboratory facilities as
services expand at health posts.
page- 459
7. Financing the Health Extension Program
CONCLUSIONS
• Investment in the HEP has been increasing in nominal terms. The share
of spending on the HEP in relation to total expenditure at the HC and
HP levels, however, has been continuously declining since 2010.
• In addition to voluntary community contribution of time and labor,
government and donors are the major financing sources of HEP. The
government’s share in financing the HEP has been increasing over the
years, but the HEP is still a highly donor-dependent program, with 77%
of its spending coming from external sources.
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Expanding government share in financing for HEP
Modify
• Increase the rate at which domestic financing schemes substitute
donors with the ambition of ensuring financial sustainability of
HEP.
Add
• Consider alternative sources of financing HEP packages
including CBHI.
page-460
8. Community engagement and ownership
CONCLUSIONS
• Community participation and ownership has been an important

component of the HEP throughout the life of the program.
• Model family training is an effective strategy for increasing the household
level implementation of the HEP. Only a very small portion of the
population, however, are aware of, have enrolled in, or have completed
the training.
• WDA and/or SMC structures are widely available. Their functionality in
supporting the HEP is, however, very limited.
• Among the primary reasons for this are:
o WDA leaders and SMC members are not models in their health
behaviors.
o The selection of WDA leaders gives very little attention to health
behavior.
o WDA leaders currently have low acceptance and are mostly
considered political agents.
o The roles and responsibilities of WDA leaders sometimes outpace
their capacity, with outsized expectations being placed on them
while ignoring that they are volunteers.
• The use of WDAs alone has resulted in the underuse of community

potential, including that of men, religious leaders, and traditional leaders.
page- 461
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Keep community engagement central to the HEP
Modify • Strengthen model family training by providing clear guidelines,

increasing HEWs’ time spent for training of families and
arranging experience sharing sessions between model families
and others.
• All community volunteers working with HEWs should be selected
only among model families.
• Introduce a system that allows HEWs to track enrollment,
progress, completion, and recognition of model families.
• Redesign community structure for HEP with renewed branding,
capacity, and responsibilities. Consider the following features to
address challenges faced by the WDA approach:
o Link HPs with all segments of the kebele population
including men, women, and youth with different roles and
positions in the society
o Allow community structures to vary across regions and within
regions depending on culture and functionality of existing
structures.
o Involve all segments of the population including youth,
women, men, traditional leaders, religious leaders, and other
Add
influential individuals.
o Use fewer (a manageable size) community volunteers to
serve as change agents and community mobilizers
o Keep community volunteers accountable to HEWs.
• Incentivize volunteerism and limit duration of service to a pre-
defined period of performance. Recognition and free enrollment
to CBHI scheme are among possible approaches to incentivize
voluntary health agents.
• Make maximal use of opportunities created by: 1) relatively
better availability of literate community members, 2) high level
of school enrollment among adolescents and youth, and 3)
increasing use of communication technologies including cellphone
and the internet.
Drop • Reliance on single approach to community participation

• Avoid creating expectations of becoming salaried workers among
community volunteers
page-462
9. Information system and monitoring & evaluation
CONCLUSIONS
• The current health information/M&E system that captures data for

measuring indicators reportable up to the federal level focuses only on
the outputs of the specific programs implemented through the HEP,
with very limited attention paid to monitoring the process of the HEP at
lower levels.
• The kebele-level indicators that are directly linked to the performance of
the HEP involve definitions with unrealistic targets (e.g., HDF, ODF, 100%
CBHI enrollment), resulting in a lack of sensitivity to the intermediate
progress of HPs.
• The data recorded and reported by HPs are largely inconsistent with
source documents, mostly resulting in the over-reporting of performance.
• The use of information is limited at the HP and higher levels.
page- 463
RECOMMENDATIONS
Level of change
Suggested changes
required
Maintain
• Data disaggregation by level of service provision (HP, HC)
Modify
• Revise definitions of indicators with unrealistic targets.
• Expand electronic CHIS with dashboard features facilitating
information use in situations with limited data processing
capacity.
• Enforce the use of family folders to record encounters between
HEWs and household members as per the guideline.
• Include process indicators of HEP for monitoring & implement-

ation of HEP service delivery modalities at least at health post,
healthcenter, and woreda levels.
• Establish data verification system including community level
verification on a random sample of service users as well as
introduction of innovative technologies in order to minimize
deliberate over reporting.
• Introduce performance management system that relies on
Add objective auditing of coverage and quality of services.
• Initiating incentive mechanisms to encourage improved data
quality and use.
• Transform information system of health posts in a way that
guarantees continuum of care that is resilient to staff turnover.
This will require making CHIS a more dependable source of
information about households than the memory of individual
HEWs.
Drop
• Eliminate formal and informal incentives to over-reporting
page-464
10. Governance and leadership
CONCLUSIONS
• There has been limited guidance on how the HEP should evolve over
time.
• The major decisions made about the HEP have not taken advantage
of the opportunities created to generate and use evidence on the
effectiveness of the HEP.
• The dual accountability of HEWs and parallel reporting are common
among HPs.
• A single standard for HPs didn’t fit the realities of populations at the
kebele level. The services provided, staffing patterns, supplies and
equipment, and infrastructure needs of the HEP in kebeles with an HC
are different from those located far from an HC. The current standard
of HPs did not acknowledge this difference, leading to the inefficient use
of available resources.
• Supervisory support from HCs has been inadequate. Whenever provided,
the team-based supervision of HPs has been more supportive of the
HEP’s implementation than individual HEP supervisors.
• Restrictions in the rights of HEWs as civil servants have frequently been
reported. This has been a major source of dissatisfaction among HEWs.
• Accountability is limited at the HP level, leading to high rate of
absenteeism and the closure of HPs.
page- 465
RECOMMENDATIONS
Level of change Suggested changes

required
Maintain
• Health center to health post linkage for technical support and

administrative oversight.
Modify
• Clarify lines of accountability of HEWs/HPs to avoid dual
accountability and overlapping responsibilities.
• Ensure alignment of priorities and targets of different health
programs with those of HEP.
• Strengthen intersectoral collaboration at all levels guided
by collaborative frameworks enforced at higher levels by the
leadership and management.
• Ensure that HEP plays vital role in facilitating Kebele level
intersectoral collaboration with the intention of addressing social
determinants of health.
page-466
• Introduce service and input standards for multiple categories

of HEP implementers. One option could be classifying HEP
structures as:
o HEP unit in a health center (for kebeles with HCs)
o HPs implementing basic set of packages
o HPs implementing comprehensive set of packages
• Develop a roadmap of evolution of HEP over the next10 to 20
years.
• Rebrand health posts along with changes in their function and
structure. Consider changing names from health post to a one
that reflects upgrading to a facility with more comprehensive
services in order to boost demand for both existing and newly
added services.
• Consider establishing and testing administrative boards for
health posts involving community members to enhance the
oversight role of the community.
Add • Provide clear guidelines on involvement of HEWs in “non-health”
activities with the purpose of keeping their involvement in
activities that:
o Create opportunities for health promotion and disease
prevention and address social determinants of health
o Do not create negative co-notation for HEWs by any
member of the society
o Can be pre-planned to avoid too much compromise in
routines of HEP
o Increase the power and acceptance of HEWs
o Are planned in a framework that facilitates
intersectoral collaboration
• Introduce a standard decision-making procedure that requires
generation and use of adequate evidences before making
decisions on major investments in HEP.
page- 467
References
1. Institute for Health Metrics and Evaluation. Global Burden of Disease

(GBD) Country Profile for Ethiopia. http://www.healthdata.org/
ethiopia. Published 2017. Accessed 20 March 2019.
2. Federal Democratic Republic of Ethiopia Ministry of Health. Health
Sector Development Program IV (2010/11 - 2014/15). In:2010.
Sector Development Program III - 2005/6 - 2009/10. 2005.
Sector Development Program II: 2002/03 – 2004/05. In:2002.
Sector Development Program I - 1997/98 - 2001/02. In:1997.
6. FMoH. Health Sector Transformation Plan (HSTP) 2014/15 - 2019/20.
In:2015.
7. Millennium Development Indicators: Country and Regional Progress
Snapshots. United Nations; 2015. https://unstats.un.org/unsd/mdg/Host.
aspx?Content=Data/snapshots.htm. Accessed 20 July 2019.
8. WHO. Declaration of Alma-Ata. Paper presented at: International
Conference on Primary Health Care1978; Alma-Ata, USSR.
9. Central Statistical Agency [Ethiopia], The DHS Program ICF Rockville.
Ethiopia Demographic and Health Survey 2016. In:2017.
10. EPHI, MoH. Ethiopia Mini-Demographic and Health Survey Key
Indicators. In:2019.
11. FMOH. Health Policy of the Transitional Government of Ethiopia. In.
Addis Ababa1993.
12. African Union. Africa Health Strategy: 2007-2015. In:2007.
page-468
13. OECD. Paris Declaration on Aid Effectiveness. 2005.
14. OECD. Accra Agenda for Action. 2008.
15. Africa Union. Abuja Declaration on HIV/AIDS, Tuberculosis, and
Other Related Infectious Diseases. In:2001.
16. WHO. Task shifting: rational redistribution of tasks among health
workforce teams: Global recommendations and guidelines. In.
Geneva2008.
17. Lin V GY, Legge D, Wu Q.,. Health policy in and for China. 1st edition.
In. Beijing, China: Peking University Medical Press; 2010.
18. Hu D, Zhu W, Fu Y, et al. Development of village doctors in China:
financial compensation and health system support. International
journal for equity in health. 2017;16(1):9.
19. GHWA Task Force on Scaling Up Education and Training for Health
Workers. Pakistan’s Lady Health Worekrs Programme: Country case
study. In:2008.
20. Witter S, Anderson I, Annear P, et al. What, why and how do health
systems learn from one another? Insights from eight low- and middle-
income country case studies. Health research policy and systems /
BioMed Central. 2019;17(1):9.
21. Admasu K, Balcha T, Ghebreyesus T. Pro-poor pathway towards
universal health coverage: Lessons from Ethiopia. Journal of global
health. 2016;6.
22. Bilal NK HC, Zhao F et al.,. Health extension workers in Ethiopia:
improved access and coverage for the rural poor. Yes Africa Can:
Success Stories from a Dynamic Continent, 433-443. 2011.
23. EPHI. Improving the Health Extension Program in Ethiopia: An
evidence-based policy brief. In:2014.
24. Wang HT, Roman; Ramana, Gandham N.V.; Chekagn, Chala Tesfaye.
Ethiopia Health Extension Program : An Institutionalized Community
Approach for Universal Health Coverage. World Bank Studies;.
page- 469
Washington, DC: World Bank. © World Bank. License: CC BY 3.0 IGO.
2016.
25. Wang H, Tesfaye R, Ramana GNV, Chekagn CT. Ethiopia Health
Extension Program: An Institutionalized Community Approach for
Universal Health Coverage. 2016.
26. United Nations. The Millennium Development Goals Report: We can
end poverty 2015, Millennium Development Goals. New York: United
Nations;2015.
27. Central Statistical Agency, ORC Macro. Ethiopia Demographic and
Health Survey 2000. In:2001.
28. Central Statistical Agency, ORC Macro. Ethiopia Demographic and
Health Survey 2005. In:2005.
29. FMOH. Guideline for Implementation of Health Extension Program. In.
Addis Ababa, Ethiopia2008.
30. Admasu K, Balcha T, Getahun H. Model villages: a platform for
community-based primary health care. The Lancet Global health.
2016;4(2):e78-79.
31. Damtew ZA, Karim AM, Chekagn CT, et al. Correlates of the Women’s
Development Army strategy implementation strength with household
reproductive, maternal, newborn and child healthcare practices: a
cross-sectional study in four regions of Ethiopia. BMC pregnancy and
childbirth. 2018;18(Suppl 1):373.
32. Assefa Y, Gelaw YA, Hill PS, Taye BW, Van Damme W. Community
health extension program of Ethiopia, 2003-2018: successes and
challenges toward universal coverage for primary healthcare services.
Globalization and health. 2019;15(1):24.
33. PHCPI. The PHCPI Conceptual Framework. 2018. https://improvingphc.
org/phcpi-conceptual-framework. Accessed 28 Feb 2019.
34. Veillard J, Cowling K, Bitton A, et al. Better Measurement for
Performance Improvement in Low- and Middle-Income Countries: The
page-470
Primary Health Care Performance Initiative (PHCPI) Experience of
Conceptual Framework Development and Indicator Selection. Milbank
Q. 2017;95(4):836-883.
35. Ethiopian Public Health Institute, Federal Ministry of Health, World
Health Organization. Ethiopia Services Availability and Readiness
Assessment 2016. In:2016.
36. Chianca T. The OECD/DAC Criteria for International Development
Evaluations: An Assessment and Ideas for Improvement Journal of
MultiDisciplinary Evaluation. 2008;5(9).
37. OECD. Guidelines for Project and Programme Evaluations In:2009.
38. Teklehaimanot HD, Teklehaimanot A, Tedella AA, Abdella M. Use
of Balanced Scorecard Methodology for Performance Measurement
of the Health Extension Program in Ethiopia. Am J Trop Med Hyg.
2016;94(5):1157-1169. doi:10.4269/ajtmh.15-0192.
39. Desta, F.A., Shifa, G.T., Dagoye, D.W. et al. Identifying gaps in the
practices of rural health extension workers in Ethiopia: a task analysis
study. BMC Health Serv Res 17, 839 (2017). https://doi.org/10.1186/
s12913-017-2804-0
40. FMOH. Second generation HEP implementation framework. In. Addis
Ababa, Ethiopia2016.
41. Assefa Y, Tesfaye D, Damme WV, Hill PS. Effectiveness and
sustainability of a diagonal investment approach to strengthen the
primary health-care system in Ethiopia. Lancet. 2018;392(10156):1473-
1481.
42. Karim AM, Admassu K, Schellenberg J, et al. Effect of ethiopia’s health
extension program on maternal and newborn health care practices in
101 rural districts: a dose-response study. PloS one. 2013;8(6):e65160.
43. FMOH. NATIONAL HUMAN RESOURCE FOR HEALTH
STRATEGIC PLAN FOR ETHIOPIA 2016-2025. 2016.
44. FMOH. National Health Workforce Update: Human Resource
Development Directorate Bulletin second edition ed 2019
page- 471
45. Ministry of Health [Ethiopia]. Health and Health-Related Indicators
EFY2010 (2017/18). In:2018.
46. FMOH. Brief Guide on the implementation of Second Generation
Health Extension Program (HEP). 2018.
47. Workie NW, Ramana GN. The Health Extension Program. In:2013.
48. Rotenstein LS, Torre M, Ramos MA, et al. Prevalence of Burnout
Among Physicians: A Systematic Review. JAMA : the journal of the
American Medical Association. 2018;320(11):1131-1150.
49. Hanlon C, Medhin G, Selamu M, et al. Validity of brief screening
questionnaires to detect depression in primary care in Ethiopia. Journal
of affective disorders. 2015;186:32-39.
50. Levis B, Benedetti A, Thombs BD. Accuracy of Patient Health
Questionnaire-9 (PHQ-9) for screening to detect major depression:
individual participant data meta-analysis. Bmj. 2019;365:l1476.
51. Ministry of Health [Ethiopia]. Health and Health Related Indicators
2011EC (2018/19). In:2019.
52. ESA. Ethiopian Standards. Health post requirements, 2012. 2012.
53. FMoH. Ethiopia Health Accounts, 2013/14. In:2017.
54. WHO, UNICEF. Progress on drinking water, sanitation and hygiene:
2017 update and SDG baselines. In. Geneva2017.
55. WHO. Indicators for assessing infant and young child feeding practices
Part 2: Measurement. In: World Heal Organization; 2007.
56. WHO, UNICEF, UNFPA, World Bank Group, United Nations
Population Division. Trends in maternal mortality: 1990 to 2015:
estimates by WHO, UNICEF, UNFPA, World Bank Group and the
United Nations Population Division. In:2015.
57. Medhanyie A, Spigt M, Dinant G, Blanco R. Knowledge and
performance of the Ethiopian health extension workers on antenatal
and delivery care: a cross-sectional study. Human resources for health.
2012;10(1):44.
page-472
58. Afework MF, Admassu K, Mekonnen A, Hagos S, Asegid M, Ahmed S.
Effect of an innovative community based health program on maternal
health service utilization in north and south central Ethiopia: a
community based cross sectional study. Reproductive health. 2014;11:28.
59. Yitayal M, Berhane Y, Worku A, Kebede Y. Health extension program
factors, frequency of household visits and being model households,
improved utilization of basic health services in Ethiopia. BMC health
services research. 2014;14(1):156.
60. Jackson R, Hailemariam A. The Role of Health Extension Workers in
Linking Pregnant Women With Health Facilities for Delivery in Rural
and Pastoralist Areas of Ethiopia. Ethiopian journal of health sciences.
2016;26(5):471-478.
61. Medhanyie A, Spigt M, Kifle Y, et al. The role of health extension
workers in improving utilization of maternal health services in rural
areas in Ethiopia: a cross sectional study. BMC health services research.
2012;12:352.
62. Gebrehiwot TG, San Sebastian M, Edin K, Goicolea I. The Health
Extension Program and Its Association with Change in Utilization
of Selected Maternal Health Services in Tigray Region, Ethiopia: A
Segmented Linear Regression Analysis. PloS one. 2015;10(7):e0131195.
63. Yitayal M, Berhane Y, Worku A, Kebede Y. The community-based
Health Extension Program significantly improved contraceptive
utilization in West Gojjam Zone, Ethiopia. Journal of multidisciplinary
healthcare. 2014;7:201-208.
64. Gebre-Egziabher D, Medhanyie AA, Alemayehu M, Tesfay FH.
Prevalence and predictors of implanon utilization among women
of reproductive age group in Tigray Region, Northern Ethiopia.
Reproductive health. 2017;14(1):62.
65. Negussie A, Girma G. Is the role of Health Extension Workers in
the delivery of maternal and child health care services a significant
attribute? The case of Dale district, southern Ethiopia. BMC health
services research. 2017;17(1):641.
page- 473
66. Gebretsadik A, Teshome M, Mekonnen M, Alemayehu A, Haji Y.
Health Extension Workers Involvement in the Utilization of Focused
Antenatal Care Service in Rural Sidama Zone, Southern Ethiopia:
A Cross-Sectional Study. Health Serv Res Manag Epidemiol.
2019;6:2333392819835138.
67. Jackson R, Tesfay FH, Godefay H, Gebrehiwot TG. Health Extension
Workers’ and Mothers’ Attitudes to Maternal Health Service Utilization
and Acceptance in Adwa Woreda, Tigray Region, Ethiopia. PloS one.
2016;11(3):e0150747.
68. Seyoum A, Urgessa K, Gobena T. Community knowledge and the role
of health extension workers on integrated diseases among households
in East Hararghe Zone, Ethiopia. Risk Manag Healthc Policy.
2016;9:135-142.
69. Admassie A, Abebaw D, Woldemichael AD. Impact evaluation of
the Ethiopian Health Services Extension Programme. Journal of
Development Effectiveness. 2009;1(4):430-449.
70. Datiko DG, Yassin MA, Theobald SJ, Cuevas LE. A community-
based isoniazid preventive therapy for the prevention of childhood
tuberculosis in Ethiopia. The international journal of tuberculosis and
lung disease : the official journal of the International Union against
Tuberculosis and Lung Disease. 2017;21(9):1002-1007.
71. Teklehaimanot HD, Teklehaimanot A, Yohannes M, Biratu D. Factors
influencing the uptake of voluntary HIV counseling and testing in rural
Ethiopia: a cross sectional study. BMC public health. 2016;16:239.
72. Gebru T, Taha M, Kassahun W. Risk factors of diarrhoeal disease in
under-five children among health extension model and non-model
families in Sheko district rural community, Southwest Ethiopia:
comparative cross-sectional study. BMC public health. 2014;14:395.
73. Ashenafi A, Karim AM, Ameha A, Erbo A, Getachew N, Betemariam
W. Effect of the health extension program and other accessibility
factors on care-seeking behaviors for common childhood illnesses in
rural Ethiopia. Ethiopian medical journal. 2014;52 Suppl 3:57-64.
page-474
74. Mathewos B, Owen H, Sitrin D, et al. Community-Based Interventions
for Newborns in Ethiopia (COMBINE): Cost-effectiveness analysis.
Health policy and planning. 2017;32(suppl_1):i21-i32.
75. Getnet F, Hashi A, Mohamud S, Mowlid H, Klinkenberg E. Low
contribution of health extension workers in identification of persons
with presumptive pulmonary tuberculosis in Ethiopian Somali Region
pastoralists. BMC health services research. 2017;17(1):193.
76. Ross RK, King JD, Damte M, et al. Evaluation of household latrine
coverage in Kewot woreda, Ethiopia, 3 years after implementing
interventions to control blinding trachoma. International health.
2011;3(4):251-258.
77. Crocker J, Saywell D, Bartram J. Sustainability of community-led total
sanitation outcomes: Evidence from Ethiopia and Ghana. Int J Hyg
Environ Health. 2017;220(3):551-557.
78. Mengistu B, Karim AM, Eniyew A, et al. Effect of performance review
and clinical mentoring meetings (PRCMM) on recording of community
case management by health extension workers in Ethiopia. Ethiopian
medical journal. 2014;52 Suppl 3:73-81.
79. Marsh DR, Hazel E, Nefdt R. Integrated Community Case
Management (iCCM) at Scale in Ethiopia: Evidence and Experience.
Ethiopian medical journal. 2014;52.
80. Birhanu Z, Godesso A, Kebede Y, Gerbaba M. Mothers’ experiences
and satisfactions with health extension program in Jimma zone,
Ethiopia: a cross sectional study. BMC health services research.
2013;13:74.
81. Banteyerga H. Ethiopia’s health extension program: improving health
through community involvement. MEDICC review. 2011;13(3):46-49.
82. National Planning Commission [Ethiopia]. Growth and Transformation
Plan II (GTP II) 2015/16 - 2019/20. In:2016.
83. MOH. Essential Health Service Packages of Ethiopia. 2019.
page- 475
Annex : Methods of Analysis of HEP
Financing Landscape
INTRODUCTION
Ethiopia has made a profound investment in strengthening its health system,

guided by its pro-poor policies and strategies in the past two decades. As a
result, significant gains have been made in improving health outcomes, including
a substantial reduction in maternal, neonatal, infant, and child mortality.
Mortality and morbidity due to HIV/AIDS, tuberculosis (TB), and malaria have
all been reduced significantly and, as a result, have contributed to an increase
in average life expectancy at birth from 45 years in 1990 to 64 years in 2014.1
The Health Extension Program (HEP) is the country’s flagship program and
the principal vehicle for expanding access to essential health services packages
to rural communities, where nearly 80% of the country’s population resides, with
a particular focus on women and children.1 The program was initially exclusively
government financed, demonstrating the government’s strong commitment to
ensuring access to healthcare.2 Implementation of the program began in 2004
as part of the second phase of the Health Sector Development Program and
is ongoing, entailing 16 health packages, which are categorized into four major
components: promotion of hygiene and environmental sanitation, prevention
and control of major communicable diseases (CDs), promoting and providing
family health services, and health education and communication.1 As a result,
the HEP has introduced a new cadre of health workers, called Health Extension
Workers (HEWs), who are to deliver defined packages of essential interventions
from village health posts (HPs) free of charge. More than 38 000 HEWs have
been trained and deployed. The HEP is also supported by an organized and
functional Women Development Army (WDA), which enhances the access to
and use of key health interventions. The HDA was launched in 2011 to further
strengthen the HEP and sustain its gains.
1 MoH. Health Sector transformation plan, 2015/16-2019/20 Addis Ababa. Ethiopia. October 2015
2 World Health Organization, GHWA task force on scaling up education and training for health
workers. Country case study: Ethiopia’s Human Resources for Health Programme. https://www.
who.int/workforcealliance/knowledge/case_studies/Ethiopia.pdf
page-476
HDA leaders are selected from model families to complement the work of the
HEWs within 1-to-5 networksand the 1:30 household ratio.3,4
Governmental commitment, together with strong leadership, community

ownership of health programs, active participation, and development partners’
unprecedented financial and technical support, has resulted in remarkable
gains over the last two decades.1 Ethiopia has increased its health resources
exponentially in terms of the overall share of domestic financing for health.
These resource improvements include the following.
First, over the last decade, the total expenditure on health has increased
dramatically, improving most Ethiopians’ quality of life. As shown in Figure
1, total nominal health expenditure (THE) has grown from US $0.23 billion
(ETB 1.4 billion) in 1995-96 to US $3.1 billion (ETB 72.05 billion) in 2015-
16), along with the steady growth of health expenditure per capita, which
increased from $4.50 in 1995-96 to $33.2 in 2016-17. Although this growth is
encouraging, the amount is still low compared with that of peer countries and
the recommendations of the World Health Organization (WHO), which has
suggested that US $86 per-capita spending would be needed for the delivery
of essential health services by 2015.5 In addition, health spending has ranged
between 3.5% and 5.2% of the Ethiopian gross domestic product (GDP) in
recent years. This is less than the 7% estimated by a recent WHO report as the
average for low-income countries.6
The substantial financial support from Ethiopia’s development partners has

become a major source of health sector financing. Ethiopia’s health sector is
financed from three major sources: government (32%), households in the form of
out-of-pocket payments (31%), and the rest of the world (35%) in EFY 2016-17.
Over the last decade (2007/08 to 2016/17), the share of government spending
increased from 21% to 32%. Household out-of-pocket spending remains a major
source of domestic financing, with a slight decrease in contribution from 33%
to 31%. By contrast, financing from abroad, although still a major source of
funding, has been declining significantly; it increased from 39% in 2007/08
to 50% in 2010-11, and declined to 35% in 2016-17. Regardless, financing from
abroad accounts for more than one third of the total health expenditure, making
3 MoH. Annual performance report of HSDP-III. EFY 1999 (2006/2007). October 2007.
4 MoH. Annual health sector performance report. Executive Summary.Version-1 EFY 2009
(2016/2017).
5 Jowett, M, Brunal, MP, Flores, G, Cylus, J. Spending Targets for Health: No Magic Number. 2016.
6 World Health Organization. New Perspectives on Global Health Spending for Universal Health
Coverage. Conference Copy for Consultation. 2017.
page- 477
it a major source of health sector financing and considerably higher than the
average low-income country share of 28%.7,8 Figure 3. Trends in total health
expenditure (THE), 2007/08 to 2016/17 (in billions of ETB).
Figure 1. Trends in total health expenditure (THE) by source (%), 2007/08 to

2016/17
Figure 3. Trends in total health expenditure (THE), 2007/08 to 2016/17 (in

billions of ETB)
7 MoH. Ethiopia Health Accounts, 2013/14. August 2017.

8 MoH. Ethiopia Health Accounts Report, 2016/17. August 2019
page-478
Figure 4, Trends in per-capita total health expenditure (THE), and THE as a
% of GDP, 2007/08 to 2016/17. (Source: Ethiopia National Health Account
Report, 2019)
The study provides information on trends in HEP spending over the last decade
and identifies the main funding flows by source, service type, service input, and
economic classification. The findings are intended to support decision-making
processes in policy development, planning, and resource allocation for the HEP
across stakeholders, as well as provide evidence for advocacy and domestic
resource mobilization.
OBJECTIVES
The main objective of the study is to generate evidence on HEP spending

nationally from 2010/11 to 2016/17. The study specifically explores total HEP
spending trends as the share of total health spending, financing sources, and
spending by service type, input, and economic classification.
page- 479
METHODS
Data sources
To estimate the national spending specific to the HEP, primary and secondary
data sources were used. The primary data sources entail service lists provided
at the HP level and the inputs required to deliver the services. Secondary data
sources include the Ministry of Health (MoH) for the standard Pharmaceutical
Supply Agency (PSA) drug cost, supply and medical equipment costs, human
resources cost, and health service coverage and use, Ethiopia Health Accounts
2010/11 to 2016/17 for health expenditures, the Health Sector Transformation
Plan (HSTP) 2015/16 to 2019/20, the Central Statistics Agency (CSA) for the
annual inflation rate, and the World Bank’s 2017 database for the conversion
of Ethiopian purchasing power parity into USD, adjusted for inflation.
HEP expenditure estimation framework
The expenditure framework and analysis follow the major steps and
recommendations outlined in the guidelines of the System of Health Accounts
(SHA) from 2011. The HEP expenditure estimations focus on recurrent and
capital expenditures; recurrent expenditures consist of drugs, supplies, salaries,
and other utility costs that are incurred on a regular basis, either direct or
indirect expenditures. Direct expenditures are those expenditures that are directly
attributable to a specific service output, while indirect expenditures are spending
that cannot be attributed directly to a certain output. This expenditure analysis
includes both direct and indirect expenditures. The top-bottom approach for
estimating HEP health expenditure steps is shown in Figure 4.
page-480
Figure 4. Health Extension Program expenditure estimation phases
SCOPE
The study used both on- and off-budget data (budget allocations and
expenditures) for all HEP activities in Ethiopia, including public revenue, external
sources of funding, and community contributions. Community contribution to
the HEP through the HDAs was also considered, but, due to the insignificant
contributions of the private sector, this analysis did not capture such out-of-
pocket payment, private insurance, and employer contributions to the HEP or
the nature of the HEP services.9
LIMITATIONS AND CHALLENGES
Estimating HEP expenditure represents a significant accomplishment in the

evaluation of the program. The estimation was far-reaching and the first project
of this type to be employed. Nevertheless, it is not without limitations.
9 All HEP services are strategically exempted services.
page- 481
Several challenges were encountered during the estimation process. Reports
of expenditures by government and from abroad did not directly match the
HEP/HP level, since HPs were not considered the cost center in the Ethiopian
health system.10 Therefore, distribution keys11 were used to disaggregate the
expenditure figures by level of care (HC and HP/HEP) and health service type.
Hence, these figures are estimates, not actual figures. Different institutions kept
information in different formats, which again caused difficulty in obtaining
data that matched the HEP’s categories and levels. For instance, the datasets
requested detailed information about expenditures on the HEP level, but
most datasets did not keep information at that level of detail and, if they
did, the formats were different from those used by the HEP. For example,
although we showed the total HDA12 contribution in the NHA 2013/14 report,
the total estimate of HEP health spending did not include the value of HDA
contributions, as the health spending or health account methodology defines
health expenditure in terms of financial transactions.
10 .The HC is the last level of cost center, which is the HC surrounded by usually five satellite HPs
11 The distribution keys are estimated based on the health services use data and unit cost study
reports obtained from the MoH and are used according to the HEP classification and
definition.
12 The community-based health delivery platform creates a network between five households and
one model family; they encourage one another to practice a healthy lifestyle and
work to empower the community to generate its own health.
page-482
page- 483
PART 3
Urban Health Extension Program
Assessment
This part presents the methods and findings of the national

assessment of the UHEP. It has seven chapters:
Chapter 1: Methods of the Urban Health

Extension Program Assessment
Chapter 2: Findings: Relevance of the UHEP
Chapter 3: Findings: Availability and adequacy of

resources for UHEP
Chapter 4: Findings: UHEP workforce analyses
Chapter 5: Findings: Implementation of the UHEP
Chapter 6: Conclusions and Recommendations
page-484
1 Methods of the Urban
CHAPTER 1
HEP Assessment
1.1 Study Design
The UHEP assessment employed a concurrent mixed-methods approach using

both quantitative and qualitative methods. The household-level quantitative
survey involved the collection and analysis of primary data from female
respondents who are heads of households in female-headed households or
spouses of heads in male-headed households. Heads of sample HCs and
sample UHEPrs were included in the HC and UHEPr assessments, respectively.
KIIs were conducted with health managers, experts, and service providers from
different levels of the health system including the MoH, RHBs, ZHDs, sub-city
health offices, WorHOs/town health offices, and HCs). Moreover, IDIs were
conducted among community members (including WDA leaders, women from
households that are UHEP service users/model households, and women and
men from non-service user households).
1.2 Study area and period
The study was conducted from October 2018 to July 2019. All 9 regions and the
2 city administrations were included in the qualitative study, HC assessment,
and urban HEPr survey. The household survey was undertaken only in the
Addis Ababa and Dire Dawa city administrations.
1.3 Source population
All HCs, HEPrs, community members (UHEP service users and non-users),
WDA leaders, administrative health facility staff, and households were the
source population for the assessment.
page- 485
Methods of the UHEP Assessment
1.4 Study participants

The study participants for the quantitative study include:
1. Heads or representatives of sample HCs in all 9 regions and all HCs in

the Addis Ababa and Dire Dawa city administrations;
2. Sample urban HEPrs from all 9 regional states, and both city
administrations;
3. Heads of households or spouses of heads of households from all sub-
cities and woredas/kebeles of the Addis Ababa and Dire Dawa city
administrations.
Purposive sampling was employed to select study participants for the qualitative
study from:
1. The Ministry of Health (directors, coordinators/team leaders, advisors
and experts);
2. RHBs (Heads/vice heads, directors, coordinators/team leaders, and
experts);
3. WorHOs (heads or representatives);
4. HCs (heads/representatives, Family Health Team leaders, and UHEPrs);
5. Partners supporting the implementation of the UHEP at different levels.
1.5 Inclusion and exclusion criteria
Households and residents with a minimum of 6 months’ stay in the study kebele
were eligible for the study. The UHEPrs with less than 1 year experience were
excluded from the study. Key informants who had inadequate exposure to and
knowledge of the UHEP were excluded.
1.6 Sample size determination and sampling

strategy for quantitative study
A separate sample size was calculated to determine the numbers of households

and HEPrs included in the quantitative study. The study primarily focuses on
the measurement of variables required for the estimation of proportions for a
single population. Therefore, the sample size was calculated using the following
formula:
page-486
ni=z2 Pi (1 – Pi) / D2
Where
ni=the required sample size for the outcome I;
Pi=the estimate of the true population proportion for variable-i (Sample
size was calculated using different variables and variable that yielded
the maximum sample size was taken);
D=level of precision.
1.6.1 Household survey
The minimum numbers of households required for the household survey was
determined using the single population proportion formula. Findings from the
2016 Ethiopia Demographic and Health Survey (EDHS) relevant to the UHEP
were used as estimates of population proportions in order to calculate the
sample size. The estimate that yielded the largest sample size—the proportion
of households with a handwashing facility—was used. The sample sizes for
Addis Ababa and Dire Dawa were calculated independently using a different
level of precision and population proportion estimate. Sample sizes were
adjusted for a design effect due to the sub-city woreda/kebele hierarchy in
selecting households. Moreover, a 90% expected response rate was considered
in calculating the sample size (Table 1-1).
Table 1-1: Sample size calculation for household survey
P (availability Adjustment for

of handwashing design effect and # of households
D
Study area facility from EDHS response rate of (final sample
(precision)
2016) 90% size)
Addis Ababa 0.674 0.04 2 1 287
Dire Dawa 0.362 0.05 1.5 625
Total sample size 1 912
Abbreviation: EDHS, Ethiopian Demographic and Health Survey.
page- 487
The maximum sample sizes of 1 287 households from Addis Ababa and 625
households from Dire Dawa were considered for the household survey.
1.6.2 Sampling Strategies for Household Survey
A systematic random sampling technique was applied to select participants

for the household survey conducted in Addis Ababa. All 10 sub-cities of Addis
Ababa were included in the household survey. Next, 90 woredas in which the
SafetyNet program was being implemented were purposely selected, with the
aim of obtaining a sampling frame with the list of poor urban residents, who
are the targets of the SafetyNet program and the primary targets of the
UHEP. The total sample size was proportionally allocated to the 90 woredas
based on their respective number of SafetyNet beneficiaries. The interval for
sample household selection was determined using a lottery method for each
woreda, dividing the total number of SafetyNet beneficiaries by the sample
allocated to the woreda. Finally, households and respondents were selected
from the list based on the interval.
Similarly, a systematic random sampling technique was applied to selected

sample households for the survey conducted in Dire Dawa. First, all 8 kebeles
of the city administration were included; then the total sample size of 625 was
proportionally allocated to each kebele based on the number of households in
each kebele. Registration books used for ITN distribution for the current fiscal
year were obtained from each kebele and used independently as a sampling
frame. The interval for the sample household selection was determined using a
lottery method for the 8 kebeles, dividing the total number of households by the
sample allocated to the respective kebele. Finally, households and respondents
were selected from the list based on the interval (Figure 1-1).
page-488
Figure 1-1. Schematic representation of sampling strategies
1.6.3 Sample size for urban HEPr assessment
A single population proportion formula was used to determine a sample size

of UHEPrs to be included from all 10 sub-cities of Addis Ababa, while a census
was employed to collect data from all UHEPrs currently working in the 8
kebeles of Dire Dawa city administration. In addition, a total of 113 UHEPrs
from other towns (whose sample size was calculated in the first phase of the
page- 489
HEP assessment) were included in the study. The following parameters and
assumptions were used to calculate the sample size of UHEPrs from Addis
Ababa to be included:
Pi=0.5 was taken to obtain the maximum sample size;

D=a 5% level of precision was used.
Considering the above parameters and the 5% non-response rate, the final
sample size for UHEPrs to be included from Addis Ababa was determined as
404.
1.6.4 Sample size for HC assessment
The study employed a census to include all 98 and 7 HCs operating under the
Addis Ababa and Dire Dawa city administrations, respectively. Moreover, 34
HCs from other towns that were assessed as part of the rural componenet were
included in the HC assessment.
1.7 Sample size for qualitative data
The sample size for qualitative data collection (focus group discussions and key
informant interviews) was determined by considering the diversity of potential
information sources about the UHEP. Data were collected at all levels of the
health system in each region, from units directly responsible for the planning,
coordination, management and implementation of the UHEP; and those who
are directly or indirectly affected by the program (user and non-users).
page-490
Table 1-2. Sample size for qualitative study
Number of FGD or KII by place

Participant category Addis Dire Other MoH and
Total
Ababa Dawa Urban Partners
FGD with had - 13 - 13
FGD with Community members users 9 4 9 - 22
FGD with Community members non-
6 - - 6
users
FGD with UHEPrs 4 2 10 - 16
KII with HC head 5 2 8 - 15
KII with HC HPDP Process Owner 3 2 8 - 13
KII with Woreda head 2 0 2 - 4
KII with sub-city UHEP team leader 3 0 0 - 3
KII with sub-city HPDP process owner 7 0 1 - 8
KII with City admin HB head/vice 0 1 3 - 4
KII with City admin HB UHEP
1 2 3 - 6
coordinator
KII with City admin HB HPDP
0 1 4 - 5
Director
KII with FHT Coordinator or Leader 2 2 0 - 4
KII with MoH - - - 5 5
KII with Partners - - - 6 6
KII with Policy Advisors - - - 2 2
Total 42 16 61 13 132
Abbreviations: FGD, focus group discussion; HAD, Health Development Army; UHEPrs, Urban Health
Extension Professionals; KII, key informant interview; HC, Health Center; HPDP; HB; HPDP; FHT,
Family Health Team; MoH, Ministry of Health.
1.8 Development of data collection tools
Qualitative data collection guides and quantitative tools were developed for
different categories of data-collection activities. The quantitative data collection
tools include a household questionnaire, UHEPr survey questionnaire, and HC
assessment tool. The key informant interview (KII) and focus group discussion
(FGD) guides were developed to collect qualitative data from all levels of the
health system and at the community level.
page- 491
All the data collection tools were prepared through a process that involved 4
major steps:
1. Sub-constructs related to the overarching research questions/constructs

of the PHCPI framework were identified through a review of literature by
a team of professionals organized for each health system building block
component (governance and leadership, human resources for health,
healthcare financing, facilities and infrastructure, drugs and medical
supplies, service delivery, and health information);
2. For each sub-construct, data need was determined at the federal,

regional, woreda, HC and community levels;
3. For each data need determined in step 2, standard questions were

identified from different sources. For those with no pre-existing source
of standard questions, new questions were formulated by the respective
teams;
4. Questions were then arranged into data-collection tools and guides

based on their respective data sources.
The household survey tool was translated into local languages and translated
back into English to ensure the accurate translation of each question. The tools
were pre-tested in communities outside of the sample woredas prior to data
collection. Observations and inputs from field-level pre-testing were used to
refine the data-collection tools and procedures. All quantitative data-collection
tools were designed in Open Data Kit (ODK), and data were collected
electronically using Android-based tablet computers.
page-492
1.9 Data collection and quality assurance team
A team of enumerators, supervisors, and coordinators was established for data

collection in each of the 9 regions and 2 city administrations. Each team was
composed of:
• A regional coordinator
• Supervisors
• Qualitative research assistants
• Household data collectors
The number of team members varied depending on the sample size in each
region. Data-collection teams identified and engaged local field guides. The
assessment team was responsible for overseeing the overall process of data
collection, starting from the recruitment of data collectors. The team also
conducted a supervisory visit with random data quality checking procedures at
the field level.
1.9.1 Training of data collectors
A comprehensive training was provided to all data-collection and quality-

assurance team members. The training covered general guidelines on data
collection methods, sampling and data collection procedures, and the content
of each data-collection tool. The training was organized for 5 days: 3 days for
discussion, 1 day for practice, including the use of tablet computers for data-
collection activities, and 1 day for fieldwork planning.
page- 493
CHAPTER 2
Findings:
Relevance of
the UHEP
page-494
CHAPTER 2
2 Findings: Relevance of the UHEP
According to Development Assistance Committee of the Organization for

Economic Cooperation and Development (OECD/DAC), relevance is the
extent to which a program is suited to the priorities and policies of the target
group, recipient, and donor1. In this study, we assessed the relevance of the
UHEP to solving the common health problems of the urban community, the
relevance of the service delivery modalities, and the community perceptions
and acceptability of the UHEP. Participants were asked to reflect on the most
common health and health-related problems in cities and whether the UHEP
was relevant and responsive to the existing health needs of the urban community.
The participants were also asked for their suggestion on services that should
be added to the existing packages, as well as services to stop, and suggestions
related to the UHEP’s implementation modality. Therefore, this section of the
report will describe the epidemiological, social, and structural relevance of the
program from community and program staff perspectives.
2.1 Epidemiological Relevance of UHEP Packages
2.1.1 Context of Disease Burden and Health Problems in Urban

Ethiopia
Ethiopia has successfully reduced deaths related to communicable, maternal,
neonatal, and nutritional deficiency diseases (CMNNDs) and injuries in the
past 2 decades. The burden of non-communicable diseases (NCDs) such as
cardiovascular disease, diabetes, cancer, and chronic respiratory disease has
increased in recent years, however, making these diseases among the leading
causes of death. NCDs account for 42% of deaths and contribute to 69% of the
Disability-Adjusted Life Years (DALYs) in 2015.2
According to the 2015 Global Burden of Disease Study (GBD), lower respiratory
infections, tuberculosis (TB), diarrheal disease, ischemic heart disease, HIV/
page- 495
Findings: Relevance of the UHEP
AIDS, hemorrhagic stroke, neonatal encephalopathy, ischemic stroke, neonatal

preterm birth, and neonatal sepsis were the top 10 leading causes of mortality
in Ethiopia in 2015. The attention given to the prevention and control of NCDs,
such as cardiovascular disease, diabetes, cancer, and chronic respiratory disease,
was very low, however, enabling these diseases to become leading causes of
premature mortality and death in 20152.
Previous studies have also indicated the major epidemiological transition

from infectious diseases to NCDs in Ethiopia. Nevertheless, NCDs were not
priorities, and the national response to NCDs remains fragmented. While the
country is progressing toward universal health coverage, disease prevention and
control strategies in Ethiopia should consider the double burden of infectious
communicable diseases (CDs) and NCDs.3,4
Various studies have suggested heterogeneity in the burden of NCDs, due to

diverse sociodemographic, lifestyle, and health risk factors between the rural
and urban populations2,3 Risk factors of the major NCDs, such as tobacco use,
alcohol abuse, hypertension, being overweight or obese, high glucose levels,
and khat use, were highly prevalent among the urban population. The existence
of those risk factors depends on the economic transition, rapid urbanization,
and changes in lifestyles of the 21st century.5
Despite the importance of NCDs, like hypertension, to public health in urban

areas, a majority of individuals with these diseases are not aware of their
status. This results in the burden of the hidden morbidity and subsequent
complications due to NCDs.3,4 Without action, Ethiopia will be the first among
the most populous nations in Africa to experience the dramatic burden of
premature deaths and disability from NCDs by 2040.6
Access to and use of WaSH amenities, such as toilets and handwashing facilities
with soap, significantly contribute to the reduction of diarrheal diseases.6 A
cross-sectional study conducted in the town of Sebeta showed that children from
households that had handwashing facilities were less likely to have diarrhea
compared to those that did not have handwashing facilities.6
The importance of sanitation to safeguarding human health is well known

and undisputed. Access to improved sanitation is a human right. On the road
page-496
to universal access to improved sanitation, more than half of the Ethiopian

population has no access to improved sanitation. In both urban and rural
Ethiopia, access to improved sanitation coverage falls far short of the Millennium
Development Goals (MDG) target, and the majority of residents are living
with high health and environmental risks. The high proportion of those using
unimproved sanitation facilities (88.6% in Addis Ababa urban slum-dwellers
and 82.5% of urban residents nationwide) indicates that the urban poor have
rates of sanitation service coverage as low as the rural population.7 Even this
may underestimate the actual coverage, which might be better gauged if the
method of estimating improved sanitation coverage considered the functioning
and use of sanitation systems and fecal sludge management (FSM) rather
than simply identifying and counting the available sanitation technologies.8
Ethiopia has achieved the 50% reduction of most of the MDG targets related
to TB. The decline of TB incidence and prevalence rates, however, has been
comparatively slow. TB remains a major public health problem; Ethiopia is
among the 22 countries with the highest TB burden, with a high number of
missed and infectious TB cases in the community.9 The country should strengthen
its TB case detection and treatment programs at the community level.9,10
Studies have also indicated that Ethiopia’s healthcare delivery system is being
challenged by the double burden of diseases owing to communicable and
NCDs. There is also a major epidemiological transition to NCDs, particularly
in the urban population, and we expect to see dramatic increases in the burden
of premature death and disability from NCDs.3
In this study, we explored whether the UHEP packages are perceived as

relevant to solving the common health and health-related problems of the
urban community. Participants in the qualitative study reported both CDs
and NCDs to be the most common health problems of the urban community.
Hypertension, mental health problems, diabetic mellitus, cancer, and heart
diseases were among the NCDs frequently reported by the participants. A
participant from Dire Dawa described the situation as follows:
page- 497
“
In the past, diabetic mellitus and hypertension were considered the
diseases of the wealthy. Nowadays, diabetic mellitus is common
among the poor community.…Hypertension, diabetes, and mental
health problem are increasing over time.…Cancer is increasing
because it is not easily diagnosable.
(HC head, Dire Dawa)
Moreover, the participants frequently mentioned HIV, diarrheal diseases,

malaria, typhoid fever, and typhus to be among the main health problems
in urban settings. In addition, overcrowding, limited access to safe and clean
water, addiction, substance abuse, and traffic accidents were reported to be
the important health problems of the urban community. This result is basically
in line with the above evidence that the urban population is facing challenges
due to the double burden of CDs and NCDs.
Other frequently mentioned problems include the shortage or/and absence of

adequate facilities and services for proper solid and liquid waste management,
the lack of a safe and clean water supply, and the challenge of proper excreta
disposal, which are reported to have a significant effect on health and well-
being of the community, particularly the urban poor. Participants in this study
also identified problems related to water, hygiene and sanitation (WaSH)
as factors predisposing them to various health problems, including diarrheal
diseases, typhoid fever, and typhus.
The findings of this study are in line with the previous studies conducted to
assess the major causes of morbidity and mortality in the urban settings of
Ethiopia. A study conducted in Addis Ababa to determine the patterns of
mortality in public and private hospitals reported that a majority of hospital
deaths are secondary to CDs, maternal conditions, and nutritional deficiencies,
followed by non-communicable causes.10
Overall, participants in the qualitative study reported that the UHEP was
an appropriate vehicle for addressing the common health problems in urban
settings and was aligned with the country’s national health policy. Participants
stated that the UHEP is relevant to promoting basic health services in urban
settings and solving the health problems of the urban community. Participants
page-498
identified various forms of evidence supporting the relevance of the program.

Similarly, the quantitative findings from Addis Ababa and Dire Dawa show
that 67% of households perceived UHEP as relevant in addressing the health
needs of the urban community. The perceived relevance of the program was
found to be higher among respondents in Addis Ababa than in Dire Dawa
(Figure 2-1).
Figure 2-1: Community Perception about Relevance of UHEP

Family planning (FP), immunization, and TB prevention and control activities
were reported to be relevant in urban settings. The UHEP was underscored as
particularly relevant in solving the health problems related to overcrowding
and migration to the cities.
On the other hand, some packages have limited relevance in solving the health
problems of the urban community. The participants described some packages,
including the WaSH package, as being inadequate to address the multifaceted
problems of the urban community. The relevance of each of the package is
described in the following sections.
According to the responses from program staff and the community, the
inadequate implementation of the program packages and the multifaceted
nature of the health problems in urban settings limited the relevance of UHEP
in addressing the health needs of the urban community. They believe that
page- 499
the UHEP implementation lacks the adequate collaboration of important

stakeholders. More specifically, participants believe that the program lacks
integration, harmonization, and alignment among various sectors working
on urban health-related activities. Moreover, stakeholder involvement during
guideline development, planning, implementation, and monitoring was reported
as inadequate, limiting the relevance of the UHEP in addressing the health
problems of the community related to hygiene and sanitation.
In general, the hygiene and sanitation and maternal health packages were
found to be more relevant in reducing the burden of CDs, maternal morbidity,
and mortality. Moreover, the health education provided at the household level
was found to be relevant in improving health literacy, which in turn affected
health-seeking behavior and health service use among the urban poor.
Several challenges were reported that limited the implementation and relevance
of the UHEP. Participants frequently reported that resources needed for the
implementation of the UHEP were usually missing. Moreover, the number of
UHEPrs is believed to be inadequate to improve community awareness on both
NCDs and CDs. In addition, rapid and unplanned urban population growth
was reported to be a burden to the health system and a challenge to the
successful implementation of the UHEP.
2.1.2 Relevance of Water, Hygiene and Sanitation Packages
The environmental and personal hygiene and sanitation package is one of

the major UHEP packages. It is intended to improve and promote proper
behavior toward personal hygiene, solid and liquid waste management, excreta
disposal (as well as latrine construction and use), food and water hygiene,
and the availability and use of handwashing facilities. This package is being
implemented mainly through health education, model family training, and home-
to-home visits with the aim of ultimately changing the knowledge, attitude, and
behavior of the community toward appropriate WaSH practices.
Ethiopia has one of the fastest growing urban populations in the world, and the
number of urban residents has more than doubled in the past 2 decades. This
rapid and unplanned urbanization has exacerbated challenges to providing
safe and adequate water, infrastructure for waste management, housing, and
page-500
health services. Studies also show that more than 70% of the urban population
in Ethiopia resides in slums, which are characterized by overcrowding, poor
housing conditions, and the absence of safe water and sanitation facilities7.
The results of the current study clearly indicate that all components of the
WaSH package are important in addressing health problems, mainly diarrheal
diseases, typhus, and other CDs. Due, however to the complexity of WaSH
interventions, which require the collaboration of different actors, the problems
remain a major challenge to the urban community. Inadequate or an absence
of space was the most frequently mentioned challenge to constructing latrine
and waste-disposal facilities. In addition, the participants reported that some
segments of the urban community lacked the space and financial capacity to
construct a latrine.
“
When you teach how to use a latrine, there has to be minimal material
support too, and then their needs can be satisfied. When you give
education, they know it, but how can they implement it? If you tell
them to construct a latrine, even if they have the money, since the
management has not prepared the site, they cannot construct it. So,
even if we educate, how can you make a community with no latrine
a model?
Sub-City Health Promotion and Disease Prevention Process Owner, Addis

Ababa
page- 501
The lack of latrines, shortage of water-treatment supplies, shortage of a clean

water supply, and poor waste disposal systems were frequently reported to be
among the problems of the urban community.
As mentioned by some participants, the service fee to empty a filled latrine is

so high that the urban poor cannot afford to pay. As a result, latrines are not
emptied regularly, and this causes environmental pollution due to the overflow
of excreta. Participants stated that a large number of communal latrines
became nonfunctional because of the inability to pay for a pit-emptying service.
Some participants reported the difficulty of emptying filled latrines due to the
absence of a road facility to access slum areas. One FGD participant from
Addis Ababa said:
“
...Sometimes latrines remain filled and flow onto the road...
Female FGD participant, Addis Ababa
The participants strongly argued that urban sanitation and waste disposal
issues are complex and need collaboration with other sectors. The program staff
argued that hygiene and sanitation ought to be handled at a multi-sectoral
level. The UHEP alone does not bring changes in urban sanitation, as the
role of the UHEPr is limited to creating awareness. The participants reported
that the water and sewerage authority and waste management offices have
major roles in enhancing sanitation and hygiene in urban settings. Thus, the
participants argued that the proper implementation of hygiene and sanitation
packages require an integrated approach in which all relevant stakeholders
should be actively involved in the planning, implementation, and monitoring of
interventions and activities.
page-502
Some participants also noted that, despite the efforts made by UHEPrs and
the community to improve hygiene and sanitation conditions, the problem of
latrine and liquid waste had not been solved so far, which is partially attributed
to lack of sectoral collaboration. One female participant in the community
“
member FGD in Addis Ababa said:
There were public toilets and full and overflowing…the UHEPrs reported
it to the responsible body (the waste, water, and sewage authorities)
and even brought the responsible person to see the situation, but
the problem has not been solved yet. We [the community] know the
UHEPrs are trying to solve our problems, but they lack support from
other government bodies like the water and sewerage authorities.…
Think what will happen if a diarrheal disease outbreak occurs in our
area.
The community and program staff most frequently mentioned the relevance of
health education and the creation of awareness about hygiene and sanitation
to reduce the incidence of diarrheal diseases. The health education provided to
the community was also reported to have contributed to improved awareness
of water treatment and handwashing. One participant in the Addis Ababa
community members FGD, explained that:
“
Following the rainy season, there was diarrheal disease, but now that
has decreased since the Health Extension professionals go house to
house to provide brochures and teach the community on how to treat
water before use. So, it [diarrheal disease] is decreasing..
A few participants explained that implementation of the UHEP has contributed

to a reduction in indoor air pollution and its health effects. An HEPr from
Adwa, Tigray described the situation as follows:
page- 503
“
Previously, as their kitchen did not have an outlet for smoke, mothers
were exposed to blindness, aging, and premature death. That is why
the house gets suffocation and the mothers are exposed to indoor
smoke day and night. Let alone this, you know what sandalwood
smoke feels like. So we got satisfaction the time we showed them
practically how to make the outlets for their kitchen smoke when they
do so. We were also happy to introduce the wood-saving stove [MIRT
ETON] by jointly working with the water resources and other sectors.
UHEPrs, Adwa, Tigray
In general, the findings of this study show that the interventions of the WaSH
package are important to addressing health problems related to poor personal
and environmental hygiene and sanitation, mainly diarrheal diseases. This
package is challenged, however, by the multifaceted problems of urban sanitation
and hygiene, whose interventions require strong collaboration among different
government sectors. A considerable number of urban community members live
in rental houses, which have no access to a latrine. Moreover, some segments
of the urban community have no space to construct a latrine, and those who
have space have no economic capacity to construct one. The findings of this
study indicate that the WaSH package, as it is now, has a limited relevance to
enhancing urban sanitation and hygiene. The absence of innovative technology
options to ensure the availability of a latrine facility that fits the existing living
conditions and economy of urban poor households is another challenge.
2.1.3 Relevance of the Family Health Package

The family health packages, including maternal and child health, FP,
immunization, nutrition, and adolescent and reproductive health, were among
the UHEP packages frequently reported to be relevant to the urban community.
In our qualitative findings, both program staff and FGD participants expressed
their belief that the UHEP played a significant role in the observed improvements
in maternal and child health service use, including FP, ANC, PNC, facility
delivery, and immunization. Overall, the participants argued that the number
of women giving birth at home has decreased and that their health-seeking
behavior has improved. Previous studies have documented that women who are
aware of the UHEP and those from households visited by UHEPrs are more
likely to visit health facilities11.
page-504
“
The maternal and child mortality shows a huge decrease, especially
when it comes to maternal mortality. It is one of the packages, and
the other is the services that will be given for pregnant women.
Previously there was home delivery, but now they are going to health
facilities for delivery.
WorHO Head, Benishangul Gumuz
Despite the reported successes of the implementation of the UHEP in the

improvement of health-seeking behavior and service use, participants believed
that there were segments of the population who have still limited access to
basic health services. For instance, a participant from Mekelle explained that
there were still women who give birth at home. This finding indicates that the
UHEP has not yet fully addressed the health needs of the segments most in
“
need in the urban community.
There was one death due to home delivery. The woman was new to
our residence. There was also another home delivery. Home delivery
is common among the newcomers to our community. We are trying
to solve their problem along with the Health Extension Workers. But
we still have the problem.
Female community member, Mekelle, Tigray
Overall, the family health package was reported to be relevant to addressing

the health needs of women and children. Poor implementation and the absence
of clinical services, however, limited the relevance of the package in addressing
the urban communities’ health needs.
As suggested by participants, to meet the changing needs of the urban

population, the UHEP should go beyond health education and referral; in addition
to community sensitization and health education, participants suggested the
provision of services like FP, home-based HIV testing, treatment of childhood
illness, and treating other common illnesses. Moreover, study participants
strongly believed that the inclusion of clinical services would increase demand
for the UHEP and the acceptance of UHEPrs by the urban community. The
proposed changes in service packages would also help UHEPrs by reducing
page- 505
the community fatigue resulting from the prolonged implementation of similar
“
services using similar methodologies.
They only distribute condoms here in Addis Ababa. It was possible

to make available services beyond contraceptives, such as short-term
family planning in the form of tablets. These kinds of things were not
done or executed. If these kinds of service were included. It would
also increase their acceptance. People would easily get the services
there instead of coming to facilities or Health Centers. So, I think this
remains from the program or service.
HC Head, Addis Ababa
2.1.4 Relevance of Disease prevention and control Packages

HIV/AIDS and TB control are among the UHEP packages implemented by
UHEPrs. Participants in the qualitative study reported that awareness creation
and the screening of cases have resulted in a reduction in new cases of TB. The
participants reported the community preference for the UHEPrs for HIV testing
due to their fear of stigma. Moreover, some participants in the qualitative study
suggested that UHEPrs provide treatment for people infected with HIV/AIDS.
Moreover, the participants believed that the program relatively neglects HIV
prevention among commercial sex workers and youths, who are considered to
“
be more vulnerable segments of the urban population
The reduction of new TB infections is made by home-to-home

education on the major symptoms of TB and early case detection.
Community awareness has increased over time.…There were more
people getting [HIV] tested in the kebele with them [UHEPrs] than
the ones at the Health Centers. Because the society does not openly
come forward.
Program officer, Tigray
page-506
The available literature indicates a high burden of HIV/AIDS in urban settings

compared to rural settings in Ethiopia. The urban prevalence rate of HIV/AIDS
is estimated at 7.7%, while the rural prevalence rate is 0.9%.13 Similar to the
findings of this study, a previous study conducted in urban settings in Ethiopia
reported that there were population groups that were particularly vulnerable
to HIV infection.14 This finding indicates that the UHEP’s HIV/AIDS package
should be designed in such a way that it addresses the segments of the urban
community most vulnerable to the disease.
One participant from Dire Dawa said that HIV is still a public health problem
“
in the city:
Currently, the diseases which are increasing are cancer and HIV. It has
been said that HIV has decreased, but the prevalence is increasing.
This is because the awareness creation education has decreased for
a long time. We can hear the prevention method of HIV on TV
sometimes: the number shows that the society is not preventing HIV.
HC Head, Dire Dawa
Malaria prevention and control is one of the UHEP’s disease prevention and
control package. In the qualitative study, participants explained how malaria
prevention measures, including awareness creation and the distribution of
insecticide-treated bed nets (ITNs) have contributed to a reduction in the
“
malaria epidemic in Dire Dawa. One participant from the city said:
Malaria has decreased. They [UHEPrs] give us bed nets and teach
us to remove stagnant water around the home. When we compare it
with the number 5 years ago, there is a difference in the prevalence
of malaria. It is decreasing.
HC Head, Dire Dawa
The prevention and control of NCDs are among the interventions of the
UHEP. In the qualitative study, program staff and the community reported
the awareness creation and screening for hypertension and breast cancer as
among the successes of UHEP. One participant from Bahir Dar describes this
condition:
page- 507
“
The health care professionals (UHEPrs) advised us to go to Felegetsion
Hospital to have a pre-cancer screening. The professionals also come
to our homes and check our blood pressure, and, if hypertension is
found, she advises us to minimize drinking coffee, eating salt, etc.
WDA member, Bahir Dar, Amhara
Overall, study participants broadly questioned the relevance of the program

in addressing NCDs. They voiced their concerns that the UHEP has paid little
attention to emerging chronic disease burdens, including cervical and breast
cancer. They also described their belief that the program, relatively speaking,
neglects HIV prevention among commercial sex workers. They believed that
HIV prevention among commercial sex workers is less emphasized despite the
increasing number of commercial sex workers in the cities. They also said that
diabetic mellitus and mental health problems were not well addressed by the
UHEP even though these conditions have become challenges in recent years.
2.2UHEP Implementation Strategies and Modalities
The UHEP uses a combination of implementation modalities and approaches

to deliver the packages to their intended target beneficiaries. This study
assessed the strategies and approaches being used by UHEPrs to implement
the program at the household level. According to our findings, home visits,
model families, the Women’s Development Army (WDA), and the Family
Health Team approaches are the major UHEP implementation modalities for
household-level implementation.
Model families and WDAs are the major support networks, who play catalytic
roles in the implementation of HEP packages at the community level. As
envisioned in the consecutive health sector development plans, the model family
and the WDA are considered extensions of the UHEPrs. To this end, the model
family and WDA are considered major UHEP implementation strategies. The
relevance of the UHEP implementation approaches and their functionality will
be described in this section.
page-508
2.2.1 Relevance of Model Family
The UHEP implementation manual notes several times that the ultimate aim
of the program is to produce model households with a higher level of health
literacy, as demonstrated by better health-seeking behavior and healthy living
conditions. To be acknowledged as a model family, a household should do 60
hours of training on the UHEP packages and implement practically at least
75% of the packages for which the household is eligible.
Participants described that the implementation of the UHEP through the

model family approach as following the diffusion strategy, in which model
households are expected to share their knowledge and experience regarding
the UHEP packages with neighboring non-model households. As reported by
some participants, the UHEPrs are working with model households to educate
“
the community and share practical experiences.
In sharing the best practices, the selected model person or group will
move from house to house and do a practical demonstration of how
to keep cleaning, preparing handwashing materials, and cleaning the
surroundings.
WDA member, SHEPrarobit, Amhara)
One WDA member from Assosa reported that model families were involved in
“
awareness-creation activities and teaching other members of the community.
There are other groups under us in the 1-to-5 networks. We are also
making them models like us. For example, discussing SRH issues,
communicable and non-communicable diseases…We teach them
about the signs and symptoms and ways of detecting diseases
like cervical and breast cancer. We also tell them to seek medical
treatment at the Health Centers, and we also follow that. We also
trained and showed how to do waste disposal, toilet use, children’s
nutrition, and a balanced diet.
WDA member, Assosa, Benishangul Gumuz
page- 509
UHEPrs reported that they had used model families to reach as many people
as possible. The use of volunteer women was one of the strategies used in
the creation of model families. These volunteers are trained to become model
women in the community. These model women, in turn, train other women in
“
their community, as explained by a UHEPr from Logia:
We recruited willing women who have a say in the community, give

them a 3-month-long training, and when they graduate, they get
the title “model women.” In turn, they have to go and teach their
community and be their role models.
UHEPr, Semera Logia, Afar
Similarly, program staff also reported that model families not only served to
create other model households but also were used as a basis for the creation of
model development groups, model kebeles or villages, model youth centers, and
model schools. Model households enhance the implementation of the UHEP by
sharing their knowledge, skills, and experiences with their communities through
home-to-home visits. They are involved in awareness-creation activities and
teach other members of the community.
According to participants, however, there remain gaps in the implementation

and use of model households. A lack of inter-sectoral collaboration, low
community capacity for some packages (e.g., to construct latrines and perform
waste disposal), and problems with integration among the Family Health
Team (FHT), WorHOs, HCs, UHEPrs, and WDA members were reported
as the limitations in the creation of model households. Moreover, decreasing
the training given and other motivations for model households was another
reported factor.
page-510
It was also reported that some model families do not practice healthy behaviors
sustainably. Once they graduated as models, they began to dispose of their
wastes in an open field. As a result, as one participant reported, there is
punishment when model families fail to do what is expected of them. A woman
“
from Bahir Dar confirmed this:
We established a group. First, we leaders learn about the packages.

So, as a leader, if we go against what we learned and throw or dump
garbage carelessly, we pay 50 birrs as a punishment because we all
are models and we took all 16 packages.
WDA member, Bahir Dar, Amhara
“
An informant from Tigray mentioned:
The performance of the UHEPrs is different from what it was before.…

Previously, the community easily accepted the packages since it was
a new initiative, but now there is resistance. Even model households
created previously are going back now.
Program officer, Tigray
Another gap reported in the creation of model household was that UHEPrs
might select and graduate model households without adequate follow-up
and support. The high expectations of the city administrations, RHBs, and the
MoH regarding model family training coverage was stated as a major reason
for compromised model family training and graduation. Program staff and
the community reported that households who did not take the training and
implement the packages were being accredited and reported as model families
for the purpose of inflating the performance of the UHEPrs and the catchment
“
HC.
As the leadership takes account of the graduating households, it

might have a push factor on encouraging UHEPrs to graduate
households without following up on the process.
Sub-City HPDP process owner, Addis Ababa
page- 511
Finally, some participants noted that the experience-sharing from model families
is not enough to scale up the implementation of UHEP packages because
experience-sharing is not done day to day but rather only when a field visit is
scheduled by supervisors or other guests.
In line with the qualitative results, the findings of a household survey conducted
in Addis Ababa and Dire Dawa showed that only 18% of households were
certified as model households. In this study, receiving a model family training
was found to have a significant association with households` participation in
sanitation campaigns, the availability of a separate kitchen, the availability of
a handwashing facility with soap or a substitute, liquid waste disposal practices,
enrollment in community-based health insurance (CBHI), ANC follow-up,
and exclusive breast-feeding practices. Model households were more likely to
participate in sanitation campaigns, have a separate kitchen, and be enrolled
in CBHI. Moreover, model family households were likely to report that they
were aware of the danger signs during pregnancy and the benefits of exclusive
breastfeeding (Table 2-1).
Table 2-1. Comparison of model and non-model households by awareness and

practice of urban residents
Model Non-model Total

P-value
N % N % N %
Participation in sanitation No 13 4.0 304 21.5 317 18.3 <.001
campaign Yes 308 96.0 1109 78.5 1417 81.7
Availability of separate Yes 174 54.2 631 44.7 805 46.4 .002*
Kitchen No 147 45.8 782 55.3 929 53.6
Handwashing facility Yes 63 20.4 186 13.4 249 14.7 .002*
with soap or substitute No 246 79.6 1198 86.6 1444 85.3
Disposal of solid waste in Yes 0 0.0 74 5.2 74 4.3 -
an open field No 321 100.0 1339 94.8 1660 95.7
Sewer line 192 59.8 565 40.0 757 43.7 <.001
Liquid disposal
Liquid waste disposal 83 25.9 312 22.1 395 22.8
pit
practice
Open field 41 12.8 489 34.6 530 30.6
Other 5 1.6 47 3.3 52 3.0
Yes 117 84.8 235 64.7 352 70.3 <.001
Enrolment in CBHI
No 21 15.2 128 35.3 149 29.7
page-512
Health Center 53 60.9 296 62.8 349 62.5 .526

Gov. hospital 31 35.6 150 31.8 181 32.4
Place of delivery for most
recent child Private clinic/
0 0.0 10 2.1 10 1.8
hospital
Home 3 3.4 15 3.2 18 3.2
Yes 257 85.4 965 72.4 1222 74.8 <.001
ANC visit for most recent
No 29 9.6 257 19.3 286 17.5
pregnancy
Not applicable 15 5.0 111 8.3 126 7.7
TT vaccination during Yes 233 77.4 887 66.5 1120 68.5 <.001
most recent pregnancy No 68 22.6 446 33.5 514 31.5
Awareness on danger Yes 196 65.1 641 48.1 837 51.2 <.001
signs of pregnancy No 105 34.9 692 51.9 797 48.8
Awareness about Yes 245 81.4 968 72.6 1213 74.2 .002*
exclusive breastfeeding
practice No 56 18.6 365 27.4 421 25.8
Didn`t
6 2.0 30 2.3 36 2.2 .013*
breastfeed
Did not
exclusively 75 24.9 400 30.0 475 29.1
EBF practice breastfeed
Exclusively
208 69.1 798 59.9 1006 61.6
breastfed
Not applicable 12 4.0 105 7.9 117 7.2
Knowledge of critical Yes 165 51.4 672 47.6 837 48.3 0.21
times for handwashing No 156 48.6 741 52.4 897 51.7
Wanted status of the Yes, wanted 268 89.0 1204 90.3 1472 90.1 0.50
recent pregnancy Not wanted 33 11.0 129 9.7 162 9.9
Abbreviations: CBHI, community-based health insurance; ANC, antenatal care; TT, tetanus-toxoid;
EBF, exclusive breastfeeding.
In general, training and implementation of the model household were found

to be relevant to and effective at implementing some of the UHEP packages
and components. There is a need, however, to scale up the number of model
households, and follow their performance over time. Inter-sectoral collaboration,
integration, the provision of continuous training, follow-up from UHEPrs with
the model households, and optimizing the overall performance of the HEP are
suggested to improve the implementation of the program at the household
level.
page- 513
2.2.2 Relevance of Women’s Development Army (WDA)
The UHEP implementation guidelines also indicate that the WDAs are the
primary collaborators of UHEPrs that can serve as anchors and facilitate the
implementation of the program at the household level. The UHEP mainly
focuses on educating the community on health and health-related issues and
creates demand for health service use through referral and linkage with the
catchment’s primary health care unit (PHCU). The UHEPrs collaborate with
WDAs in identifying the health needs or problems of households and work
together as a team.
The findings of this study indicated that WDAs, as the UHEP’s implementation
collaborator, were relevant to awareness creation and communication and
community mobilization activities at the community level. Participants had
reservations, however, about the relevance of WDAs due to their limited
technical capacity to educate the community about the packages. In addition,
some participants were concerned about the negative consequence of WDAs’
involvement in politics and administrative activities.
Participants in the current study reported that the program emphasized

community sensitization, which resulted in improved health-seeking behavior
and personal and environmental hygiene. Some participants also believed that
community structures like the WDA have contributed to the improved uptake
of ANC, facility delivery, and PNC. Use of the WDA has also strengthened
hygiene and sanitation practices at the household and community level.
Participants, mainly UHEPrs, mentioned that working with WDAs and the
1-to-5 groups helped them disseminate health information and facilitated the
implementation of the UHEP packages at the household level. Some participants
also mentioned that these groups helped them reach the target group easily
and provide the service. They also reported that the WDAs were useful and
supportive in areas where there is an inadequate number of UHEPrs.
page-514
“
It [HEP] is implemented at individual, family, and household levels…
The first principal people we use to provide the package are the
Women’s Developmental Army. Within this WDA, at least 30 women
are grouped together. Under this group, there are 1-to-5 groups.…This
wing enables us to meet people easily and deliver the HEP through
this method.
RHB Head
Some participants agreed with the relevance of the WDA in implementing the
program but suggested the provision of training on health-related topics to
enhance WDAs’ knowledge. On the other hand, some program staff believed
that the WDA was not functional and that it would be better to stop this
“
approach.
I wish the WDA was dismissed, I am sorry to say this. It has led
them [UHEPrs] to dependency: the structure is there, but it is not
functional. Sorry, I am not against the politics.…I would love it if they
functioned properly ... The truth is that it is a lie, and nobody is
leading or following the WDAs. The second factor is due to their
social life, WDAs don’t discuss weekly; it is a lie. Is it clear? They have
many personal issues to take care of, and they are not functioning.
It would be great if UHEPrs went home to home by identifying gaps
so that behavioral change would come. That was how the initial tone
was.
RHB HEP coordinator
Moreover, the participants also suggested better integration between the

UHEPrs and WDAs. The participants also mentioned that sharing experiences
and lessons among WDAs would improve the implementation of the UHEP.
Setting clear criteria for the selection of the WDA members and volunteers
would contribute to better HEP implementation.
page- 515
“
It is important that WDAs get the chance to review each other’s

performance and share lessons. Experience-sharing and performance
review among WDA leaders would improve our work. There is also
a need for closer monitoring of what we do at the community level.
So far, we have never been supervised at the community level except
by UHEPrs.
WDA member, Assosa, Benishangul Gumuz
Most participants perceived the WDAs as a critical part of the program, yet
there was concern that the existing WDAs were inactive due to their poor
support and recognition by the government. Other participants reflected their
idea that the WDA was the backbone of activities implemented by the woredas
“
and kebeles.
The Women’s Development Army is the backbone of the kebele and

woreda. Most of the work to be done in the community is by the
Women’s Development Army.
WDA member, Adwa town, Tigray
2.2.3 Relevance of Home Visits
In the case of the UHEP, most of its packages and services are designed to be
rendered by providing health education and information, counseling, screening
for CDs and NCDs, and providing follow-up (with pregnant mothers, under-5
children, chronic patients, and model households) through home-to-home visits.
As per the UHEP implementation guidelines, UHEPrs are expected to spend
significant working time (3 days per week) implementing the UHEP through
home-to-home visits to target households (economically poor households and
those with under-5 children, pregnant and lactating mothers, and individuals
with NCDs and CDs). The UHEPrs are also expected to reach every household
assigned under their catchment at least once a year, regardless of their health
needs or economic status.
As one of the means for family health promotion approaches, a home visit was
found to be relevant to providing health education and visiting pregnant and
postpartum women. In addition, the tracing of unvaccinated children, screening
for hypertension, and provision of other clinical services were relevant, as were
follow-up with patients at home, the provision of TB directly observed therapy
page-516
(DOT) at home for bedridden patients, the identification and referral of

people with disabilities and mental illness, ITN distribution, breastfeeding, and
nutrition. A participant from Benishangul-Gumuz said:
“
They [UHEPrs] go home to home and visit sick people. They also
support economically deprived people in getting medical treatment.…
Regarding health, let me tell you my experience. Once the UHEPrs
came to my house and I explained to them the health problems I
had. They told me how to handle my family, how to interact with my
neighbors, how to take care of my child, and about the use of bed
nets, nutrition, and other things.
Female community member, Assosa, Beinshangul Gumuz
“
One UHEPr from the SNNPR identified the activities related to home visits:
On our home-to-home visits, we provide health education, behavioral

change communication, and teach them how to keep their areas
clean and lead a healthy lifestyle. We also do supervisory and
support visits. For example, the Health Extension Worker explains all
the details needed for a household to make a drainage path…like its
width, depth, and material needs, or like a handwashing setup after
using the toilet. Later on, she will go for a follow-up visit to check
whether the household implemented what was said or not.
UHEPr, Hawassa, SNNPR
While home visits were reported to be relevant to the implementation of the

UHEP, it was widely reported that there were several challenges with them.
The first and widely shared challenge with a home visit is that urban residents
are not available during work days. Moreover, UHEPrs reported challenges
finding households who are employed, due to the overlap in the working hours
of UHEPrs and those of the employed households. This issue remains one of the
barriers that pose a question about the relevance of the home-to-home service
delivery modality for households with an employed spouse.
page- 517
“
It is reported that the houses of most urban people are closed during
the day, and they are not available whenever UHEPrs go to their
houses to provide services. Some households tell their maids not to
open the door whenever UHEPrs knock at their door. Because of this
reason, the UHEPrs couldn’t accomplish the planned house-to-house
service provision.
“
HC Head, Bahir Dar, Amhara
Government workers are spending all their time in their offices, so this
will make it difficult to meet them.
UHEPr, Assosa, Benishangul Gumuz
Another challenge related to home visits is the inadequacy of the number of

UHEPrs. The participants believe that the number of UHEPrs is not proportional
“
to the number of households they are supposed to visit.
Due to large number of households in the kebele, it is difficult to reach

all of them because, if you want to support households adequately, it
may take to you at least 2 hours at each household.
UHEPr, Sheraro, Tigray
Another FGD participant from Amhara similarly explained that it was difficult
“
for the UHEPrs to visit all the households:
I think there are even times they will not make [a home visit] once a
month unless they work Saturday and Sunday.
Female community member, Ebinat, Amhara
It was also reported that performance of home visits was declining over time.
“
An FGD participant from Addis Ababa complained about this:
They were teaching about health and the ways of protecting against
pregnancy, 2 years ago. But I have not seen them for the last 2 years.
Female community members, Addis Ababa
page-518
The household survey conducted in Addis Ababa and Dire Dawa showed that
only 45.7% of households were visited by UHEPrs within the past one year prior
to the survey. Whereas, 36% of households were never visited. The quantitative
analysis showed that home visits have a significant association with household
participation in the sanitation campaign, the availability of a handwashing
facility with soap or a substitute, liquid waste disposal practices, whether the
most recent pregnancy is wanted, ANC follow-up, and awareness of exclusive
breastfeeding. Households visited within 1 year of the survey were more likely to
report that they had participated in the sanitation campaign, had a separate
kitchen, and had a hand-washing facility with soap or a substitute. Moreover,
households visited by UHEPrs within 1 year of the survey were more likely to
report that they had received ANC for the most recent pregnancy, got TT
vaccination, and were aware of danger signs during pregnancy (Table 2-2).
Table 2-2. Effect of home visits on awareness and practices of urban residents
Household visit status
Visited Not visited P-value
Total
within a year within a year
N % N % N
Participation in Sanitation No 97 12.2 220 23.4 317 <.001
campaign Yes 697 87.8 720 76.6 1417
Yes 371 46.7 434 46.2 805 0.817
Availability of separate kitchen
No 423 53.3 506 53.8 929
Availability of handwashing Yes 132 17.2 117 12.6 249 .008
facility with soap or substitute No 635 82.8 809 87.4 1444
Disposal of solid waste in an Yes 30 3.8 44 4.7 74 0.354
open field No 764 96.2 896 95.3 1660
Sewer line 359 45.2 398 42.3 757 <.001
Liquid
181 22.8 214 22.8 395
Liquid waste disposal practice disposal pit
Open field 251 31.6 279 29.7 530
Others 3 0.4 49 5.2 52
Yes 188 72.9 164 67.5 352 0.188
Enrollment in CBHI
No 70 27.1 79 32.5 149
Health
177 64.1 172 61.0 349 -
Center
Place of delivery for most Gov`t hospital 87 31.5 94 33.3 181
recent child Private clinic/
2 0.7 8 2.8 10
hospital
Home 10 3.6 8 2.8 18
page- 519
Yes 602 78.7 620 71.3 1222 .002

ANC visit for the most recent No 117 15.3 169 19.4 286
pregnancy Not
46 6.0 80 9.2 126
applicable
TT vaccination during most Yes 544 71.1 576 66.3 1120 .036
recent visit No 221 28.9 293 33.7 514
Awareness of danger signs Yes 448 58.6 389 44.8 837 <.001
during pregnancy No 317 41.4 480 55.2 797
Awareness of exclusive Yes 597 78.0 616 70.9 1213 .001
breastfeeding No 168 22.0 253 29.1 421
Didn`t breast
16 2.1 20 2.3 36 0.123
feed
Not
207 27.1 268 30.8 475
exclusively
Breastfeeding practice
Exclusively
494 64.6 512 58.9 1006
breastfed
Not
48 6.3 69 7.9 117
applicable
Knowledge about critical times Yes 391 49.2 446 47.4 837 0.455
for handwashing No 403 50.8 494 52.6 897
Wanted status of the recent Wanted 702 91.8 770 88.6 1472 .033
pregnancy Not wanted 63 8.2 99 11.4 162
Abbreviations: CBHI, community-based health insurance; ANC, antenatal care; TT, tetanus-toxoid;
EBF, exclusive breastfeeding.
2.2.4 Relevance of Family Health Team (FHT) Approach
The Family Health Team (FHT) was reported to offer many advantages to
the community, to service provider teams, and to the success of HEP. First,
it benefited poor people, as they could now receive services in their home. It
was reported that the service given by the FHT is of better quality since it is
delivered by a team of experts. As part of the FHT, UHEPrs identify families in
need of home-based treatment and participate in the provision of treatment
along with other professionals. As a result, community acceptance for UHEPrs
increased. Furthermore, as part of the FHT, UHEPrs have started to work at
HCs and be seen at OPD by the community, helping the community to see
page-520
them as health professionals. The FHT has also created a friendly relationship
between UHEPrs and other professionals because the FHT provides financial
support to the poor households identified during home visits. Providing a supply
of blankets, paying house rent, and offering community-based health insurance
are some additional activities that FHT deliver to the poor households.
Even though it is difficult to say that the FHT approach has disadvantages,
key informants reported that there is a huge burden in its management. Its
demands on human resources, transportation, and new health care delivery
systems (in relation to home-based drug supply and data sharing) were some
of the challenges observed during its application. In addition to the shortage of
labor for working at health facilities and in the community, physicians’ tolerance
for working under difficult conditions (e.g., traveling on foot, exposure to the
sun) was observed as a challenge.
2.3 Social Relevance and Acceptability of the UHEP

2.3.1 Acceptability of UHEPrs and UHEP services in the community
Participants were asked whether the HEP and UHEPrs were accepted by the
community. Most FGD participants said that the program and UHEPrs were
acceptable to the community. These participants related the acceptability of the
HEP to the positive contribution it has made and the benefits that community
members have received. The participants believed that acceptability had
improved over time. A participant from one of the health offices in Benishangul
“
Gumuz describes the situation:
When we look into community acceptance of the Health Extension

Program, I think it is better than in previous years. It is better than
the previous thoughts the community had about it, as, the community
saw the benefits and the changes that this program provided for
them.
Program officer, WorHO, Benishangul Gumuz
page- 521
On the other hand, some participants reported unacceptability of UHEPrs

among the community. Some of the reasons frequently mentioned for this
unacceptability were identified as the economic status of the households,
relatively better access to media as a source of similar information, and the
perception that the HEP was used as a political tool. Most participants raised
the idea that 1-to-5 networks and the HDA were considered political among
the community.
“
The 1-to-5 and development group discussion programs are the ones
that I believe I have not accomplished well. I haven’t done well on it
because the society rejects it and it is beyond my capacity.
“
UHEP supervisor, Bahir Dar, Amhara
I don’t think they accepted it as an important thing because if they

accepted it, they would implement it. Nowadays, everybody can
watch TV and maintain their health by themselves. They hate to
hear from the Health Extension Program itself.
UHEPr, Assosa, Benishangul-Gumuz
One UHEPr described situations in which the UHEPrs were verbally abused. The
“
UHEPrs reported that they were commonly insulted.
Most of the time, our communication is with stay-at-home mothers.

At times, these mothers may have adult children in their houses. If
arguments happen between us and the mothers because of their
shortcomings on the packages, lots of insults may come our way from
them. Sometimes the insults follow us outside after hours. But there
were no measures taken; sometimes I get confused on how we can
continue doing the work that we do. So yes, it is working some to me.
UHEPr, Bahir Dar, Amhara
page-522
Some participants believed that there was more political commitment a few
years back and that the acceptability of the HEP and UHEPrs was high. They
explained that the acceptability among the community has dropped over
time. The level of knowledge and the media access of urban inhabitants was
mentioned as among the challenges UHEPrs face. It was repeatedly mentioned
that resistance to the program has occurred even among model families and
model households. One reason mentioned for the declining acceptance of
the program was fatigue and higher community demand. The existing HEP
“
packages and services do not satisfy the demands of the community.
Now the community has a perception of relating HEP to political

issues. It is even challenging to organize discussions with the Health
Development Army groups. The HEP is weaker since 2008.…In our
catchment area, there are communities who have returned back to
open defecation after declaring open defecation free some years
ago, which can be an indication that the HEP is not going with same
energy as it was at the beginning.
“
How can a UHEPr advise a woman who has her own private doctor
by knocking on a door? There are UHEPrs who were told, “We don’t
like you” after they knocked on the door. There are rich households
who have their own private doctors; there are households who do
not open their door. Even after they enter the house, they are asked
questions beyond their capacity, and they feel like the town residents
don’t cover our gap. Then, they went to those households with average
or lower economic levels.
FGD participant, program officers, Tigray RHB
It is also common for inhabitants to resist some HEP services. For instance, it
was reported that people fail to visit a health facility for a cough and fail to use
ITNs. Furthermore, women are not empowered; thus, it is the partner’s decision
for a woman to seek maternal health services.
page- 523
“
Regarding malaria, they provide us with bed nets on a 6-month basis,
but we don’t use it properly. Some people even say, “I feel like as if I
am in a grave” and don’t want to use it. As a result, both children and
adults are suffering. That is why UHEPrs have focused on this. They
have worked a lot on this topic.
“
FGD participant, Assosa, Benishangul Gumuz
There are many who do not let their pregnant wives go to the HC for
an ANC and follow-up. There are many who insult them. They say,
“Is it because the UHEPrs or health offices advise you?” It is difficult
to convince husbands. There are many women who cannot convince
their husbands.
FGD participant, Assosa, Benishangul-Gumuz
There were participants who discussed UHEPrs’ transfer system and selection
as related to trust-building with the community and their acceptability by it.
These participants mentioned that the UHEPrs were assigned to the area and
“
the community where they grew up, costing them acceptability and respect.
Most of them were born and grew up there, .... And there is one
proverb from the Holy Bible that a prophet isn’t respected in their
home country. So, there is lack of trust. They will say, “How can
she help me to deliver when I raised her?” There is a challenge that
comes from the community.
Program officer, Assosa, Benishangul Gumuz
UHEPrs were also asked to rate their perception of the UHEP’s acceptance by
the community, based on their day-to-day observations and contact with the
community. The finding showed that 69% of UHEPrs perceived the program as
acceptable by the community; the result is highest for other towns (88%) and
lowest in Dire Dawa (26%) (Figure 2-2).
page-524
Figure 2-2: UHEPrs’ Perceptions about Acceptability of the Program
2.3.2 Community Preference and Perception about UHEPrs as

professionals
Participants had a mixed perception of UHEPrs as professionals. Some

participants believed that UHEPrs failed to follow professional ethics and are
hard to find during an appointment. On the other hand, some participants
believed that UHEPrs were committed and had a work plan that they followed.
These participants believed that the UHEPrs’ commitment was improving.
The participants mentioned taking care of elders, providing care to pregnant
women, treating sick children, supporting patients during referrals, providing
vaccinations, and supporting patients with chronic illnesses as the r esponsibilities
“
of UHEPrs.
In the past 6 months, things are going great. I have seen that they
are working with confidence. This is satisfactory. Previously, they set
appointments, and the appointments got canceled, but nowadays
this has changed. You don’t know why. It might be a problem among
the professionals. But now everything is going according to schedule.
Vaccinations, follow-ups, and other services are also given according
to the schedule.
FGD participant, community, Assosa, Benishangul Gumuz
page- 525
“
The UHEPrs have a serious plan. They have divided the year into
quarters. Their major activity is following up or doing village visits.
If there are pregnant women who need to attend follow-ups, if there
are severely sick people who cannot afford treatment at the hospitals,
they follow up and help them get a letter from the kebele to seek the
appropriate treatment.
FGD participant, community, Assosa, Benishangul Gumuz
In most of the FGDs, it was reported that the community preferred the UHEPrs
for services like HIV testing, under the assumption that they would be less
likely to be noticed and that better confidentiality would be maintained. The
participants explained that it was natural to trust someone who is closer.
UHEPrs’ home visits and friendly service were among the reasons frequently
“
mentioned for the UHEPrs’ being preferred over other health service providers.
There were more people getting tested in the kebele with them
[UHEPrs] than the ones at the Health Centers because the society
does not openly come forward.
WDA leader, Ebinat, Amhara
A policy advisor also said that having female UHEPrs was an important
strategy in addressing the large segment of the population’s problems and the
“
cultural contexts during home visit.
The first reason females have been selected is, obviously, because the
common health problems we have affect children and mothers.…The
other reason is that, if females go house-to-house, even if it varies
from culture to culture, mostly females have acceptance.
Policy advisor, MoH
The preferences of physicians and hospitals, however, were reported to be

common among people with a relatively higher income. There is also doubt
about the competence of the UHEPrs and quality of the services they provide,
which leads some residents to prefer other health service providers.
page-526
“
There are some challenges in urban areas. For example, in Bahir
Dar town, people with better incomes go to specialized hospitals for
medical services. So they don’t give credit for the advice or education
given by the HEP.
2.3.3 Community’s gender preference of UHEPrs
In this study, communities’ preference for UHEPrs’ gender was assessed. Overall,
the respondents said that it would be good for the UHEPrs to serve clients of
the same sex. Religion and culture were mentioned as important factors in
determining UHEPrs’ sex preference. Some participants argued strongly that
UHEPrs should be female, as it is much easier for women to share private
information with female health service providers. These participants argued
that the HEP requires frequent home-to-home visits, and women were much
likely to be visited; thus, UHEPrs should be females.
“
Most of our activities focus on women’s health and improving the
living standard at household level, which is mostly found during
women’s home-to-home visit. Under these conditions, it is better if
female UHEPrs are approaching them. So women clients are more
open with female than male UHEPrs, even for skilled delivery.
“
I don’t mind. But for us, what makes us choose one over the other is
just our culture. Both are professionals, but our culture has an effect
on it. For example, if I am pregnant and have to get checked by male
professionals, I will refuse because of the cultural influence. This needs
awareness. Even they are afraid of the women because they think she
will judge them tomorrow. This is the culture in the community.
page- 527
Some participants who said that female UHEPrs were preferred explained
that the sex of the UHEPrs would not be a problem. A considerable number of
participants said that the sex of the UHEPrs did not matter, but that their skill
“
in providing services should be taken into consideration.
I don’t mind if the UHEPr is male or female. The main thing is they
have to be skilled. We want that. If I tell my problem to either a male
or female UHEPr, it doesn’t matter. My transparency matters. As
long as he or she is a doctor or a UHEPr, it is a must that I can talk
about my problems openly. Because they bring me solution only if I
tell them my problems. Otherwise, they will not have a solution for
my problems. I believe in this, and I am happy if they are either sex.
Few respondents reported that it was easier to share secrets with male health
service providers and thus preferred UHEPrs to be males. They argued that
“
females do not respect each other:
For me, I prefer men, because when I tell my secrets to women, we

don’t respect each other. I am not ashamed to tell all my problems to
men. I prefer telling my secrets to men. I prefer if they are mixed. But
mostly I prefer male UHEPrs.
The UHEPrs recommended having a team of male and female UHEPrs to

be assigned to each catchment so that they could serve clients based on their
“
preferences.
Now we all are females who are working, so I will make the gender
distribution 1 female and 1 male in the catchment. It is well known
that females are more compassionate. The community feels at ease
sharing their problems with females. I will make male to female
distributions because the woman will tell what she wants to tell me
alone separately, and also as there are male customers who will
possibly tell their problems to male professional that they do not
want to tell female professionals.
UHEPr coordinator, Addis Ababa
page-528
2.3.4 Trust in the UHEPrs

In this study, we assessed whether the community trusts UHEPrs. Most participants
stated that the UHEPrs were trusted in the community. Respondents explained
that community members, including those infected with HIV, trust the UHEPrs
“
and share their private information with them.
The community has a familial relationship. For example, Antiretroviral

therapy users tell their secrets when UHEPrs go home to home. Then,
they give education and ask why they discontinued ART drugs,
saying that it is not appropriate. 1-to-5 networks inform UHEPrs
about who is pregnant. Again, bed nets are given at the Health
Center and in their catchment area. The people request bed nets
from the UHEPrs. For those who have lost their mothers and fathers
because of HIV, UHEPrs link with NGOs. The society also tells the
UHEPr that someone is weak and requests her support. For example,
those who move using wheelchairs request that the UHEPr support
them in joining the organization called…is it “Kannshyer?” That is why
I say they have trust in UHEPrs.
Community member, Bahir Dar, Amhara
Some respondents attributed community acceptance of and trust in the UHEPrs

to the positive changes that the HEP has made, including reduced occurrence
of CDs and maternal health problems. One UHEPr from the SNNPR describes
“
the situation as follows:
The community has more trust in the Health Extension because of

the change they have made in different ways, in urban and rural
areas too. They have implemented the HEP packages at household
level. It’s not easy to convince the community, so they have trust in
the HEP. Because all transmitted diseases, mother and child, and
all the burden of diseases have decreased because of the Health
Extension Program. So, the community has a great belief in the
Health Extension.
UHEPr, SNNP
page- 529
Another respondent mentioned the involvement of the UHEPrs in the referral

and transportation of a laboring woman as one of the contributing factors to
“
the community’s trust in the UHEPrs.
When a woman is going into labor, they call and ask me for the
ambulance number. They call me late at night. They think I can solve
their problems, help them, there is trust. They also call and ask me to
come when something happens, and I go. This is trust. They believe
that I can solve their problems, that my teachings are good for them,
and that they have changed them. So, the trust is there in all of us.
UHEPr supervisor, SNNP
A participant from Assosa similarly describes the acceptability of the UHEPrs

given their contributions to improving the lives of the community and personal
“
hygiene.
Their advice has made us live an improved life. My neighbors also

wish them long lives. Littering has decreased now, I have also heard
a mother scolding her child when he came out of the toilet without
washing his hands. This education is from the UHEPrs, and everybody
is happy with their work. I have never heard people saying negative
things about them, even from the initiation of the project.
“
Female community member, Assosa, Benishangul-Gumuz
There is tremendous progress because they teach us how to properly

use the lintel, we fully trust them. They give us advice and teach in
a good way. They cleaned us up and helped us with toilets. Because
they taught us about dry and wet, it has correctly improved, and
there is a benefit, and they continued. It is good, and we trust them.
Some participants related the acceptability of and trust in the UHEPrs to

their selection and recruitment process. A participant from Benishangul-Gumuz
describes the situation:
page-530
“
In the first place, their recruitment process has involved the community,
and they were selected because they were believed to know the
community and its problems very well. And it was believed that after
they came back from training to where they have were assigned, they
would serve their community very well, and they are doing that. Thus,
the community trusts them.
Similar findings were reported by the UHEPrs. One UHEPr from Harari
“
explained the situation as follows:
When I appoint them, they respect me and fulfill my appointment.

When I inform them about receiving a service, they come to where
I told them to come, and when I call them for a meeting, they are
never absent from that meeting.
UHEPr, Harar
On the other hand, one female UHEPr reported as common that some UHEPrs
fail to cover all the areas in their catchment given the size of the population,
which she believed to be one source of distrust of the UHEPrs and their reports:
“
The community doesn’t have trust in us. Because of the size of the
kebele, which is vast for me, there might be areas that I didn’t cover,
so when I am presenting my reports that I have done it, they will say
that’s a lie because I didn’t address of all. They thought I did not do
any of it.
Lack of community awareness of the UHEPrs’ duties and responsibilities was

mentioned as one source of distrust of the UHEPrs. Repeatedly mentioned was
that many communities had residents more knowledgeable than the UHEPrs.
UHEPrs’ absenteeism from work was also mentioned as one possible source of
distrust for the UHEPrs among the community.
page- 531
“
The other thing is that to sustain the program, sectoral supports are
needed. For example, if a woman tells a UHEPr that she doesn’t
have access to water, the UHEPr is not a water resource official, so
she cannot provide her that, but she advises the woman about the
importance of water, like in sanitation. If there is no access to clean
drinking water, the community blames her for not providing that. The
society does not differentiate the tasks of different sectors’ offices.
They will just say to her, “Without providing the water, you just nag
us about clean water.” They blame her for not being able to provide
that.
“
RHB head
People who are living in the urban areas have better understanding
because they are more exposed to the media and so on. So, the
main thing is that we need to come up with better knowledge. So the
Health Extension Workers should have better knowledge than urban
people. They need to provide the education and be prepared for the
community to be expecting that. The other thing is that, other than
the information that they get from the media, the community needs
practical service. For example, it may need to be able to measure
blood pressure, and they may need the blood sugar test service in
their homes. Nutrition and other things. They need technical service
that they can get from the professional. Because of the lack of this
integration, the community trust decreases. So if you go to people
who know better, while only having theoretical knowledge, the
acceptability is known. So they need more service than this. And
because of that, the acceptability is difficult. They will not say okay if
you ask him to provide education. But the acceptability of measuring
those things is relatively better.
HC Head, Assosa, Benishangul Gumuz
Refusal of service use, like child vaccinations, was reported to be a sign of

distrust in the UHEPrs’ qualifications.
page-532
“
They may hesitate to accept services that are important for them. For
example, they hesitate to have vaccinations, and refuse to take the,
considering that it causes the problem. They distrusted the capability
of the UHEPr.
UHEPr, Kokosa, Oromia
2.3.5 Friendliness of UHEPrs toward community members
Survey participants were asked to rate UHEPrs’ friendliness while providing

services for the community at the household level. The result showed that the
vast majority of respondents (70%) perceived that the UHEPrs assigned to
their villages were friendly. This figure varies regionally; a higher proportion
(79%) of respondents from Addis Ababa perceived UHEPrs as friendly, versus
54% of those from Dire Dawa. Similarly, 69% of household survey respondents
agreed that UHEPrs were courteous and polite while providing service (Figure
2-3 and Figure 2-4).
Figure 2-3: Community Perception of UHEPrs’ Friendliness
page- 533
Figure 2-4: Community Perceptions of UHEPr’s Courtesy

In line with the quantitative results, most of the community members (i.e., FGD
participants) described the UHEPrs as passionate, motivated, honest, friendly,
and providing services with hospitality. The participants reported that the
UHEPrs tried to build rapport with the community by taking part in different
“
social events, like the coffee ceremony.
Well, in their work motivation, their love for the work, by their behavior,
like getting tea or coffee out of their own pockets, they are trying for
the community to be closer. Just their honesty is seen by that. By their
behaviors and having a great relationship with the community while
drinking tea and coffee and saying goodbye with love. Just even
sacrificing their time. Well, nobody forces them to go in on Saturdays
and Sundays, but every Saturday and Sunday they hold it down and
call a meeting. That is what gives them work motivation. No matter
what, they receive us with a smile.…When I go to them, they get up
from their chairs and ask, “What is wrong, Mother, are you okay?” So,
this means they’ve made me closer, and I tell them my worries.…The
first thing from a human being is their smile to get you closer.
page-534
2.3.6 Community satisfaction with the UHEP services
Most participants described the community as being satisfied with most of

the UHEP packages and services. More specifically, community satisfaction
was reported for immunization service and follow-up for patients with chronic
diseases (Figure 2-5). There was, however, a concern that the services provided
were below the expectations of the community. The participants reported
that health problems like cancer were not given due emphasis by the UHEP.
It was also repeatedly mentioned that the UHEPrs lack important medical
equipment and supplies that are required to provide different clinical services
and chronic patient follow-up, such as blood sugar tests, BP apparatus, and
HIV test kits. The lack of this equipment and these supplies were reported
as a major implementation challenge that affected UHEPrs acceptability by
the community and remained a major factor that reduced the community’s
satisfaction with the program.
UHEPrs have no equipment for measuring blood pressure or blood glucose

levels. So, when she knocks on a door and measures their blood pressure, they
are happy because we have seen it in the Wukro pilot. Urban JSI provided
them with materials to pilot in one kebele. When the HEPr measures their
blood pressure or blood glucose level and tells them the result, they are happy,
and they go to the health facility for further services. So, the lack of timely
required materials and expected service from the UHEPrs may compromise the
quality of their service.
page- 535
“
There is a huge gap between what the community demands and what
they are expected to deliver. Some communities are faster than the
UHEPr. So, the curriculum should be revised. Why didn’t the UHEPrs
assist with delivery and investigate if the curriculum is revised? So,
I think there is a problem with the curriculum since it is too old.
Diseases are becoming more advanced. They may not learn details
about cancer, but they have to identify any cancer case present in
the community. If they didn’t learn that in the curriculum, how could
they do that? There is no timely health-seeking behavior in the rural
population. Another issue is that, is 1 year’s training enough for those
all HEP packages?
HC Head, Aboker, Harar
Figure 2-5: Overall Satisfaction of the Community with the UHEP
page-536
page- 537
CHAPTER 3
Findings:
Availability and
Adequacy of
Resources for UHEP
page-538
CHAPTER 3
3 Findings: Availability and
Adequacy of Resources for UHEP
3.1 Availability of Human Resources for

Implementation of the UHEP

According to the newly revised UHEP implementation manual, UHEPrs are
directly accountable to the catchment HC, while HC management is responsible
for both technical and administrative issues related to the program. HCs deliver
all required pharmaceuticals, supplies, and equipment important for UHEPrs
as per the implementation manual. HCs are also expected to assign experts to
coordinate the implementation of the UHEP.
According to the findings of the HC assessment, only 64.7% of HCs reported

that they were responsible for both administrative and technical issues in the
implementation of the UHEP. Staff responsible for the coordination of the
UHEP was available in 85% of HCs. Only 53.1% of the HCs provided a separate
office space and furniture for the HEP/HEPrs (Table 3-1).
Table 3-1: Administrative and Technical Roles of Health Centers for

implementation of the UHEP
Addis Ababa Dire Dawa Other Towns National

Characteristics
N % N % N % N %
Role of HC (n=133)
Administrative oversight 51 55.4 7 100 28 82.4 86 64.7
Technical oversight 71 77.2 7 100 29 85.3 107 80.5
Both administrative and technical
51 55.4 7 100 28 82.4 86 64.7
oversight
Availability of UHEP coordinator
(n=133)
Yes, UHEP coordinator 78 84.8 5 71.4 30 88.2 113 85.0
Yes, other HC staff 10 10.9 0 0.0 4 11.8 14 10.5
No 4 4.3 2 28.6 0 0.0 6 4.5
page- 539
Findings: Availability and Adequacy of Resources for UHEP
Allocation of UHEP budget at HC

(n=131)
Yes 9 9.8 3 42.9 2 6.3 14 10.7
No 83 90.2 4 57.1 30 93.8 117 89.3
Availability of office space and
furniture for UHEP at HC (n=128)
Yes, adequate 60 65.2 3 42.9 5 17.2 68 53.1
Yes, but not adequate 30 32.6 2 28.6 8 27.6 40 31.3
No 2 2.2 2 28.6 16 55.2 20 15.6
Abbreviation: HC, Health Center.
3.1.1 Training and Career Development for the UHEP
About 60% of HCs reported that the UHEP coordinators had attended
training on the UHEP packages, 56.4% on HEP implementation guidelines,
56.4% on Integrated Refresher Training (IRT) for the HEP, 57.1% on supportive
supervision, and 30.8% on the Community Health Information System (CHIS;
see Figure 3-1).
In nearly three quarters of HCs, short-term training for UHEPrs was provided
in the last 2 years. The mechanisms of training initiation, as mentioned by the
heads of the HCs, were: 52.4% initiated based on supervision feedback, 41.3%
on the directives from sub-city/WorHO, and 32.5% and 20.6% on the UHEPrs’
and partners’ requests, respectively (Table 3-2).
This study also assessed the availability of opportunities for UHEPrs to upgrade
their education. More than half (58.1%) of respondents reported that the HC,
along with other concerned bodies, provides an opportunity for UHEPrs to
upgrade their education: i.e., from level IV (diploma) to level V (degree).
Regarding the development of HEPrs across the career structure, 78.3% of
respondents reported that HCs periodically opened opportunities for HEPrs,
based on the rules and proclamations of the Ethiopian civil service. (Table 3-2).
page-540
Table 3-2: Training, Education, and Career Development Opportunities for

UHEPrs at HC Level
Addis Ababa Dire Dawa Other towns Total

N % N % N % N %
Provision of trainings by the HC (n=126)
Yes 72 78.3 2 66.7 18 58.1 92 73.0
No 20 21.7 1 33.3 13 41.9 34 27.0
Mechanisms of training initiation (n=126)
Feedback from supervision 51 55.4 2 66.7 13 41.9 66 52.4
Partner’s request 16 17.4 1 33.3 9 29.0 26 20.6
UHEPrs’ request 31 33.7 1 33.3 9 29.0 41 32.5
Directives from sub-city/
33 35.9 2 66.7 17 54.8 52 41.3
WorHO
Other mechanisms 1 1.1 1 33.3 3 9.7 5 4.0
Availability of opportunity for UHEPrs to upgrade their education (n=129)
Yes 59 64.1 0 0.0 16 53.3 75 58.1

No 33 35.9 7 100.0 14 46.7 54 41.9
Provision of career structure for UHEPrs (n=129)
Yes 75 81.5 7 100.0 19 63.3 101 78.3
No 17 18.5 0 0.0 11 36.7 28 21.7
Abbreviations: HC, Health Center; WorHO, Woreda Health Office.
page- 541
Figure 3-1: Training Status of UHEP Coordinators at Health Center level
The UHEPrs were reported to have educational and career development

opportunities. UHEPrs are provided the opportunity to join health offices. A
committee uses a predetermined set of criteria to select and offer UHEPrs
educational opportunities. UHEPrs’ performance was reported to be among
the criteria used to provide educational opportunities and career development.
Some UHEPrs, however, complained that the selection process lacked clarity
“
and believed that there was partiality.
The office workers doesn’t pay attention to us or consider us regarding

career structure or professional development. They provide the
opportunity to someone who is close to them. If I work 3 or 4 years in
this kebele, they won’t give me a transfer opportunity.
UHEPr, Surupha, Oromia
On the other hand, limited educational and career development opportunities

create frustration and demotivation among UHEPrs. Participants from Oromia
and Tigray noted that it takes more than a decade for some UHEPrs to get
an opportunity for further education. Another important concern mentioned by
participants was that their career development is not considered even though
page-542
they have degrees from private colleges. Even technical people from RHBs
believe that the education and career development opportunities available to
“
UHEPrs are inadequate.
The situation is discouraging. Sometimes they prefer to learn in

private schools rather than public schools. But if they attend a
private school, it will not help them for promotion.
UHEPr supervisor, Amhara
More importantly, UHEPrs reported often lacking the freedom to ask for the
benefits to which they are entitled One participant from Assosa described
“
thesituation as follows:
Even we asked for other benefits, the Health Centers and health
offices would say, “You can resign.” They always think about firing us.
The qualitative findings also broadly explain the inadequacy of training

opportunities for UHEPrs. Generic UHEPrs are trained for 3 years on the
UHEP. After graduation, they are provided on-the-job and refresher training.
Participants acknowledged that there are refresher trainings but felt that
they were inadequate. Short-term trainings were reportedly provided in the
areas of HIV, immunization, nutrition, TB, ANC, and family planning. IRT was
reported to be the most useful training for addressing the UHEPrs’ gaps in
the implementation of the UHEPrs. The participants supported the generic
UHEPr training program and felt that it could solve the problems related to
the deployment of diploma nurses. The limited capacity of the training colleges,
however, was mentioned as a reason for the poor competency of the UHEPrs,
“
one of whom explained the importance of IRT thus:
I learned a lot from the IRT program. I have received 2 different

trainings since I joined. It was really helpful to update what we
already knew.
UHEPr supervisor, Bahirdar, Amhara
page- 543
There were reports that rural HEWs were transferred to towns/cities and
deployed as UHEPrs without additional training on the UHEP. This trend is
“
believed to create a gap in the implementation of the UHEP.
Even those who came here via promotion from rural HPs want only to
enter the town without any interest in the work by moving from place
to place within the town. It would have been better if the UHEPrs
were trained generically to work as urban UHEPrs because towns
have their own unique characteristics.
Program officer, Tigray RHB
Short-term training is organized to motivate and build the capacity of

UHEPrs, and create an experience-sharing platform. Moreover, whenever there
is new development, trainings are arranged to familiarize UHEPrs with it. The
increasing demands of the community also necessitates additional trainings to
satisfy the health service needs of the community. The IRT program is considered
valuable for UHEPrs because it saves time that would have been needed for
separate refresher training.
The participants also identified several limitations of the existing training

system. On the one hand, the opportunities to participate in refresher training
are limited. Participants also reported that the trainings were not effective in
creating attitudinal changes or motivation among some UHEPrs, necessitating
the evaluation of different trainings, including the curriculum for pre-service
training. Moreover, the limited opportunities to upgrade their current education
level were reported by some UHEPrs.
The participants advised building the capacity of the UHEPrs to better satisfy
the needs of the urban community. There was also a strong recommendation
to evaluate the existing UHEPr training curriculum. The participants also
suggested evaluating the relevance of the various training for UHEPrs and
considering training the FHT.
As shown in the graph below, the proportion of UHEPrs who did IRT was low
for most packages; only 59.7% of UHEPrs received refresher training on the
social behavioral change communication (SBCC) module, and the proportion
for the first aid, maternal and child health, and WaSH modules was 67.4%,
71.9%, and 72.3%, respectively (Figure 3-2).
page-544
Figure 3-2: Integrated Refresher Training Status of UHEPrs
3.1.2 Adequacy of UHEPrs (UHEPr-to-Household Ratio)
As described in the revised UHEP implementation manual, a UHEPr is expected

to serve 400-500 households. The number of visits for the household might vary
depending on their health needs and economic status. Households with poor
economic conditions and maternal and child health needs are the first priority,
while households with the highest income are the lowest priority. Nevertheless,
the UHEPrs should visit all households assigned to them, regardless of their
economic status and health needs, at least once a year, to collect basic
profile data about the household members and housing conditions.
According to the qualitative findings, the high workload due to the inadequate
number of UHEPrs was the major reason mentioned for UHEPrs’ demotivation
and dissatisfaction. The shortage of UHEPrs was also mentioned as a major
reason for the insufficient implementation of the program, mainly with respect
to the low proportion of households visited. Like the qualitative findings, the
quantitative assessment showed that a higher proportion of UHEPrs were forced
to serve more than 500 households. A relatively lower proportion of UHEPrs
page- 545
(25.7%) from Addis Ababa serves more than 500 households; the proportion
is much higher in Dire Dawa, in which 78.2% of UHEPrs were assigned to serve
more than 500 households. In general, both the qualitative and quantitative
findings clearly indicate the inadequacy of the number of UHEPrs assigned to
the existing households (Figure 3-3).
Figure 3-3: Urban Health Extension Professional-to-Household Ratio
3.2 Availability and Adequacy of medical supplies

and equipment
The availability of medical equipment and their use during home visits was
reported to be vital to increasing the satisfaction of the community. Medical
equipment, however, was not adequately available so that UHEPrs could
implement the packages during their home visits. For instance, only a few
UHEPrs had BP apparatus for screening hypertension during home visits,
although UHEPrs are expected to do screening of NCDs during home visits and
refer patients to health facilities as needed. Similarly, a shortage of HIV test
kits was cited as a major impediment to providing HIV screening services. As
reported by the majority of key informants, a shortage of medical supplies and
equipment remained the major impediment to the implementation of UHEP;
this resulted in poor implementation of the program, demotivated UHEPrs, and
decreased their acceptance by the community.
page-546
“
…shortage of HIV tests in the Health Center. Voluntarily, I work in the
Amanuel group to encourage people to do HIV testing, but in the
majority of cases, clients returned home without getting the services
because of HIV test kit limitations.
FGD, WDA member, Shewa Robit, Amhara
Moreover, shortages of drugs and medical equipment were also reported, which
resulted in unnecessary referrals of some clients and patients. Some patients
were forced to buy drugs from expensive private health facilities because of a
shortage of such drugs in public health facilities. The unavailability of drugs and
other supplies was also reported to be contributing to community dissatisfaction.
Key informants also highlighted the reasons for the unavailability and shortage
of drugs and other medical equipment and supplies, including a shortage of
drugs on the market, a lack of separate budget for the HEP, and an irregular
purchase plan and absence of an effective drug supply management system.
The quantitative assessment also revealed that only 54.6% of HCs had a
standard list of medicines and supplies for the UHEP. As shown in the figure
below, a majority of HCs consider the provision of medical equipment for
UHEPrs. Availability of the most important equipment and supplies was more
than 80%: for instance, a BP apparatus (84.1%), MUAC tape (94.7%), vitamin A
(90.9%), and a first aid kit (88.6%). Nevertheless, the current study discovered
an inadequate availability of equipment, such as a weight scale (37%) and
height scales (38%), which are critical for conducting nutritional assessment
and growth monitoring (Figure 3-4).
page- 547
Figure 3-4: Availability of Pharmaceuticals, Medical Equipment, and Supplies
HC heads were also asked about the frequency with which medical supplies (e.g.,
medicines, test kits, contraceptives, MUAC tape, gloves, and alcohol, among
others) were refilled or supplied to UHEPrs. A considerable number (38.2%)
of respondents mentioned that they filled supplies monthly & whenever
requested by UHErs. More than 90% of HCs provided stationery and
recording andreporting tools and forms for their respective UHEPrs (Table
3-3).
page-548
Table 3-3: Logistics and Supply Chain Management of UHEP at Health Center
Level
Addis
Dire Dawa Other Towns Total
Ababa
N % N % N % N %
Standard list of medicines and supplies for UHEP

Available verified 60 65.2 0 0.0 11 35.5 71 54.6
Available but not verified 25 27.2 7 100.0 6 19.4 38 29.2
Not available 7 7.6 0 0.0 14 45.2 21 16.2
HC supplies catchment UHEPrs with drugs and/or supplies
Yes 89 96.7 6 85.7 21 65.6 116 88.5
No 3 3.3 1 14.3 11 34.4 15 11.5
Forecasting and quantification of UHEP
Yes 74 80.4 7 100.0 16 50.0 97 74.0
No 18 19.6 0 0.0 16 50.0 34 26.0
Frequency of refill
Monthly 34 37.0 4 57.1 12 38.7 50 38.2
Quarterly 8 8.7 1 14.3 6 19.4 15 11.5
Bi-annual and annually 5 5.4 0 0.0 0 0.0 5 3.8
Whenever requested by UHEPr 41 44.6 2 28.6 7 22.6 50 38.2
There is no defined refill frequency 5 5.4 0 0.0 6 19.4 11 8.4
Supply of inputs for UHEP
Stationery materials 91 98.9 7 100.0 28 84.8 126 95.5
Recording tools/forms 90 97.8 7 100.0 31 93.9 128 97.0
Reporting formats 90 97.8 7 100.0 32 97.0 129 97.7
Guidelines for UHEP 83 90.2 7 100.0 22 66.7 112 84.8
page- 549
3.3 Availability of UHEP Service Standards, Manuals,

and Guidelines
According to program staff, materials and supplies for the implementation of

the UHEP were obtained from the catchment HCs. Key informants mentioned
that necessary materials, including medicine and equipment, were obtained
from catchment HCs, as recommended by the UHEP implementation guideline.
They also reported, however, that logistics and supplies were not available in
adequate amounts or with an acceptable level of quality. For instance, key
informants frequently mentioned the unavailability of manuals and guidelines
on the different packages of the UHEP, shortages of essential drugs, and bags
for the FHT to use for carrying drugs and supplies during home visits.
Guidelines and strategies are the key factors for effective implementation
of the health care policy. Participants reported that those HEP standards
and strategies considered the context, such as population dynamics, during
recruitment of UHEPrs.
Regarding the development of guidelines and strategies, participants reported

that they were prepared at the national level by the MoH and then distributed
to RHBs, ZHDs, WorHOs, and HCs. Simple brochures for health education
are also prepared at the RHB level.In general, the HEP and PHC directorate
in collaboration with other directorates, is responsible for the preparation of
standards, manuals, and guidelines.
Development of HEP standards was reported to have considered inputs at

different levels. The participants reported that the development of the standards
considered SDGs, Health Sector Transformation Plan (HSTP), professional
“
experience, and the needs of the target groups.
When we design policy, we should first survey the community’s

problems, priorities, and purposes.…This one [the current strategy] is
good; it has considered the SDG and HSTP, even though we have
implementation problems at all levels…
HPDP process owner, Addis Ababa
page-550
The participants also stated that the manuals were developed based on
“
knowledge, attitudes, and practices in a specific package.
Let us take hygiene and sanitation….It has 7 components. If we need

to bring behavioral change in a person as well as implement lectures
taught at their home, we have to teach knowledge-, attitude-, and
skill-related issues regarding hygiene and sanitation. So, we develop
manuals that include the 3 aspects, knowledge, attitude, and skill….
That is how we deliver the urban and rural program to bring
behavioral change.
HEPr coordinator, RHB
Concerning intervention at schools, participants indicated that the guideline

had been developed recently to facilitate implementation. The manuals are
reported to be comprehensive and distributed widely to their intended users.
“
However, implementation is still very limited because of other challenges.
The basic problem with the school health service is the absence of
guidelines on how service is to be offered. Recently, the guidelines
were developed, and orientation was given to the HEPrs and other
responsible bodies….There were some limitations, like fulfilling hygiene
and sanitation facilities, carrying out regular check-up of students,
and making available the immunization service in a timely way.
UHEPr, Dire Dawa
“
Before these manuals were sent down, professionals from every Health
Center, like family health coordinators, disease prevention people,
reviewed the manual, assessed its strengths and its weaknesses.…
The manuals are comprehensive, and they consist of the programs
executed at the Health Center.
HPDP process owner
The availability of the UHEP guidelines was assessed during the HC survey,
and the result showed that about 85% of HCs had UHEP implementation
page- 551
guidelines; availability of the guidelines at the HC, however, was very low in
other towns (66.7%) compared to those in Addis Ababa and Dire Dawa (Figure
3-5).
Figure 3-5: Availability of UHEP Implementation Guidelines among Health

Centers
3.4 Periodic update of guidelines
In the process of providing health care service, manuals and strategies were
updated or developed based on the health needs of the community, the
prevalence and burden of disease and learned experience. Participants also
expressed that new health care packages were added.
“
There were 16 health care packages at the beginning, but it has now
been increased to 18 packages….One is epidemiological transition.
In particular, non-communicable diseases have been included in the
points of intervention.
RHB Head
page-552
Participants also stated that manuals like the FHT manual were updated in
“
2009 EFY (2017) and also updated every 3 years.
The manuals have been revised in the case of Addis Ababa....The HEP
implementation manual was revised in 2009. It clearly elaborated
how the Family Health Team approach should be, how the team
should be organized and execute…
UHEP sub-city team leader, Addis Ababa
Some participants stressed that the strategies need to consider the difference
in special cases at city and regional levels during amendments and periodic
“
modifications.
It is different in the cities and at the regional level. So, the design has
to be considerate of these differences and has to be a design that
enables the data of special cases to be tracked. Especially for the
indicators, I think it is good.
PPMED, MoH
3.5 Clarity and comprehensibility of HEP guidelines

The clarity of guidelines and strategies was the other point raised by most
participants. Participants reported that there were no problems regarding
clarity; it included the roles and responsibilities of each stakeholder, but effective
use of the guidelines during implementation was the challenge.
page- 553
“
Most of the guidelines strengthen the HEP, so they are very important
if we implement them properly. There are no problems with the
guidelines….There may be with the implementation, but not with the
guidelines themselves.
HEP coordinator, RHB
Participants also identified the preparation of manuals in local languages as

one factor for improving the clarity and comprehensibility of the guidelines.
“
It [the guideline] has a national and regional standard….It is prepared
in local languages to make it more understandable by users and
more familiar for the UHEPrs using it.
page-554
page- 555
CHAPTER 4
Findings: UHEP
Workforce
Analysis
page-556
CHAPTER 4
4 Findings: UHEP Workforce Analysis
4.1 Knowledge and Skills of UHEPrs
The study assessed the knowledge and skills of UHEPrs using a self-reported
competency tool with different questions to assess the skills of UHEPrs at
implementing UHEP packages. UHEPrs were asked to rate their own competency
using a scale ranging from 0 (not able to try it at all) to 4 (can trainothers
to do it). Later, responses of “I am not able to do it” and “I can do it withsome
guidance” were recorded as “not competent,” and “I can do it by myself”and “I
can train others on how to do it” were recoded as “competent.”
This study revealed that only 69% of UHEPrs reported having the knowledge
and skill to manage common childhood illnesses, and only 60% could perform
ANC. The percentage of UHEPrs who could provide FP was lower than in other
areas of competency, and insertion of IUDs was the lowest (29.5%), followed
by providing injectable FP methods (45.7%). The proportion of UHEPrs who
reported competency in providing child immunization (BCG or Heptavalent)
was also among the lowest (60.7%);(Table 4-1).
Table 4-1. Self-Reported Level of UHEPrs’ Competency to Conduct or Provide

Health Services
Addis
Dire Dawa Other towns Total
Self-reported level of UHEPrs’ Ababa
competency N % N % N % N %
Develop individual plan (weekly,
353 92.7 69 79.3 82 76.0 504 87.5
monthly, quarterly, and annual plan)
Collect and analyze population health
347 91.1 59 67.8 80 74.1 486 84.4
data
Perform disease surveillance 340 89.2 60 68.9 82 76.0 482 83.7
Conduct growth monitoring 354 92.9 74 85.0 82 76.0 510 88.5
Manage common childhood illnesses 261 68.5 67 77.0 70 64.9 398 69.1
page- 557
Conduct nutritional screening for

361 94.7 76 87.4 91 86.7 528 92.2
under-5 children
Take history of pregnant mothers 323 84.8 72 82.8 79 73.8 474 82.4
Perform antenatal care/examinations 227 59.5 52 59.8 76 77.5 355 62.7
Identify danger signs during ANC
visits and refer clients to health 337 88.5 72 82.7 87 81.3 496 86.3
facilities
Provide child immunization (BCG or
189 49.6 69 79.3 91 85.1 349 60.7
Heptavalent)
Provide vitamin A supplements for
362 95.0 73 83.9 94 87.1 529 91.9
children
Provide injectable family planning
174 45.6 73 83.9 16 14.9 263 45.7
services
Provide counseling service on
342 89.8 70 80.4 85 78.7 497 86.3
adolescent and youth RH
Abbreviations: ANC, antenatal care; BCG, Bacillus Calmette-Guerin.
Community members who were included in the household survey were also
asked to rate the level of UHEPrs’ knowledge and skill in providing health
education and counseling services during their home-to-home visits. The results
showed that majority (68%) of respondents perceived UHEPrs as having
adequate knowledge and skills regarding health and health-related issues.
Some respondents (13%), however, showed reservations (Figure 4-1).
Figure 4-1: Community Perceptions about UHEPrs’ Knowledge and Skills
page-558
The quantitative findings are supplemented by the qualitative results.

Participants repeatedly mentioned knowledge and skills gaps among UHEPrs
in providing counseling and health education on emerging health problems
like NCDs. Others, mainly program staff, suggested the provision of extensive
technical support to enhance the competency of UHEPrs and thereby enable
them to respond to the changing health service needs of the urban community.
Participants also suggested the need to revise the UHEP pre-service training
curriculum to fit UHEPrs with the current epidemiological transition to NCDs,
as the urban population is more affected by them due to lifestyle factors.
Study participants also raised their concerns regarding the short duration of
pre-service training, the quality of the training approach, and the selection
procedure for recruiting UHEPrs as a possible source of their low competency.
They also mentioned that the failure to pass the Certificate of Competency
(CoC) exam as an indicator of UHEPrs’ lower competency.
UHEPrs also mentioned their need for further training on clinical issues, as they
had forgotten what they learned at college (in their pre-service training). One
“
UHEPr from Benishangul Gumuz said:
It is insufficient to provide medical care. And we have almost forgotten

what we learned. The training is not enough, and we need additional
trainings to provide clinical care.
UHEPr, Benishangul-Gumuz
Similarly, program staff perceived that UHEPrs lack adequate knowledge

for educating some segments of the population, as some households have
better knowledge than the UHEPrs; thus, academic support should be given.
UHEPrs face challenges from the highest-income families, as they are unable
to address questions raised by them during home visits. On the other hand,
some participants perceived that UHEPrs had no knowledge or skills problems
because they were receiving short-term training, which helps increase their
knowledge and skills.
page- 559
“
There are people in the community who have better knowledge than
the UHEPrs. They might be challenged by these people whenever they
conduct health education. So, there is a need to provide academic
support to the UHEPrs
HC Head
“
I suggest that, if they get education or do long- or short-term trainings
periodically, they will benefit, as their current level is good, but not
sufficient. I don’t think it is adequate. I want them to be more than this.
It is good if any service provider gets educated or trained because we
are talking about medications and life-related things here. I say that
they should have the opportunity to pursue their education.
HC Head
4.2 Job Satisfaction and Motivation
UHEPrs’ job satisfaction was assessed using a Likert-type scale instrument

consisting of 27 items, with a 5-point scale ranging from 1 (strongly dissatisfied)
to 5 (strongly satisfied). The items were categorized into 8 dimensions of
satisfaction: leadership, promotion, autonomy, work environment, professional
training opportunities, job security and salary, recognition at work, and perceived
other job opportunities. The reliability test of the instrument indicated good
internal consistency, as shown by a Cronbach’s alpha value of 0.88.
Overall satisfaction was calculated by using the mean value of the scale as
a cut-off point to categorize respondents as either satisfied or dissatisfied.
Accordingly, respondents scoring above the mean were categorized as satisfied,
and those scoring at and below the mean were considered dissatisfied. Similar
procedures were followed to calculate HEPrs’ level of satisfaction by each
dimension.
page-560
The assessment showed that the overall satisfaction of HEPrs was 59.4%. The
highest satisfaction was observed in Dire Dawa (79.3%), followed by Addis
Ababa (58.3%). Only 11.6% of UHEPrs were satisfied with their salaries. The
level of satisfaction with job security was lower (5.3%) in Addis Ababa than Dire
Dawa and other towns. Similarly, one third (34.6%)of UHEPrs were satisfied
with their opportunities for promotion, and 36.6% of UHEPrs were satisfied with
the availability and appropriateness of their professional training opportunities,
orientation for new staff, and opportunities to participate in research activities. In
addition, about half of UHEPrs were unsatisfied with the level of administrative
support and recognition they received from the leadership.
On the other hand, a majority (76.4%) of the respondents reported that they
were satisfied with their level of autonomy to make decisions and be fully
accountable for those decisions. The vast majority (83.4%) of UHEPrs stated
being satisfied with the work environment and the relationship among staff
members (the extent to which the working environment encouragedthem
to adjust their practice to suit community needs, provided a stimulating
intellectual environment, and expanded their scope of practice).
UHEPrs were also asked whether their level of job satisfaction was increasing
or decreasing over time, and the majority (66.5%) said that it had been
decreasing. Addis Ababa had the highest proportion of UHEPrs (71.3%)
reporting a decreasing level of satisfaction compared to Dire Dawa and other
towns (Table 4-2).
The findings from the qualitative study were largely supportive of the quantitative
results. Serving the community and solving its health problems were identified
as the main sources of satisfaction for UHEPrs, whereas the inadequacy of
salary and lack of incentives were their major sources of dissatisfaction. UHEPrs
perceived their salary as being inadequate for the type and volume of work
they were doing.
page- 561
“
The other issue is salary….The other reason for them [UHEprs)] to
lose hope is that there is no nationally-set transfer criteria [in a loud
tone and angry]. We just set our own a year ago. NOTHING!
UHEP expert, Addis Ababa
The participants also reported a weak recognition system for the best-performing
“
UHEPrs.
There are no motivational packages at every level; the system treats

those who perform well and badly the same. There is not even
anything that recognizes or incentivizes the best performers according
to the balanced scorecard. At the facility level, they give certificates,
and money is given at the sub-city level, but the question would be
whether we are providing the desired incentives or not.
HPDP process owner, Sub-City Health Office, Addis Ababa
Participants stated that the availability of training opportunities was a major

reason that most UHEPrs endured their jobs and was said to be a retention
mechanism. They mentioned that training was an enticement that could help
them handle the urban community challenges they believed the job demanded
and qualify them to take on more advanced duties. In addition, participants
indicated that training was an indispensable motivational factor, as it nurtures
their profession.
On the other hand, program staff repeatedly mentioned the inadequacy of
“
opportunities for UHEPrs to advance their education.
The UHEPrs are disappointed and bored because they serve 13 years
without an education. They reach only level III or IV. I do not think
they should have to wait 13 years to get education; they could have
done their degree after they served 5 years. When you talk to them
about education now, they tell you, “We are old enough by now. How
we can learn? Let the youngsters learn.”
page-562
Despite the limited opportunities and demotivation of UHEPrs, participants

reasoned that the attrition rate was low because HEPrs have no other job
opportunities. Thus, it seems that it is not their interest that keeps them in the
“
UHEP program but rather a lack of other options.
Currently, the UHEPr is assumed to be a job for someone who

couldn’t otherwise get hired.
Participants described UHEPrs’ demotivation due to their high workload and

the hardship of working in the community. They stated that they were losing
“
hope because of the increasing burden of the job.
HEPrs have started to say that their job is no different from a military
job. They have started looking for another job. In the beginning, it
[the HEP] gave good job opportunities, and the salary was attractive,
but not anymore. Their salary is very low. It cannot meet basic needs.
Many have left the job after schooling in various health and non-
health colleges. They don’t have the prior momentum. I am afraid
that if we don’t act quickly, we may totally lose the program. Most
senior Health Extension Workers are leaving the job, and they are
not happy.
Program Officer, RHB
Program staffs mentioned that UHEPrs expressed feeling frustrated and

demotivated when they thought that their importance to the health system
was ignored and when they received negative feedback from their supervisors.
In addition, the absence of frequent supportive supervision was reported as
a source of dissatisfaction and demotivation. On the other hand, community
satisfaction and acceptability were repeatedly mentioned by UHEPrs as a
source of their satisfaction.
page- 563
Table 4-2. Level of satisfaction of UHEPrs by City Administration
City Administration
Addis Ababa Dire Dawa Other towns Total

N % N % N % N %
Leadership
Not satisfied 192 50.8 41 47.7 53 50.0 286 50.2
Satisfied 186 49.2 45 52.3 53 50.0 284 49.8
Promotion
Not satisfied 253 68.4 43 49.4 72 67.9 368 65.4
Satisfied 117 31.6 44 50.6 34 32.1 195 34.6
Autonomy
Not satisfied 95 25.3 14 16.1 25 23.6 134 23.6
Satisfied 281 74.7 73 83.9 81 76.4 435 76.4
Work environment and
cohesion
Not satisfied 59 15.9 14 16.1 20 19.6 93 16.6
Satisfied 311 84.1 73 83.9 82 80.4 466 83.4
Professional training
Not satisfied 236 68.2 28 32.2 74 74.0 338 63.4
Satisfied 110 31.8 59 67.8 26 26.0 195 36.6
Job security and salary
Not satisfied 358 94.7 62 71.3 84 80.0 504 88.4
Satisfied 20 5.3 25 28.7 21 20.0 66 11.6
Recognition at work
Not satisfied 113 30.4 20 23.0 43 42.2 176 31.4
Satisfied 259 69.6 67 77.0 59 57.8 385 68.6
Presence of alternative
employment opportunities
Not satisfied 103 27.5 16 19.0 53 53.5 172 30.8
Satisfied 272 72.5 68 81.0 46 46.5 386 69.2
Overall satisfaction
Not satisfied 159 41.7 18 20.7 58 52.3 235 40.6
Satisfied 222 58.3 69 79.3 53 47.7 344 59.4
page-564
Figure 4-2: General Satisfaction of UHEPrs Working in UHEP
Figure 4-3: UHEPrs’ Perceived Trends in their Level of Satisfaction
The distribution of UHEPrs’ overall satisfaction by selected background

characteristics shows that the level of satisfaction differs significantly by age
category and level of education. UHEPrs in the age category of 25 to 29 had
page- 565
a slightly lower probability of being satisfied compared to other age groups,

and UHEPrs with a level V (degree) education had a higher probability of
being satisfied compared to level IV UHEPrs (68.8% versus 57.4%). There was
no substantial difference, however, in the level of satisfaction by marital status,
CoC certification status, or work experience (Table 4-3).
Table 4-3. Overall Satisfaction as UHEPr by some important background

characteristics
Overall satisfaction
Not satisfied Satisfied Total
Characteristics P-value
N % N % N
City administration/town
Addis Ababa 159 41.7 222 58.3 381 < 0.001
Dire Dawa 18 20.7 69 79.3 87
Other towns 58 52.3 53 47.7 111
Total 235 40.6 344 59.4 579
Age category of UHEPrs (In years)
18-24 27 36.5 47 63.5 74 0.155
25-29 131 44.7 162 55.3 293
>=30 76 36.9 130 63.1 206
Total 234 40.8 339 59.2 573
Marital status category
Currently married 147 40.1 220 59.9 367 0.857
Never married 79 41.8 110 58.2 189
Other 9 45.0 11 55.0 20
Total 235 40.8 341 59.2 576
Educational status
Level IV (diploma) 202 42.6 272 57.4 474 0.039
Level V (degree) 30 31.3 66 68.8 96
Total 232 40.7 338 59.3 570
Work experience
<=2 years 38 39.2 59 60.8 97 0.695
3-5 years 64 39.8 97 60.2 161
>=6 years 75 35.7 135 64.3 210
Total 177 37.8 291 62.2 468
page-566
CoC for highest level of

training as UHEPr
Yes - verified 129 40.2 192 59.8 321 0.948
Yes - not verified 76 40.9 110 59.1 186
Not taken 30 42.3 41 57.7 71
Total 235 40.7 343 59.3 578
Currently studying
Yes 147 40.5 216 59.5 363 0.918
No 88 40.9 127 59.1 215
Total 235 40.7 343 59.3 578
Abbreviation: CoC, Certificate of Competency.
4.3 Intention to Leave and Attrition
A document review was performed for 648 UHEPrs in Addis Ababa, Dire
Dawa, and Harari. Nearly two thirds (60.2%) were appointed from 2016 to
2019, and the rest were appointed from 2010 to 2015. About 65.6% were born in
urban areas, and almost all (99.7%) were female. By certification, 41% of them
were diploma nurses. More than half (55.9%) were in the age range of 20 to
24 years. Almost half (48.6%) were single at the time of the survey, and 49.7%
had no biological children. 15.2% of UHEPrs had a history of administrative
reprimands, but only 5% had a history of recognition; half (50.7%) did not get
annual leave during their stay. 25.6% were not CoC certified.
Of the 648 UHEPrs, 176 left their work, an attrition rate of 21.15%, with the
highest attrition in Addis Ababa (38.48%) and the lowest in Harari 6.12%. Of
those who left their work, 37.9% simply disappeared from their jobs, 30.6%
resigned, 27.4% left their work due to a change in qualification, 3.4% left due
to transfer elsewhere, and 0.6% were dismissed. The earliest attrition occurred
in 2007, with an attrition of 66 UHEPrs per 10 000. Attrition increased steadily
until 2011, but began to peak in 2014, reaching its peak in 2018 (Figure 4-4).
page- 567
Figure 4-4. Attrition per 10,000 HEPrs since the Implementation of the HEP
from 2007 to 2018
The earliest and longest streak of attrition occurred in Dire Dawa, beginning
in the year 2007, with 588 per 10 000 UHEPrs. The shortest streak of attrition
was seen in the Harari region beginning in the year 2017, with 5000 per
10 000 UHEPrs (Figure 4-5).
Figure 4-5: Attrition per 10,000 since the Implementation of the HEP from
2007 to 2018
page-568
The median time to attrition was 4 years, with an inter-quartile range (IQR) of
[1.5 to 6.6 years], indicating that 50% of UHEPrs served 4 years before leaving
their work. The overall median time of service, however, was 4.7 years [IQR=2.2
– 8.6]. The overall incidence rate of attrition was found to be 517.8 per 10 000
person-years [95% CI, 446.3 -600.7]. The highest incidence was seen during the
fifth to tenth years of implementation: 786.8 per 10 000 person-years [95% CI
626.5- 988.2]. Attrition is high among the age group of 25-29 and 20-24 HEPrs
in respective manner. Single UHEPrs, UHEPrs from urban areas, UHEPrs with
diploma and above certification, and UHEPrs who have no children left their
jobs more common than their counterparts (Table 4-4).
The qualitative finding depicts 2 major themes as causes of attrition: personal

causes, such as a lack of educational opportunities or personal conflicts, and
work-related causes, which include the denial of annual leave, unfair wages,
judgmental appraisal, and poor support habits. Addis Ababa has the largest
attrition than Dire dawa and Harari.
“
They always think of us [UHEPrs] as mischievous and inferior, and
they do not understand us at all. In addition to that, when training
opportunities come…they would not send us, but they demand that
we take up the task. I do not know if other UHEPrs [the older ones]
may have lots of trainings, but I was new, and I need those trainings,
but they do not give me the chance. To my surprise, they asked me
to bring a report just a week after my employment. Without showing
me anything....How could it be possible? Preparing a monthly report
without knowing any of the WDAs is very difficult. And there was a
problem in the inductive introduction when I was there.
UHEPr, Addis Ababa
page- 569
Table 4-4. Magnitude of attrition by different background variables
Attrition rate Unweighted

Variable Category
(weighted %) frequency
Harari 6.12 49
City Addis Ababa 38.48 408
Dire Dawa 8.38 191
18-19 12.5 55
20-24 22.8 331
Age
25-29 25.25 184
>30 13.8 78
Urban 27.03 527
Birthplace
Rural 10 121
Single 22.48 479
Married 17.88 154
Marital status at deployment
Divorce/separated/
0 15
widowed
0 23.8 345
Biological children
1+ children 17.44 292
Level III 8.18 96
Level IV 6.6 95
Current certification
Diploma nurse or midwife 28.03 353
Degree 50.26 104
No recognition 20.5 606
Recognition
1+ recognition 32.4 42
<7 years 24.6 438
Experience in HEP
>7 years 12.8 206
No reprimand 22.2 570
Punishment
At least 1 reprimand 15.09 78
No annual leave 17.4 316
Annual leave 1-3 annual leaves 18.3 181
4+ annual leaves 34.5 151
CoC-certified 21.6 474
CoC status
Not certified 19.8 174
page-570
The quantitative findings showed that, of 573 UHEPrs, 28.6% were currently
looking for another job. The main reasons to look for alternative job opportunities
were low salary (82%), the absence of career development (64.4%), and the
absence of motivational schemes (60.7%). There was not significant variation
among cities (Figure 4-6 and Figure 4-7).
Figure 4-6: Percentage of UHEPrs Currently Looking for Another Job

The qualitative findings from Addis Ababa, Harari, Dire Dawa, and other
towns in Ethiopia are widely supportive of the above findings. Participants
frequently mentioned high staff turnover as one of the major challenges that
affected the implementation of the program. As explained by one expert from
an RHB, the selection process and preference of UHEPrs to work as clinicians
“
contribute to higher staff turnover.
The main problem related to the Urban Extension Program is in

the selection process of UHEPrs. Clinical nurse professionals were
trained for 3 months and assigned as UHEPrs. They preferred to stay
at the office rather than going home-to-home and performing their
activities.
page- 571
Figure 4-7: Major Reasons for Intention To Leave Job

The study investigated the intended attrition of UHEPrs in an urban setting, and
the majority of respondents responded that the main reason for attrition was the
work burden. They perceived the workload is a result of inadequate numbers of
HEPrs. Thus, they are supposed to perform different activities simultaneously.
Participants also described work conditions that involved walking long hours to
reach all households, executing the program, ending up exhausted, and looking
elsewhere for other job opportunities. The intimidating behavior of other health
professionals was also perceived as another reason for UHEPrs’ intention to
leave their jobs, and, in spite of their hard work, incentives were being given to
nurses and others.
“
There is no attention, as I mentioned, to the high workload.
Everybody complains about our achievement; we are considered
farmers [description of tiresome unskilled labor], but we graduated
as diploma nurses. That is why I wanted to leave this work.
Their acceptance and recognition by the community were the reasons most
frequently mentioned by some of the UHEPrs and their supervisors as reasons
to continue working with the UHEP.
page-572
“
Even if the pay isn’t plentiful compared to the job you do, it’s still your
monthly income. Gaining acceptance from the community initiates
you. This acceptance makes you want to stay longer on the job.
UHEP supervisor, SNNPR
UHEPrs resign because they perceive that there would be a higher salary
elsewhere, perhaps in non-health-related fields. Moreover, the lack of recognition
for their performance is also a reason for high attrition, as it limits their career
growth possibilities, and there would be no professional advancement in the
field in which they are working. Participants also remarked that, even though
they ask to advance in their profession, the civil service requests CoC approval,
which is challenging for them to meet. By contrast, some participants stated
that clinical nurse professionals were trained for 3 months and assigned as
UHEPrs.
UHEPrs perceived a shortage of in-service training, which might be specific to

any packages needed presently. UHEPrs did, however, respond positively to
the fact that training lays the path for professional advancement, and working
longer as a HEPr when it is provided at their workplaces reduce the need for
them to resign from their jobs. Respondents also stated that UHEPrs might
leave their jobs if their residence is far from their workplace.
“
The former UHEPrs were competent enough, because at that time
there was enough in-service training. But now they resign due to low
incentives for overtime work. The professional background of Urban
Health Workers is as nurses, so they upgraded their professions and
looked for other options with better incentives. Currently, almost
all of the former UHEPrs have resigned and new Health Extension
Workers have been employed.
UHEP-supervisor, Aboker, HA
Despite the limited opportunities for and demotivation of UHEPrs, the

participants reported that the attrition rate was low because UHEPrs had no
other job opportunities. Thus, it seems that they continue to work for the UHEP
not out of interest but out a lack of other options.
page- 573
“
Currently, the UHEPr is assumed to be a job for someone who
couldn’t get hired.
The vast majority of UHEPrs also do not recommend other people who are
qualified for the position to work as UHEPrs (Figure 4-8).
Figure 4-8: UHEPrs’ Tendency to Recommend Others to Work as UHEPrs
4.4 Degree of Burnout among UHEPrs

Burnout was assessed by an instrument containing 15 items, with a scale ranging
from 1, “rarely”, 2, “sometimes”, 3, “often” and 4, “very often.” Risk of burnout was
analyzed by categorizing the scores between 15 to 18 as “no risk of burnout,” 19
to 22 as “little sign of burnout,” 33 to 49 as “risk of burnout,” and 50 to 79 as
“severe risk of burnout.”
In the current assessment, the majority (77%) of UHEPrs had either little sign
of burnout or were at risk of burnout. A small but meaningful percentage of
UHEPrs (4.5%) had a severe risk of burnout. A slightly higher proportion of
level V UHEPrs showed little sign of burnout, risk of burnout, or severe risk
of burnout compared to level IV UHEPrs. There was no significant difference
page-574
in the level of burnout across city administrations. Similarly, burnout was not
significantly associated with UHEPrs’ age, marital status, work experience, or
CoC certification (Table 4-5).
Table 4-5. Level of Burnout among Urban Health Extension Professionals
UHEPr’s level of burnout

Severe
Little sign Risk of
Characteristics No burnout risk of
of burnout burnout P-value
burnout
N % N % N % N %
City administration
Addis Ababa 75 19.7 185 48.6 104 27.3 17 4.5 0.444
Dire Dawa 18 20.7 39 44.8 27 31.0 3 3.4
Other towns 15 14.4 44 42.3 40 38.5 5 4.8
Total 108 19 268 47 171 30 25 4
Age of UHEPrs
18-24 15 20.8 35 48.6 22 30.6 0 0.0 0.142
25-29 46 15.9 138 47.8 86 29.8 19 6.6
>=30 44 21.5 94 45.9 61 29.8 6 2.9
Marital status
Currently married 69 19.1 171 47.4 106 29.4 15 4.2 0.869
Never married 34 18.1 88 46.8 56 29.8 10 5.3
Other 3 15.8 8 42.1 8 42.1 0 0.0
Educational status
Level IV (diploma) 93 19.9 216 46.3 142 30.4 16 3.4 0.031
Level V (degree) 12 12.5 47 49.0 28 29.2 9 9.4
Experience as UHEPrs
<=2 years 17 17.5 47 48.5 28 28.9 5 5.2 0.828
3-5 years 37 23.0 71 44.1 45 28.0 8 5.0
>=6 years 39 18.6 106 50.5 58 27.6 7 3.3
CoC for highest level of training as UHEPr
Yes (verified) 57 17.9 151 47.5 90 28.3 20 6.3 0.270
Yes (not verified) 37 20.2 82 44.8 60 32.8 4 2.2
No 14 20.0 35 50.0 20 28.6 1 1.4
Currently on study
Yes 67 18.8 176 49.3 98 27.5 16 4.5 0.343
No 41 19.2 91 42.5 73 34.1 9 4.2

page- 575
Dissatisfaction and burnout are usually associated with turnover, low morale,
poor quality of care, low productivity, absenteeism, and compromised social
interactions due to poor interpersonal relationships. Various studies have
indicated that burnout can lead to insomnia, perceptions of physical exhaustion,
increased substance abuse, and ultimately mental health problems. The effects
of having staff with a high level of burnout and dissatisfaction goes beyond the
personal; they may affect the reputation and acceptability of the program and
the health system in general.
4.5Mental health status of UHEPrs

The study evaluated the mental health status of UHEPrs using a standard tool
called a patient health questionnaire (PHQ-9). The tool contains 9 questions
that measure one’s level of depression on scales, ranging from 0 (notat all)
to 3 (nearly every day). According to the reliability test result, the instru-
ment has an acceptable consistency, with a Cronbach’s alpha value of0.86.
The mental health status was measured against a total of 27 points,
where 10 is the cut-off point. Accordingly, individuals scoring 10 or more points
were categorized as having probable symptoms of depression.
The findings of the current study showed that 8.9% of UHEPrs had probable
symptoms of depression. A slightly higher percentage of UHEPrs from other
towns had probable symptoms of depression compared to Addis Ababa and
Dire Dawa. Compared to other age categories, UHEPrs aged between 25 and
29 had relatively higher symptoms of depression. Similarly, UHEPrs with an
educational status of Level V (degree) had a slightly higher risk of probable
depression (12.8%) compared to those with a Level IV (diploma) educational
status (8.0%) (Table 4-6).
page-576
Table 4-6. Mental Health Status of UHEPr
Depression status of UHEPrs
Probable symptoms
No depression
of depression
N % N %
City Administration
Addis Ababa 350 91.9 31 8.1
Dire Dawa 81 93.1 6 6.9
Other towns 83 86.5 13 13.5
Total 514 91.1 50 8.9
Age of UHEPrs
18-24 65 95.6 3 4.4
25-29 256 88.9 32 11.1
>=30 189 93.6 13 6.4
Marital status
Currently married 328 92.4 27 7.6
Never married 165 89.2 20 10.8
Other 18 90.0 2 10.0
Educational status
Level IV (diploma) 425 92.0 37 8.0
Level V (degree) 82 87.2 12 12.8
Work experience as UHEPrs
<=2 years 91 93.8 6 6.2
3-5 years 147 91.3 14 8.7
>=6 years 193 91.9 17 8.1
CoC for highest level of training as a UHEPr
Yes (verified) 284 90.4 30 9.6
Yes (not verified) 166 91.7 15 8.3
No 63 92.6 5 7.4
Status of current study
Yes 325 91.5 30 8.5
No 188 90.4 20 9.6
page- 577
CHAPTER 5
Findings:
Implementation
of the UHEP
page-578
CHAPTER 5
5 Findings: Implementation of the
UHEP
As clearly described in the UHEP implementation manual, the program has 4

major packages and 15 sub-packages, which are intended to improve access
to health services, mainly by improving the hygiene and sanitation conditions
of urban residents, maternal and child health, and youth and adolescent
reproductive health through health promotion, disease prevention, accident
prevention, first aid, referral, and linkage. The UHEP packages and services are
delivered in households, school, youth centers and workplaces. Major service
delivery strategies include home-to-home visits, community mobilization during
the sanitation and immunization campaign; training of model households,
teachers, students, and youth, and collaboration with formal and informal
community structures (WDAs, religious leaders, and civic associations).
This study assessed the implementation of the UHEP using qualitative and
quantitative approaches. This section describes the perceived and actual levels
of UHEP implementation from providers and beneficiaries’ (i.e., households’)
perspectives. The results of the study highlight the extent to which the UHEP
packages and services have reached the urban households, households’ and
UHEPrs’ perception about the implementation of the packages in the area,
and the judgement of the UHEPrs regarding the level of UHEP implementation
among the service delivery modalities. Moreover, the facilitators and challenges
of the program are described in detail. The level of community access to
the UHEP was assessed using the knowledge, attitudes, and practices of
households related to program packages, services, behavioral changes, and
health and health-related outcomes that were expected to be improved due to
the implementation of the UHEP.
page- 579
Findings: Implementation of the UHEP
5.1 Home-to-Home Visits by UHEPrs
In the case of the UHEP, most of the packages and services are designed
to be rendered by providing health education or information, counseling,
screening for CDs and NCDs, and follow-ups (pregnant mothers, under-5
children, chronic patients, and model households) through home-to-home visits.
As per the UHEP implementation guidelines, UHEPrs are expected to spend
significant working time (3 days per week) to implement the UHEP through
home-to-home visits to target households (economically poor and households
with under-5 children, pregnant and lactating mothers, and individuals with
NCDs and CDs). The UHEPrs are also expected to reach every household
assigned under their catchment, at least annually, regardless of their health
needs or economic status.
This study’s findings revealed that about 36% of households included in the
survey had never been visited by UHEPrs, and that more than 18% of households
were visited more than a year prior. This result shows that more than half of
the households included in this study were not contacted or visited by UHEPrs
on the recommended annual basis. There is no significant variation in the level
of household visits by UHEPrs among the respondents of Addis Ababa or Dire
Dawa (Figure 5-1).
Among a total of 1110 respondents who were asked about the most recent time
their house was visited by UHEPrs, about 71% reported that they had received
at least 1 visit from UHEPrs 6 months prior to the survey. Among households
visited in the past year, 15.8% and 22.7% were visited only once or twice per
year, respectively (Table 5-1).
In general, the results of quantitative survey indicated less frequent household

visits by UHEPrs, below the minimum standard set in the implementation
manual.
page-580
Figure 5-1: Performance of UHEPrs in Conducting Routine Household Visits
Table 5-1. Frequency of Household Visit by UHEPrs
City administration
Status of Household Visit Addis Ababa Dire Dawa Total
by UHEPrs N % N % N %
Time of most recent visit
(n=1110)
Within a week 136 18.8 36 9.3 172 15.5
Within a month 168 23.2 51 13.2 219 19.7
3 months earlier 128 17.7 70 18.1 198 17.8
6 months earlier 96 13.3 109 28.2 205 18.5
A year earlier 196 27.1 120 31.1 316 28.5
Frequency of household visits
(n=747)
Once a year 57 11.5 61 24.4 118 15.8
Twice a year 90 18.1 80 32.0 170 22.7
3 times a year 75 15.1 42 16.8 117 15.7
4 times a year 57 11.5 24 9.6 81 10.8
5 times a year 44 8.9 13 5.2 57 7.6
6 times a year 44 8.9 10 4.0 54 7.2
7+ times a year 130 26.2 20 8.0 150 20.1
page- 581
The findings from the qualitative study revealed that UHEPrs provide a wide
range of services for community members during home visits. Program staff
indicated that home-to-home services by UHEPrs were not limited to specific
days or times. They are also accessible through phone calls. Regarding the
frequency of visits, respondents indicated that home-to-home visits were held on
a monthly basis. There are also groups who indicated that specific households
identified for a follow-up visit would be visited once or twice a week, but under-5
children and mothers were visited daily.
“
They do their activities moving home to home. They do not have rest,
even on the weekends. When they are required by the community
members even at midnight, they go to the one who demanded their
service.
UHEP supervisor, Tigray
Including the nutritional assessments, UHEPrs provide screening services for

HTN and diabetic cases during home visits. UHEPrs weigh children and chart
their growth when they make home-to-home visits. If children are identified as
malnourished, they provide them with supplementary foods or refer them to
HCs. Similarly, UHEPrs provide nutritional support for pregnant mothers after
their home visit inspections. They advise pregnant women to visit HPs for ANC
services, including iron supplementation and vaccinations, when they make
home visits. They also provide food-preparation lessons through demonstrations.
“
UHEPrs also show mothers how to breastfeed.
There is nothing we can say the UHEPrs lack. They even go home to
home to give lessons for mothers on how to breastfeed. When we tell
them that a woman is pregnant in our area, they go to their homes
and visit them until the child reaches 6 or 7 months old.
WDA member, Assosa, Benishangul-Gumuz
UHEPrs provide health education on HEP packages and provide home-to-

home services for CDs and NCDs. Some committed UHEPrs engage in a
home-to-home TB DOT program together with HC staff. As the respondents
indicated, UHEPrs focus on changing awareness and also verify whether the
community has implemented the packages by visiting households.
page-582
During their home visits, UHEPrs provide training for groups of women and ask
about any problems the community faces. They also provide first aid services
and treat children with pneumonia and diarrhea. During epidemics, UHEPrs
go home-to-home for screenings. UHEPrs also ensure that ITNs are distributed
to all homes.
UHEPrs provide treatment and hygiene services and facilitate free treatment
at HCs for elders who cannot afford the service. UHEPrs with FHTs care for
bedridden patients. WDAs also accompany UHEPrs during home visits and
educate household members about the packages.
UHEPrs collaborate with FHTs and provide rehabilitation services for people
with mobility problems. In the same way, they advise people seeking mental
health services and refer them to facilities. On the other hand, some respondents
indicated that UHEPrs were not giving the appropriate home visit services at
the household level for different reasons. According to the respondents, due to
the large population size, UHEPrs cannot address all households. They spend
at least 2 hours per household, and their engagement in non-health areas,
like meetings at the woreda level and their lack of skill and equipment restrict
UHEPrs from providing the desired home-to-home services as intended.
“
...The other thing is that we are occupied with different meetings
arranged by the woreda health office. For these reasons, we cannot
conduct regular home-to-home visits.
UHEPr, Adwa, Tigray
page- 583
5.2 Model Family and Women’s/Health Development

Army
As described in the previous section, model families and WDAs are the major
support networks through which the UHEP is implemented at the community
level. Model family training is designed to create role models who can educate
other non-model households and influence them through positive attitudes and
skills toward healthy behaviors. To be considered a model, a member of the
household should receive theoretical and practical training on the 15 UHEP
sub-packages and should implement at least 75% of the packages that apply
to their household.
Participants in the qualitative study stated that the number of model families
was considered one of the efficient strategies for implementing UHEP
packages through the diffusion of knowledge and skills from model to non-
model households in the HEP. In this study, the criteria for a household to be
considered a model were assessed. These criteria included properly performing
all the HEP packages, like building separate kitchens, properly disposing of
wastes, maintaining environmental hygiene, learning and teaching others, and
preparing food and using latrines properly. Model households were reported to
“
be awarded a certificate or gift.
I am a model family. We are given this title because we translated

what we learned to our lives …. Model families were chosen based on
health behavior related criteria, and they were given handwashing
jugs and plate as a gift to motivate them. The model is the one who
applies in his or her life what is learned from the UHEPr. Women who
always attend a meeting are also called model.
WDA leader, Bahir Dar, Amhara
The UHEPrs organize an event wherein model families can share their experiences
with the community. Model households also share their experience with the
community through home-to-home visits and demonstrate how to maintain
cleaning, prepare handwashing materials, and clean their surroundings. They
are involved in awareness-creation activities and teach other members of the
community.
page-584
“
There are other groups under us in the 1-to-5 groups. We are also
making them models like us. For example, discussing SRH issues,
communicable and non-communicable diseases, for example, we
teach them about signs and symptoms and ways of detecting diseases
like cervical and breast cancer. We also tell them to seek medical
treatment at the Health Centers, and we also follow that. We also
train and show practically waste disposal, toilet use, child nutrition,
and a balanced diet.
“
WDA leader, Assosa, Benishangul Gumuz
She brings models as an example for us, to learn from them how they
keep their environment clean, how they use their toilets properly, how
they maintain their personal hygiene, and how to handle waste.
WDA leader, Hawassa, SNNP
The communities are inspired by, learn from, and follow in the footsteps of
the model families. The participants reported that there was punishment when
model families failed to follow what was expected of them. Some participants
reported that there were some model families who do not properly dispose of
“
wastes.
They try to see what the model family does, and they do the same. If
they are sending their children to school, the others also try to do the
same. They learn from each other. Yes, he will learn, for example, if
he works, what I worked like. If he sees my doorstep clean and putting
my waste separately and sees my house cleaned well, he will get
motivated to do the same, I hope. So, he will learn like why couldn’t
we do this before. It was spilling everywhere, but now we put it in a
bag, and if there is no safety net, we pay 5 or 10 birr to dispose of it.
WDA leader, Aboker, Harar
page- 585
The UHEPrs reported that they had used model families to reach as many
people as possible. The use of volunteer women is one strategy used in the
creation of model families. Volunteer women are trained and become model
women in the community. These model women, in turn, train other women in
“
their communities, as explained by a UHEPr from Logia:
We recruited willing women who have a say from the community,

give them a 3-month-long training, and when they graduate, they
get the title “model women.” In turn, they have to go and teach their
communities and be their role models. If in case the Health Extension
Service is somehow interrupted, they ask questions. They demand the
availability of the Health Extension Workers.
This study assessed the extent of model family implementation among the
participants in the survey. As depicted in the graph below, model family
training was provided for only 18.5% of households included in the survey. This
figure is relatively low in Dire Dawa (8.7%). More than 25% of respondents
were unaware of the model family program. Participants were also asked their
reason for not doing the training. Among the 967 households who had no
model family training, a majority (58%) were not asked to do the training, 15%
had no access to the training, and 13% had busy schedules (Figure 5-2).
Figure 5-2. Households’ Model Family Training Status
page-586
Figure 5-3. Households` Reasons for Not Taking Model Family Training
5.3 Family Health Team Approach

5.3.1 FHT Establishment and composition
According to the Ethiopian urban primary health care reform implementation

(PHC) guideline, the number of FHTs per HC is determined by the size of the
catchment population. The number of staff available at the HCs determines
the composition of each FHT. Each FHT, however, is expected to employ 5-10
health professionals. In the qualitative study, participants reported that each
FHT was composed of multi-disciplinary health workers. The participants
reported that each FHT was supposed to encompass physicians or health
officers, midwives, nurses, and UHEPrs. The team composition was, however,
reported to be affected by the shortage of health service providers.
In the qualitative findings, participants reported that they believed the

implementation of the FHT approach had improved access to basic health
services among the less advantaged segments of the urban population.
page- 587
“
Mostly, there were people who had no one to care about or where
the Health Extension Worker was unable to take them to the Health
Center to be treated. So when they are becoming integrated into
the community, there is a patient carer with then so that they can
diagnose the case and treat it right there. If necessary, the lab
professional on call will come to collect samples, and appropriate
treatment will be offered.
5.3.2 Involvement of the UHEPrs in the implementation of the FHT
This section describes different forms of UHEPr involvement in the implementation

of community-based primary health care services through the FHT approach.
The Ethiopian urban PHC guidelines indicate that UHEPrs are responsible for
collecting a baseline population profile and timely updates, conducting home
visits, identifying households that need close follow-up, and arranging referrals
when needed. As members of the FHT, the UHEPrs are also responsible
for identifying and providing health services at schools, youth centers, and
workplaces, and on the streets.
Overall, about three fourths (75.3%) of the UHEPrs reported that they
participated in the conduction of baseline assessment to collect a population
profile. This figure was relatively higher in Dire Dawa (95.8%) than Addis Ababa
(73.9%). The majority (95.3%) of UHEPrs reported identifying and scheduling
households for FHT services as one of their roles and responsibilities in the
implementation of community-based primary healthcare services through the
FHT approach. In general, most (79.8%) UHEPrs mentioned the identification
and planning of activities to be conducted at schools and youth centers as
two of their roles and responsibilities. This figure is relatively lower in Addis
Ababa, where only 58.3% of the participants reported participating in the
identification and planning of activities conducted at schools and youth centers.
A comparable number of UHEPrs from Addis Ababa (60.2%) and Dire Dawa
(58.3%) reported that they were involved in the identification of the homeless
community. Compared to Dire Dawa (62.5%), a higher number (89.9%) of
UHEPrs from Addis Ababa reported recording activities and performances of
page-588
the FTH approach. 82.4% and 54.2% of the UHEPrs from Addis Ababa and
Dire Dawa, respectively, reported participating in the preparation of a weekly
FHT performance report. On the other hand, 70.6% of UHEPrs (69.7% in Addis
Ababa and 87.5% in Dire Dawa) reported that they were involved in updating
the catchment population profile (Table 5-2).
Similarly, in the qualitative findings, participants reported that the activities
“
implemented under the FHT model were documented and evaluated weekly.
Every Friday, we sit to evaluate what team members had done in

the presence of their focal person and identify the gaps we had. So I
believe we did better than before.
Table 5-2: Preparation, Planning and Monitoring of FHT Activities
Addis Ababa Dire Dawa Total

N % N % N %
Conduct a baseline assessment to collect a population profile data
Yes 264 73.9% 23 95.8% 287 75.3%

No 93 26.1% 1 4.2% 94 24.7%
Identify and schedule households for FHT visit or services (weekly planning)
Yes 340 95.2% 23 95.8% 363 95.3%

No 17 4.8% 1 4.2% 18 4.7%
Identify and plan activities to be conducted at schools and youth centers
Yes 290 81.2% 14 58.3% 304 79.8%

No 67 18.8% 10 41.7% 77 20.2%
Identify the health needs of the homeless community
Yes 215 60.2% 14 58.3% 229 60.1%
No 142 39.8% 10 41.7% 152 39.9%
page- 589
Record activities and performances of the FHT

Yes 321 89.9% 15 62.5% 336 88.2%
No 36 10.1% 9 37.5% 45 11.8%
Prepare the weekly FHT performance report
Yes 294 82.4% 13 54.2% 307 80.6%
No 63 17.6% 11 45.8% 74 19.4%
Update the catchment population profile
Yes 248 69.5% 21 87.5% 269 70.6%
No 109 30.5% 3 12.5% 112 29.4%
Abbreviation: FHT, Family Health Team.
5.3.3 Establishment of baseline community profile
Only 60.9% of UHEPrs reported that their respective FHT had collected a
population profile at baseline. The collection of a baseline population was
relatively better in Dire Dawa (80.3%) than Addis Ababa (59.4%). Of the
232 UHEPrs who reported that their respective FTHs had collected a baseline
population profile, 88.4% mentioned that they had categorized the population
according to economic status. Three fourths (75%) of UHEPrs reported that
they updated the catchment population profile of their respective FHTs, the
majority (58.6%) of whom made updates to the population profile 3 times per
year (Table 5-3).
The qualitative findings also indicate a limited practice of the collection of a

baseline population profile. Participants said that the FHT did not accommodate
the heterogeneity of the living conditions of the urban population. The lack of
categorization of the population was reported as having contributed to less
effective implementation of the FHT approach.
page-590
Table 5-3: Establishment of baseline community profile and regular updating

practices by the Family Health Team by City Administration

N % N % N %
FHT collected a population profile at baseline
Yes 212 59.4 20 83.3 232 60.9
No 145 40.6 4 16.7 149 39.1
Categorization of households based on economic status and health needs
Yes 185 87.3 20 100.0 205 88.4

No 27 12.7 0 0.0 27 11.6
FHT updated the catchment population profile
Yes 158 74.5 16 80.0 174 75.0
No 54 25.5 4 20.0 58 25.0
5.3.4 Services provided by the FHT
The types of services that were reported by the UHEPrs to be provided by

their respective FHTs are presented in Table 5-4. In this study, 98.7%, 96.1%,
and 82.2% of the UHEPrs reported that their respective FHTs provided health
education and counseling, community mobilization for environmental sanitation,
and FP services, respectively. Similarly, a majority of UHEPrs reported that
their respective FHTs provided ANC services (78.8%), PNC services (79.5%),
child health services (89.5%), TB screening and referral (93.7%), psychological
support for patients with mental illness (89.2%), and first aid services (83.7%).
Screening and referral of cases of chronic diseases were among the services
FHTs provided that were most frequently mentioned. Only 45.4% of the UHEPrs,
however, reported that their respective FHTs provided malaria diagnosis and
treatment.
page- 591
Likewise, the qualitative findings indicate that disease prevention and control,
health education, screening, mental health services, hygiene and sanitation,
and referral and linkage are widely provided by the FHTs. The qualitative study
participants reported that the FHTs provided education on reproductive health
as well as sanitation and hygiene. Moreover, palliative care of bedridden patients
was among the services reported to be widely provided by the FHTs. Similarly,
screenings for malnutrition, visual and hearing impairments, hypertension, and
diabetes mellitus were among the most frequently mentioned UHEP packages
implemented through the FHT approach.
Table 5-4: Distribution of services provided through the Family Health Team
by City Administration
City administration
Dire Dawa Total (Yes)
Addis Ababa
Services provided by FHT (Yes)
(Yes)
N % N % N %
Health education and counseling service 356 99.7% 20 83.3% 376 98.7%
Community mobilization for environmental
347 97.2% 19 79.2% 366 96.1%
sanitation
Family planning/contraceptive 293 82.1% 20 83.3% 313 82.2%
Antenatal care service 280 78.4% 20 83.3% 300 78.7%
Post-natal care service 283 79.3% 20 83.3% 303 79.5%
HIV testing and counseling 263 73.7% 20 83.3% 283 74.3%
Child immunization 240 67.2% 18 75.0% 258 67.7%
Child health services (nutrition screening,
supplementation of micronutrients, and 322 90.2% 19 79.2% 341 89.5%
deworming)
Diagnosis and treatment of children with
278 77.9% 20 83.3% 298 78.2%
diarrhea
Diagnosis and treatment of children with
267 74.8% 20 83.3% 287 75.3%
pneumonia
Malaria diagnosis and treatment 154 43.1% 19 79.2% 173 45.4%
TB screening and referral 338 94.7% 19 79.2% 357 93.7%
NCD screening/identification and referral

339 95.0% 20 83.3% 359 94.2%
(diabetics, hypertension and heart disease
page-592
Cancer screening and referral, including

311 87.1% 20 83.3% 331 86.9%
self-breast examination for breast cancer
Provide psychological support for patients
321 89.9% 19 79.2% 340 89.2%
with mental illness
Palliative care service for bedridden
318 89.1% 19 79.2% 337 88.5%
patients
Wound care service 259 72.5% 20 83.3% 279 73.2%
First aid service 299 83.8% 20 83.3% 319 83.7%
Referral and linkage for health and
308 86.3% 20 83.3% 328 86.1%
psychosocial support
Abbreviations: FHT, Family Health Team; TB, tuberculosis; NCD, non-communiable disease.
5.3.5 FHT service delivery points
Table 5-5 indicates the different points at which the FHTs provide health
services to the community. Almost all (99.5%) of the UHEPrs reported that the
FHTs used home visits to provide services, and 80.8% reported that they used
schools; these were the most widely reported outlets. Only 55.4% of respondents
mentioned that FHTs provided services at workplaces. More importantly, FHT
health provision is limited at youth centers and to the homeless community.
Similarly, the qualitative findings indicate that the FHT provides health services
at different outlets, including schools, households, workplaces, and youth centers.
Moreover, the FHT provides services to homeless people and children who
have no support. Use of several service-delivery approaches was reported to
have benefited the economically disadvantaged segments of the population,
homeless individuals, people with chronic diseases, bedridden patients, and
elders.
page- 593
“
In schools, we have implemented eye check-ups for visual acuity with
the Health Extensions and Family Health. Of the checked individuals,
about 146 came to a Health Center, and we referred them to Ras
Desta eye specialists. We did not have these opportunities before. For
example, a sick person might stay home for a long period of time and
nobody would inspect that. But now the Health Extension Worker
goes house to house and reports cases like these to Family Health,
and Family Health gives treatment and brings them to hospital if it
is beyond their capacity.
FHT coordinator, Addis Ababa
Table 5-5: Distribution of delivery points used by the Family Health Team to
provide services
City Administrations
Total
Addis Ababa Dire Dawa
Service delivery points for FHT
N % N % N %
Household 355 99.4 24 100.0 379 99.5
School 297 83.2 11 45.8 308 80.8
Youth center 123 34.5 6 25.0 129 33.9
Workplace 200 56.0 11 45.8 211 55.4
Street (homeless community) 175 49.0 18 75.0 193 50.7
5.3.6 Planning, supervision, and evaluation practice of the FHT
The majority (96.6%) of UHEPrs reported that their respective FHTs used a
standard recording tool to capture data for the outreach services they provided
to the community. The daily activity registration format was the single most
widely used type of recording tool used by FHTs. Most (85.8%) UHEPrs reported
that their FHT had a weekly plan. Similarly, 85.6% of the respondents mentioned
that their FHT had a regular meeting with their FHT coordinator, and 84.4%
stated that their FHT met with the team coordinator yearly. Most (77.7%)
UHEPrs stated that their HC performance monitoring team (PMT) conducted
a performance review of their FHT’s community (outreach) activities. Similarly,
page-594
63.0% of participants stated that their respective FHT had received supportive
supervision in the last 6 months. Only 32.3%, however, of the UHEPrs reported
that HCs conducted supportive supervision, and only and 37% reported that
their FHT had its own kit (Table 5-6).
Similarly, the participants in the qualitative study said that performance appraisal
was conducted every week. The participants believed that this practice was of
paramount importance to the timely identification of the implementation gaps
and the ability to take prompt measures in response. Moreover, the participants
argued that there had been strong supportive supervision, monitoring, and
evaluation of the implementation of the UHEP activity. They reported that
the use of the findings from routine monitoring and evaluation had positively
contributed to the implementation of the FHT approach.
page- 595
Table 5-6: Planning, monitoring and evaluation practice of the Family Health
Team, UHEP assessment

N % N % N %
Use of standard recording tools
Yes 348 97.5% 20 83.3% 368 96.6%
No 9 2.5% 4 16.7% 13 3.4%
Recording tools used by FHT
CHIS cards and registers 54 15.1% 4 16.7% 58 15.2%
Daily activity registration forms 231 64.7% 18 75.0% 249 65.4%
Referral slips 68 19.0% 1 4.2% 69 18.1%
Others 4 1.1% 1 4.2% 5 1.3%
FHT has a weekly plan
Yes 311 87.1% 16 66.7% 327 85.8%
No 46 12.9% 8 33.3% 54 14.2%
Regular meeting scheduled with FHT coordinators
Yes 309 86.6% 17 70.8% 326 85.6%
No 48 13.4% 7 29.2% 55 14.4%
PMT of the HC review of FHT performance
Yes 280 78.4% 16 66.7% 296 77.7%
No 58 16.2% 6 25.0% 64 16.8%
PMT is not available 19 5.3% 2 8.3% 21 5.5%
FHT received supportive supervision in the last 6 months
Yes 230 64.4% 10 41.7% 240 63.0%
No 127 35.6% 14 58.3% 141 37.0%
Responsibility for conducting supervision
HC head/deputy/process owners 178 77.4% 6 60.0% 184 76.7%
WorHO 19 8.3% 4 40.0% 23 9.6%
Sub-city health office 30 13.0% 0 0.0% 30 12.5%
City administration Health Bureau 3 1.3% 0 0.0% 3 1.3%
HCs conduct supportive supervision
Yes, regularly 117 32.8% 6 25.0% 123 32.3%
Yes, not regular 169 47.3% 8 33.3% 177 46.5%
No 71 19.9% 10 41.7% 81 21.3%
FHT has its own kit
Yes 124 34.7% 17 70.8% 141 37.0%
Yes, partially 182 51.0% 5 20.8% 187 49.1%
No 51 14.3% 2 8.3% 53 13.9%
Abbreviations: FHT, Family Health Team; CHIS, community health information system; HC, Health
Center; WorHO, Woreda Health Office; PMT, Performance monitoring team
page-596
5.3.7 Availability of drugs and supplies needed for FHT activities
UHEPrs were asked about the availability of drugs, equipment, and supplies
needed for the FHTs to provide the service packages. In the quantitative study,
most participants reported that they have blood pressure apparatus, MUAC
tape, and vitamin A. A significant number of the UHEPrs, however, reported
that most of the drugs, equipment, and supplies needed for FHTs were not
available (Table 5-7).
The qualitative findings also show the inadequacy of the inputs and budget
needed to provide community-based health services through the FHT approach.
The inadequacy of drugs and supplies were reported as obstacles to the
“
implementation of the FHT approach. One participant describes the situation
thus:
We are going to the community having inadequate resources. For

example, after measuring blood pressure, we have nothing to do if
the person has elevated BP, because we do not have a medication
when we go out for the services. But if we were treating the person
right there, the possible complications would be prevented until the
client reached the health facilities.
UHEPr, Gulele Sub-City, Addis Ababa
page- 597
Table 5-7. Availability of basic equipment for Family Health Team activities
Availability of Pharmaceuticals Addis Ababa Dire Dawa Total

and Supplies N % N % N %
Functional BP apparatus 312 87.4 24 100 336 88.2
Thermometer 268 75.1 24 100 292 76.6
MUAC tape 337 94.4 23 95.8 360 94.5
Functional otoscope 109 30.5 20 83.3 129 33.9
Functional statoscope 191 53.5 20 83.3 211 55.4
Spatula 161 45.1 16 66.7 177 46.5
Functional glucometer 127 35.6 23 95.8 150 39.4
HCG test 31 8.7 22 91.7 53 13.9
Blood group testing kit 23 6.4 16 66.7 39 10.2
HIV test kit 30 8.4 22 91.7 52 13.6
Functional weight scale 34 9.5 14 58.3 48 12.6
Height scale (measuring tape) 43 12 14 58.3 57 15
Normal saline 92 25.8 15 62.5 107 28.1
Syringe and needle 140 39.2 21 87.5 161 42.3
Safety box 129 36.1 23 95.8 152 39.9
Hazard bag 76 21.3 23 95.8 99 26
Gloves 300 84 23 95.8 323 84.8
Oral and injectable contraceptives 115 32.2 23 95.8 138 36.2
Bandages 193 54.1 22 91.7 215 56.4
Cotton 255 71.4 23 95.8 278 73
Antiseptics or alcohol 227 63.6 23 95.8 250 65.6
Scissors 153 42.9 21 87.5 174 45.7
Forceps 95 26.6 17 70.8 112 29.4
ORS and zinc 152 42.6 22 91.7 174 45.7
Vitamin A 321 89.9 24 100 345 90.6
Abbreviations: BP, blood pressure; MUAC, Mid-upper arm circumference; ORS, oral rehydration salt.
When asked about the supply system, almost half (48.8%) of the UHEPrs
reported that the essential drugs and supplies needed for the FHT activities
were refilled based on the availability at an HC. Regarding the availability of
guidelines, 77.7% of the UHEPrs reported that the FHT implementation manual
was available, and 91.6% reported that the UHEP implementation manual
page-598
was available. Most participants also reported that the SBCC materials and
referral slips were available. Similarly, the qualitative findings indicated the
availability of guidelines and manuals, which were reported to have facilitated
the FHT implementation approach (Table 5-8).
Table 5-8: Availability of supply management system, manuals and IEC–BCC

materials for Family Health Team

Variables
N % N % N %
Supply chain management
Refilled whenever requested 113 31.7 11 45.8 124 32.5
Yes, refilled based on availability at
181 50.7 5 20.8 186 48.8
HC
Not refilled 63 17.6 8 33.3 71 18.6
FHT implementation manual
Available 277 77.6 19 79.2 296 77.7
Not available 80 22.4 5 20.8 85 22.3
UHEP implementation manual
Available 329 92.2 20 83.3 349 91.6
Not available 28 7.8 4 16.7 32 8.4
Social and behavioral change
communication material
Available 269 75.4 19 79.2 288 75.6
Not available 88 24.6 5 20.8 93 24.4
Referral pad/slip
Available 327 91.6 21 87.5 348 91.3
Not available 30 8.4 3 12.5 33 8.7
Abbreviations: HC, Health Center; FHT, Family Health Team.
5.3.8 Community involvement in the implementation of Family

Health Team activities
Most (78.7%) of the UHEPrs described the WDAs as being involved in the
outreach activities implemented by the FHTs. In this study, 87.7% of the UHEPrs
reported that the WDAs were involved in selecting the priority households to be
visited by the FHT, 95.7% said they were involved in mobilizing the community
page- 599
for sanitation campaigns, and 81.7% said they were involved in baseline data
collection. Moreover, the WDAs were reported as receiving support in training
model households, assessing households’ eligibility for free medical services,
mobilizing the community for enrollment in CBHI, and providing up-to-date
information on health and health-related issues (Table 5-9).
Table 5-9. Role of WDAs in the implementation of Family Health Team activities

Roles of WDAs
N % N % N %
Selecting priority/target households for
259 87.5 4 100 263 87.7
FHT visits
Mobilizing community for sanitation
283 95.6 4 100 287 95.7
campaigns
Involved in baseline data collection 242 81.8 3 75 245 81.7
Supporting UHEPrs/FHT in providing
242 81.8 4 100 246 82
model household training
Identifying households for free
254 85.8 4 100 258 86
healthcare
Mobilizing households to enroll in the
218 73.6 4 100 222 74
community health insurance scheme
Providing up-to-date information about
238 80.4 4 100 242 80.7
health and health-related issues
Abbreviations: WDA, Women’s Development Army; FHT, Family Health Team.
5.3.9 Government sector involvement in the implementation of FHT

approach
Relatively high government sector involvement in the implementation of the FHT

approach was documented for WorHO (78%), and solid waste management
and beautification office (60.4%). The participants reported fairly low or low
involvement of government sectors including woreda administration, education
office/schools, water and sewage authority, woreda youth and sports office,
and private health offices. Nearly half of the UHEPrs expressed that youth
centers, and food medicine and health care authority are closely working with
the FHTs (Table 5-10).
page-600
Table 5-10. Involvement of different government sector offices in the

implementation of the FHT approaches
City Administration
Stakeholders Addis Ababa Dire Dawa Total
N % N % N %
Woreda Health Office 282 79 15 62.5 297 78
Woreda Administration 129 36.1 13 54.2 142 37.3
Education Office and schools 141 39.5 10 41.7 151 39.6
Water and Sewerage Authority 111 31.1 6 25 117 30.7
Solid Waste Management and
221 61.9 9 37.5 230 60.4
Beautification
Health Insurance Agency 125 35 9 37.5 134 35.2
Woreda Youth and Sport Office 80 22.4 9 37.5 89 23.4
Woreda Women’s and Children’s
179 50.1 12 50 191 50.1
Affairs Office Youth Centers
Youth Centers 147 41.2 8 33.3 155 40.7
Food Medicine and Health Care
178 49.9% 9 37.5% 187 49.1%
Authority
Private health facilities 51 14.3% 5 20.8% 56 14.7%
5.3.10 UHEPrs’ feelings about and impressions of the FHT approach
The UHEPrs were asked about their impressions of the implementation of the
FHT approach. Overall, most agreed with the constructs used to assess their
impressions. In this study, most UHEPrs stated that they felt motivated working
with the FHT (83.5%) and believed that they were more confident providing
community-based health services or UHEP packages through the FHT approach
(81.9%). Similarly, 81.1% of the UHEPrs believed that FHT approach had better
community acceptance compared to the previous UHEP approach; 77.2% felt
that they had the opportunity to learn from their day-to-day activity when
working with the FHT.
A slight difference in UHEPrs’ impressions of the FHT approach was observed

between Addis Ababa and Dire Dawa. For instance, compared to UHEPrs
from Dire Dawa, most of the UHEPrs from Addis Ababa reported that they
felt motivated working with the FHT (85.4% in Addis Ababa vs. 54.2% in Dire
Dawa). Most (87.7%) UHEPrs from Addis Ababa stated that they recommend
scaling up the FHT approach in other similar setups (Figure 5-4 and 5-5).
page- 601
Figure 5-4: UHEPrs’ Feelings about the FHT
Figure 5-5: Recommendation of UHEPrs to Scale Up the FHT Approach to

Other Settings
page-602
Similarly, findings from the qualitative study indicate that the FHT has
contributed to the UHEPrs’ motivation and confidence. Moreover, implementation
of the FHT was reported to have helped in reaching out to the segment of
the population most in need of care, thus improving community-based health
service coverage. The participants pointed out that the FHT approach had
contributed to improving volunteerism among the health service providers to
undertake outreach activities.
5.3.11 Facilitators of the implementation of the FHT approach
Several factors were disclosed to have contributed to the effective implementation

of the FHT approach. Community-based health insurance (CBHI) was among
the facilitators of the FHT approach. In the qualitative study, CBHI was reported
to have minimized out-of-pocket expenditure and believed to have resulted
in enhanced health-seeking behavior by the community. The establishment of
steering and technical committees was said to have been among the facilitators
of the FHT approach. These committees comprise leaders of the sub-city and
professionals. Moreover, a weekly performance appraisal, strong supervision,
monitoring and evaluation and the use of the findings were reported to have
positively contributed to the implementation of the FHT.
page- 603
“
When the FHT was established, we discussed how to control it. It is
not working alone. We planned that, every week after doing home-to-
home visits, the team (the team has a sub-team) would go together
or be divided in two and discuss the challenges of the weekly work
every Friday in a review meeting and evaluate it. Also, every 15 days
we report to the higher level. There is also supervision.
HPDP process owner, Dire Dawa
Other factors reported to be facilitators of FTH implementation include

integration and collaboration of the UHEP with other sectors, including woreda
administrations and schools, the availability of infrastructure (like a generator),
and facilities for UHEPrs at the HCs, which improved community awareness,
working in a team, and the availability of guidelines.
In addition, the participants stated that the MoH and RHB provided training to
build the capacity of the UHEPr and develop manuals, which further strengthen
“
and facilitate the implementation of the UHEP and FHTs.
The management is with them, the Health Bureau is with them,

professionals are with them. All the support that they need is within
reach. The relationship of the Health Extension professionals and
other health professionals is strong. The trainings provided for us all
have made it possible for us to build strong relationships. More than
before, we now work toward the same goal. They don’t go empty-
handed into the community. And when they come back, the problems
they bring back with them are discussed and solved in a unified
manner.
HC Head, Dire Dawa
5.3.12 Barriers to implementation of FHT approach
The positive effects of the implementation of the FHT were said to be challenged
by numerous factors. In the qualitative study, the participants explained that
the inadequacy of the budget and of logistics were common obstacles in the
page-604
implementation of the FHT approach. Staff attrition was also reported to be a

challenge to establish FHTs, as per the PHC reform guideline. It also increase
the workload for health workers. The limited integration and coordination of
activities were identified as barriers to the implementation of the FHT approach,
“
as were the limited commitment among some FHT members.
The major reason is the lack of logistics and human resources. For
example, at the Health Center level, 5 to 7 teams are required to set
up. One team contains 8 to 12 team members. No Health Center has
this level of adequacy. There is turnover, displacement, resignation,
and other stuff by workers that make it challenging. The other one
is logistics. Logistics should also be fulfilled. Medical equipment,
glucometers, and other apparatuses are important. These can be
found at Health Centers, but there are medicines—we call them
“essential drugs”—that cannot be easily reached by the community.
These problems prevent the activities from functioning in accordance
with the manual.
UHEPr team leader, Addis Ababa
“
It is better not to lie about the FHT approach. The health professionals
did not accept working outside. This is the truth. They think that
all the community work has to be done 100% by Health Extension
Workers….The main reason is the lack of acceptance and willingness
to work in the community. So, in the kebele where I work, there is no
FHT.
UHEPr, Dire Dawa
The other challenges to the implementation of the FHT approach included

the lack of a clear annual implementation plan, community mobility, less-
established CBHI, and the long walking distance needed to provide services.
Absence of follow-up training on the FHT and transportation issues, the lack
of or inadequacy of guidelines for the school health service, the lack of regular
evaluation, and the imbalance between the community’s expectations and the
service packages were also among the barriers to the effective implementation
of the FHT approach.
page- 605
Moreover, the achievements attained so far are challenged by the resistance

to outreach health services among some health workers. It was also stated
that the other members of the FHT do not make home visits in the absence of
UHEPrs.
5.4 Community engagement and ownership
Findings from the qualitative study showed that different segments of the
community, including religious leaders and the Women’s Development
Army, participate in UHEP activities. Though their participation is limited
and decreasing over time, the study participants mentioned different ways
of community participation, which mainly include environmental hygiene and
sanitation, awareness creation about health and health-related issues, and the
dissemination of health information. In particular religious leaders were said to
be involved in the implementation of the HEP by promoting the use of FP and
tackling misconceptions regarding FP methods.
In contrast, few participants said that the community did not support the
UHEPrs when they conducted home-to-home visits and mobilize the community
for sanitation campaigns. The lack of awareness of the UHEP was reported as a
contributing factor in the limited level of community engagement. Participants
“
suggested that the media should give coverage to the program.
The media has to cover different meetings, and other means should
be used to familiarize the community with the program more. Then,
the community will ask for it. They will say that we need these things,
and they will show you the gaps in what you are doing or telling
them to do. They will ask for supplies to do what they need and ask
for professional help when they need it. So, for the community to
participate, coverage is necessary. For instance, the Family Health
Team is new.
Moreover, participants reported that the current political instability and reduced
emphasis on community mobilization and sensitization were contributing to the
page-606
ongoing decline in community participation and ownership of the HEP. Some

participants attributed the level of community engagement to the residents’
occupations and working hours. There were reports that day-laborers were
interested but failed to participate given the nature of their jobs.
Implementation of the Family Health Team approach was said to have

contributed to the improvement of community participation and acceptability
“
of the UHEP.
The community participation is positive, since they are the first

beneficiaries, especially after the implementation of the Family
Health Team approach. There were many complaints from the
Health Extension professionals regarding community resistance,
before implementation of the FHT approach.
UHEPr coordinator, Dire Dawa
On the other hand, some participants argued that community participation

was limited, sub-optimal, and worsening. The current political instability and
reduced emphasis on community mobilization and sensitization were mentioned
as contributors to decreasing community participation in and ownership of
the HEP. Some participants attributed the low community engagement to the
residents’ occupations. There were reports that day-laborers were interested
but failed to participate, given the nature of their jobs. Perceptions about
and awareness of the UHEP were also mentioned among the challenges to
community engagement in the program.
There were also expectations of incentives for participating in UHEP-related

activities. UHEPrs reported that community engagement was linked with the
availability of incentives (per diems).
page- 607
“
Previously, incentives were given to the community for attending a
particular event organized at the [UHEP] Youth Center. Now, they
may not show up if you call them for a meeting.
UHEPr, Addis Ababa
In this study, 1 734 respondents representing their respective households were

asked to rate the level of UHEP package implementation in their area. The
result showed that 55.4% of respondents perceived the UHEP as being well
implemented, while 38.4% perceived it as poorly implemented in general. As
depicted in Figure 5-6, a greater proportion of respondents from Dire Dawa
perceived the implementation of the program as poor.
Figure 5-6: Implementation of UHEP as Perceived by the Community, Response

from Household Survey
As perceived by households, the highest implementation of UHEP packages
was 57% (implementation of solid waste management), and the lowest (28%)
was for youth reproductive health services. The results indicated that the
communities perceives none of the UHEP packages or services as adequately
implemented (Table 5-11).
page-608
Table 5-11. Perceived Level of UHEP Package implementation among Households
Perceived Household-Level Addis Ababa Dire Dawa Total

Implementation of UHEP Packages/
Services (well implemented) N % N % N %
Personal hygiene 662 59% 222 36% 884 51%
Food and water hygiene 622 55% 212 35% 834 48%
Latrine use and construction 622 55% 221 36% 843 49%
Solid waste management 777 69% 213 35% 990 57%
Maternal health (antenatal and
581 52% 213 35% 794 46%
postnatal care)
Child immunization (deworming and
616 55% 228 37% 844 49%
vitamin A)
Nutrition (growth monitoring and
451 40% 179 29% 630 36%
nutritional screening)
Youth reproductive health 314 28% 164 27% 478 28%
Family planning 601 54% 202 33% 803 46%
HIV prevention and control 589 52% 196 32% 785 45%
Tuberculosis and leprosy 455 41% 208 34% 663 38%
Malaria prevention and control 312 28% 250 41% 562 32%
Non-communicable disease 460 41% 199 33% 659 38%
Mental health 336 30% 157 26% 493 28%
5.5 Leadership and Governance of the UHEP

POLICY, STANDARDS AND STRATEGIES OF THE HEP
The national health policy focuses mainly on providing well-promoted, preventive,

and selected curative health care services in an accessible and equitable
manner to reach all segments of the population. The Health Extension Program
serves as a primary vehicle for prevention, health promotion, behavioral change
communication, and basic curative care. Participants reported that they were
focusing on prevention.
page- 609
“
Priority is given to the health extension program, Ethiopian health
policy focuses on disease prevention and health promotion.
Program officer, SNNPR
Some participants reported that there was a manual in their workplaces

and that proper use of the manual was assessed quarterly during supportive
supervisions and during reporting.
“
Sometimes they [UHEPrs] are not following the guideline. We redirect
and tell them that they are not following the guideline. Quarterly
supportive supervision is given by our staff, and these things are
assessed during reporting.
RHB Head
Some participants reported that they were not following the standards,
especially the multidisciplinary approach, in their FHT because of a lack of
personnel at the HC, a lack of motivation and management skills on the part
of the professionals, budget constraints, and the work burden.
“
When the Family Health Team is established, the number of
professionals should be at least 8 to 12, according to the standard,
but do all implement the FHT according to that? No, because
the civil service office has already banned hiring, so the program
designers themselves forced us not to work based on standards. So
we can’t control it.…So, it is also determined by human resources and
the professionals’ motivation.
HPDP, process owner, Addis Ababa
page-610
Participants described gaps related to the manual’s development, distribution,

and availability, implementation for a certain target group, personnel, and
commitment to work as per the guideline, supervision, monitoring and evaluation,
administration, and collaboration.
CHALLENGES RELATED TO THE MANUAL AND ITS STRATEGIES
Related to the manual’s development, participants mentioned that the study

was conducted before the HEP was optimized, but it was the qualitative
study that did not allow them to see the magnitude of the problem and plan
“
appropriate action.
On the optimization of the HEP, all the existing evidence was

very useful, and they benefitted a lot. Many data were collected
qualitatively.… Based on the existing evidence, mainly by engaging
stakeholders, the action points that were drawn were good. On some
action points, additional data might be needed, though….
Program Officer, HEP & PHC Directorate
In addition, using the experience of the RHEP in preparing the UHEP manual
“
was reported by the participants as a gap:
I think a survey was conducted. But it was not necessary for the city
to copy the rural HEP directly. The approach should be different…
shifting the program for the UHEP to the FHT approach is not
appropriate, because of the inadequacy of resources, the number of
people or households per team is not considered. For example, 60 000
people in a woreda are served in a Health Center. So there should an
assessment of the resources and personnel before designing it.
UHEP team leader, Addis Ababa
The participants also suggested increasing the number of UHEPrs in the urban
settings to address the geographical disparity between the rural and UHEPs.
page- 611
“
The landscapes in rural and urban areas is not similar…If the number
of UHEPrs is sufficient in urban areas, a good benefit will be obtained
from the program, because the training that the UHEPrs received and
the way they implement the urban and rural programs is different.
HEP coordinator, Oromia
Regarding the strategies of the HEP, participants stated that it lacked a

strategy for NCDs. They also mentioned that a gap in strategies related to
UHEPrs’ transfer opportunities is now solved and that they are now allowed to
work where they wanted to work.
GAPS RELATED TO THE AVAILABILITY, DISTRIBUTION, AND USE OF

THE MANUAL AND GUIDELINES
Participants mentioned that, even though the HEP guidelines were developed,
there is a problem in its availability and distribution for implementers and a
problem putting it in place and using it even after it was distributed.
“
We then urgently printed it [the second-generation HEP manual]
out and distributed it to the UHEPrs. As a bureau, we support the
woreda and Health Centers’ working based on the standards and
guidelines by giving them follow-up. But when we travel to Health
Centers and Health Posts and check the availability of the guidelines,
we didn’t see the guidelines in places….and when we go to the Health
Center, there is a problem with Health Center heads, as they do not
distribute properly what they have already received from us.
HEP coordinator, Harar
page-612
“
Due to inadequate support and monitoring, there is a gap in use of the manual:
You are right, there are problems in the use of the existing guidelines;
this is because of poor or inadequate monitoring and support related
to owning the program in order to achieve the benefits, as I mentioned
earlier, starting from the top down, including myself.
UHEP coordinator, Dire Dawa
GAPS RELATED TO WORK BURDEN
The work burden related to coverage of a large population and low personnel
“
was also reported as a challenge to working according to the guidelines.
Even if the manual says that one Health Extension Worker is expected
to serve 500 people…in reality, some of us are serving more than
800. This makes it difficult to implement the program and to reach
everyone.
UHEPr, Mekelle, Tigray
In the case of the FHT, the manual suggests intervention in the community
4 times per week and specific schedules for home-to-home visits and visits
to workplaces, schools and the homeless, which are compromised because of
inadequate human resources.
“
The standard dictates that 1 FHT conduct community work 4 times a
week. But that cannot happen because of the workload at the Health
Center. But it still addresses the program perfectly in the two days.…
It says that the professionals visit the youth centers twice a month,
specific schedules are also available for home-to-home, workplace,
schools and homeless tasks. This has not been met.…This is because
of the professional shortage at the Health Center.
page- 613
CHALLENGES TO IMPLEMENTING THE HEP FOR SOME TARGET

GROUPS
Participants mentioned that some manuals were difficult to implement for

some beneficiaries like homeless, people of low economic status, youths, and
schoolchildren.
Regarding the homeless, it is a challenge to track homeless individuals because

they have no address for curative services and follow-up. Getting free medication
“
to them was also mentioned as an additional challenge.
When they [homeless individuals] come for healthcare services,

and free health service is authorized by the kebele, since they don’t
have a house number, this halts the activities among the homeless….
There is no implementation on the street. Even if we go to give them
healthcare service and referral, they don’t come, and, again, we can’t
go to the same place and bring them….
FGD participant, UHEPr, Dire Dawa
Participants also reported challenges related to the affordability of medication,

getting free treatment for people of low economic status, and the professionals’
“
commitment.
Especially there are no clear guidelines on the budget to provide

drugs.…Even if we provided free drugs in an emergency, health
insurance service is not available at all levels.
Regarding youths and school, participants reported that the implementation

was hindered by a lack of cooperation from the schools and a lack of acceptance
of the HEP service by students, who even ask for incentives for the services they
receive.
page-614
“
Satisfactory work is not done at the school level.…At some places,
they don’t even acknowledge you, especially private schools, even the
schools refuse to vaccinate. In these places, unless they are forced by
other bodies, we cannot do it alone. They refuse vaccinations that are
done in a national campaign.
UHEPr, Dire Dawa
Some participants reported that other implementers were incentivizing students

and, thus, students expected incentives from the HEP, which in turn creates a
“
barrier to good communication.
There is a girls’ club at school, so we planned to educate them about

family planning and how to prevent abortion. When we did that, the
children asked for money. They said, “Are you paying me?” This is
because they are given 50 or 100 birr when other organizations or
SOS are teaching them. So, they participate only once if it doesn’t
have money, and there is no coordination of this by the director or
other coordinators to learn from it.
UHEPr, Dire Dawa
Related to the HEP service for youth, some participants mentioned that the
youth centers were rented for other activities so that UHEPrs had difficulty
finding a place to meet with the youth and give them proper education about
“
health and health-related problems.
Another is the youth center. It is not functional, and there is one in

each kebele, and instead of using that for youth, the hall is rented for
meetings. We had started previously, but we didn’t find anyone, so
we sit. And we get back, if the youths are not present, whom will you
will work on? So we are not doing anything on the youth center, and
even if you go, you will see the hall rented for different meetings.…
There is nothing that attracts the youth.
UHEPr, Dire Dawa
page- 615
CHALLENGES RELATED TO COLLABORATION
Community acceptance is a key factor in the effectiveness of the HEP. Some

participants stated that, in urban settings, due to increased community
awareness about health and health-related issues and other factors, they are
reluctant to accept the service provided by UHEPrs. In addition, people may
“
not be at home during home-to-home visits.
It is not implemented as per the guidelines due to the challenges posed

by the beliefs, attitudes, and perceptions of urban communities.…They
think that they are well informed. Thus, they never open their doors
to us [UHEPrs] to talk to them.
FHT coordinator, Dire Dawa
Participants also mentioned that the standards did not consider the actual
status of the country, especially the laws and regulations that can affect the
implementation of HEP directly or indirectly, like water, sanitation, and latrine
“
construction.
Latrine and fluid infiltration: the standard says 3 months. As to our

woredas, we cannot allow the building of latrine and water infiltrations
because it is the jurisdiction of the house and building that authorize
building latrines. The owner of the house will not allow building a
latrine unless permission is given by the authority.
UHEPr, Addis Ababa
Some participants, however, reported that stakeholder involvement was

inadequate during guideline development and that there was a lack of proper
handover of the document.
page-616
“
When program-specific guidelines were developed, they may have
good knowledge about the program but poor on M&E, or vice-
versa…so when we want to develop one manual, people from M&E
have to be invited...because the gap that we have now is here…
Second, the biggest problem is that we face high turnover of those
who participated in the first discussion. There is no proper handover
when the new staff come.
Program Officer, MoH
“
I do not expect people who are involved in planning and evaluation to
know as much about the HEP as a health program officer. Generally,
the program lacks harmonization in terms of resources. That is why
we have a lack of budget, human resources, and materials and the
support in planning, monitoring and evaluation get less attention.
RHB Head
CHALLENGES RELATED TO IMPLEMENTATION
Participants described different challenges related to the implementation of

the HEP. Among them, the inability to make the UHEPrs work where they were
born or in the community they were living as was originally planned to ensure
“
sustainability was a challenge.
When the program was designed…UHEPrs born in each community

were selected and trained to serve the community by residing
within the community…but the UHEPrs started to resist working in
the designed way. Because of this, the community members are not
getting the expected support from the UHEPr.
WorHO head, Oromia
page- 617
Moreover, they also stated that not all packages were implemented equally.
Only selected topics were emphasized, like TT immunization, FP, and HIV
screening.
“
It would be difficult to say that full packages that need to be given
to households are being provided. One piece of evidence that will
prove this is the less significant increments in the number of model
households. It is not what was expected. Model households are those
who complete all packages.
Participants also suggested that the information system to be used would

inform policy.
“
When we plan for 5 years, the guiding principle has to do for 5 years
and get the outputs…so that the information system that can serve
for this purpose could be designed. You cannot change every time
new things come rather than align to them. If we do so, then we
cannot measure the health system.
page-618
The participants also offered their view that emergency cases and road traffic
accident management should be included in the manual. Additionally, they
suggested that UHEPr supervisors should have a role in the FHT and should
“
also be included in the manual.
There are areas that need improvements. For example, the Health
Extension supervisor doesn’t have a role in the Family Health Team.
Process owner, Addis Ababa
Participants mentioned that there were 3 different guidelines for the UHEP:
the UHEP implementation guidelines, the FHT implementation guidelines, and
the urban CBHIS manual. This creates ambiguities regarding which one to use.
They proposed that these 3 manuals be merged and updated.
“
When we come to the UHEP, there are 3 types of manuals, namely
the urban program implementation guidelines, the Family Health
Team implementation guidelines, and the urban community-based
health information system manual. These 3 manuals have created
some ambiguities in some places. Therefore, there is a need to
consolidate them and make one single guideline.
5.6 Existence of an enabling non-HEP policy,

guidelines, and strategies that support the HEP
Participants stated that the community-based health insurance (CBHI) policy

was a facilitating factor for the HEP.
page- 619
“
The new initiatives that can contribute to the HEP are the CBHIS,
community-based health insurance. This system was introduced after
we started the primary health care unit. When you go out to the
community, non-communicable diseases are common. The medicines
for these illnesses are costly. When you do community work, you focus
on the poor. Once you give them medical services, they should be able
to use medicines that they can’t afford…When they get registered
in the insurance [CBHI], they pay around 350 to 370 birrs. That
payment will serve them for a whole year. This system supports the
Family Health Team as well.
Participants also acknowledged that other programs started by the government

initiative, like the school feeding and safety net program, helped facilitate the
“
HEP.
The Federal Program and Primary Program were recently structured.

They are working on what we are going to support and where the
problem is. For example, they supported us in the 2010 assessment to
be effective, in school nutrition and the health program…but in terms
of finance, I believe it has not been enough, and there is a gap in
verifying and validating the reports and the veracity and accuracy
of the report.
RHB Head
page-620
5.7 Organization of the HEP, Coordination, and

Collaboration
5.7.1 Coordination of the HEP among the health sectors
The Ethiopian health system has 3 linked levels of healthcare: primary,
“
secondary, and tertiary.
As is known, our system has 3 guiding systems; primary, secondary,

and tertiary.…I believe that it is all linked, since coming to the Health
Centers is inevitable due to diseases like TB, malaria, and HIV.
RHB Head
Regarding the UHEP, Health Extension professionals are working with other
health professionals at the HC as FHTs and targeting 5 population groups:
home-to-home, school, youth, the homeless, and workplaces. The participants
also described each team as being composed of professionals from different
“
disciplines and working in rotation in the community.
The team works on the 5 areas, which are the home-to-home, schools,
youth centers, homeless, and workplace. That means that the last
2 are added to the Family Health Team. But the regular programs
are home-to-home, schools, and youth centers. On these other 2, the
Extension professionals work together with the Family Health Team.
Still, participants reported that the link among the various level of the health
system and among departments were not as strong as expected and perceived
that the HC staff were not as concerned about the community as the HP staff
were. Some participants also reported that there was no direct linkage between
the HC and WorHO that enabled regular communication between them.
page- 621
“
It is possible to say that no linkage exists. The linkage may be when
we meet with the Health Center medical director on meeting, some
political issue that need to be solved politically. Rather than this,
no direct linkage between the HC and the Woreda Health Office
because the Health Center is administratively responsible for the
sub-city administration.
“
WorHO, Yeka sub-city, Addis Ababa
There was no agreement with other programs. There was no

integration between the departments. One department will allow the
health providers to give the service and not the others. There is a big
gap in the integration and uniform activity of the system.
UHEP supervisor, Amhara
COLLABORATION WITH OTHER SECTORS
Collaboration is the process of working together with different sectors to

achieve a common goal, as well as jointly working with sectors including sharing
activities, working based on a schedule, reporting, and evaluating each other
in review meeting. The HEP is working on Health Promotion and Disease
Prevention, which are associated with different social, economic, environmental,
and psychological well-being factors of the community. Therefore, it needs
collaboration among various sectors that contribute to the HEP directly or
indirectly to improve the HEP’s implementation.
Participants suggested different strategies for collaborating with different

social and religious associations and governmental and non-governmental
organizations (NGOs), such as preparing discussions with officials, forming
a task force, using existing social, religious, and political institutions like
Idir–a traditional burial society that supports orphaned children–signing a
memorandum of understanding, and doing community mobilization.
page-622
“
There are a lot of task forces that we formed. There are technical
working groups.…We have different things to do together with them.
Participants reported that they were working with different governmental

organizations, like the Education Bureau, technical and vocational and training
institutes and universities, the sanitation and environmental office, and the water
and sewage office, and working jointly with police to intervene in situations of
domestic violence. They also reported that those sectors and organizations
“
played different roles.
The kebele is our first stakeholder. Schools, the Women’s and Children’s
office, especially related to the women’s developmental teams,
cleanliness and beauty with respect to the disposal of dry wastes, the
water and sewerage authorities with respect to the disposal of liquid
wastes and access to clean drinking water, universities with respect
to identifying problems and carrying out research; TVET training
institutions with respect to training the Health Extension professionals,
community representatives, elders, religious institutions—all these are
our stakeholders.
Participants stated that they were working with the Women’s and Children’s
Affairs Office, the Education Office, the Sanitation Administration, the
Water and Sewage Office, and municipalities as important stakeholders. The
participants reported that the HC was responsible for establishing collaborations
with other sectors.
Participants believed that working with other sectors had both benefits and
challenges. They reported that collaboration was used as a means to get faster
and better results, while some reported that it also necessitated effort and
extra time to incorporate different opinions.
page- 623
“
When you work in groups, you get better results, but time management
does not always work, because everyone’s opinion is included. But to
get quick, better results, I believe working in groups is ideal.
People are not equally responsible, however, when they participate in health-
related projects.
“
“It is written in the book,” they said, but there was nothing when we
came to practice. For example, there was a women’s team, which
was selected from the Agriculture Office, the Women’s Development
Team, and the Education Office, but since the health sector has more
tasks to do, the tasks are being done only by the health sector. For
example, when we came to Agriculture, they were doing few tasks
since it’s an urban area. But they don’t participate with us.
“
UHEPr, Assosa, Benishangul-Gumuz
The Health Center invites political leaders who work at kebeles and
other stakeholders, such as schools and celebrities. But they decline
our invitations. We are ineffective in this area. They never participate
in our work because of negligence, resistance, and lack of focus.
FHT coordinator, Dire Dawa
People also mentioned that collaboration was crucial in the UHEP since
the community needs and other health-related problems cannot be solved
by UHEPrs alone irrespective of their effort. They also mentioned that the
communities didn’t understand UHEPrs’ duties and responsibilities.
page-624
“
She [the UHEPr] advises the woman about the importance of water,
like sanitation. If there is no access to clean drinking water, the
community blames her for not providing that. The society doesn’t
differentiate the tasks of the different sector offices.…So, they
[UHEPrs] need a supporter on this.
“
RHB Head
There are problems that couldn’t be solved by UHEPrs irrespective of

their efforts. For example, to control problems related to latrine excreta,
which was being disposed in the open fields and rivers and ditches,
is the responsibility of the administrators...They [UHEPrs] report to
the concerned body that the problem is beyond their capacity.…The
UHEPrs have been doing their best, but the administrative body is
not solving the problem.
UHEP supervisor, Addis Ababa
For the packages to be implemented well and be effective, it is necessary

to work in collaboration with other sectors by motivating people through a
preset plan and awareness-creation in an integrated way. Participants also
mentioned the need for collaboration to discourage those who are disregarding
“
the UHEPrs’ teachings.
[A] lack of…, motivating or sensitizing the community is not being done.…
Packages require the active participation of different stakeholders.…
Those packages are not well implemented, and consequently, their
acceptance is decreasing. For example, if a person disposes of dry
wastes inappropriately, he or she should be charged and fined by the
responsible bodies.
They also mentioned that some government sectors were not responsive as
quickly as expected for health and health-related problems or with respect to
working in collaboration with UHEPrs.
page- 625
“
The other thing is that the city administration or municipality should
support them.…Garbage is collected regularly, but there is no vehicle
for transporting it. If they didn’t take it in a timely way, it is all
the same. The garbage is out of my house and polluting the next
household. The UHEPr cannot solve this alone. The garbage can be
collected, but if good measurements are not taken, it has no benefit.
Female community member, Assosa, Benishangul-Gumuz
Participants also recommended that, in addition to collaborating and forming

a committee, proper regulation needs to be in place to tackle the problem.
Additionally, they recommended a strong link between the UHEP and
“
municipalities to work on liquid- and solid-waste management.
In urban areas, there is a 7-sector committee. Based on the urban

sanitation strategies, steering committees are formed. But, the problem
of sanitation cannot be solved by education and other committees.…
The committee is not functional or productive. The problem can be
solved only by regulation, as it is everywhere else in the world.
Program officer, MoH
Participants also complained about the lack of responsible bodies to govern,
“
direct, and implement the collaboration among HEP and other stakeholders.
It is known that this program demands the contributions of many

stakeholders.…For example, the existence of schools leads health
and educational offices to work together. When we take a latrine,
there is a standard set in the manual for how many students should
be served by a single hole and handwashing facility near to the
latrine. The educational office is responsible for making these things
available. But sometimes there are mandate issues arising between
these two offices.…There is no responsible body that governs, directs,
and monitors their activities in terms of collaborative actions.
page-626
WORKING WITH OTHER RELIGIOUS AND SOCIAL ORGANIZATIONS
Some participants stated that UHEP works with faith- and community-based
associations, like Idir, which allowed them to use their halls for HEP-related
“
meetings.
We work with Idirs…to support orphans and the needy. We register

and link them [orphans] to supporting organizations. There is one Idir
here - we gave them 10 children last year and 5 this year.
UHEPr, Kirkos Sub-City, Addis Ababa
There are also NGOs working with the UHEP and supporting HEP activities in
various ways for a specific period of time. Some participants reported that NGOs
supported HEP in terms of financial assistance and logistics like healthcare kits
and drug supplies for specific diseases and drug management. Participants also
reported that there were NGOs that provided technical support, like capacity-
building for UHEPrs and the community through education, post- and pre-
service training, mentorship, skills development, campaigns, technical advice,
“
and manual preparation.
They [NGOs] used to support infrastructure, training. They had

professionals at each woreda who used to go for supportive supervision,
and they gave financial support to the trainings. They supported
the 5 woredas in competency-based programs; they supported the
manual preparation and covered all related costs of training.
Other NGOs are involved in different activities, such as system-strengthening

through FP, immunization, ambulance service, and the transformation of
“
woredas.
Ambulance service runs day and night, and the UHEPrs are facilitating
this by communicating with the Red Cross.
page- 627
5.8 Health Information System Assessment

5.8.1 Data collection and reporting formats, tools, and guidelines
Using similar and standardized data collection and reporting formats is critical
to collecting valid and reliable data, enabling the comparison of findings in
the standard set and with findings in a similar setting, and making decisions.
In Ethiopia, the Health Management Information System (HMIS) is the data-
collection system designed to support management, and decision-making in the
health system. At the community or HP level, the Community HealthInformation
System (CHIS) is used. It organizes information on individuals’ andfamilies’
information related to vaccines, FP, maternal and child health, HIV
treatment and support, and other services and makes it available all in one
place, such as a Family Folder (FF).
Participants from the MoH also reported that District Health Information
System, version 2 (DHIS-2) software was implemented to enable the MoH to
receive data from the primary sources, which minimizes information inconsistency
“
and improves the data management system.
The Ministry of Health has implemented an effective mechanism in

the DHIS-2 information management that permits the information to
be entered directly into the Federal Ministry of Health’s server from
the primary source, which minimizes the information inconsistency
greatly. It also can help us have better data management.
On the other hand, some participants from HCs said that UHEPrs and FHTs
reported using the HMIS(which has a minimum standard for their weekly,
monthly, and quarterly performance) which sends to the HC, then the HC
“
organizes it and sends it to the Health Bureau.
The HMIS is collected monthly and quarterly. The format is similar to

the urban and rural….This is also organized quarterly to monitor their
[UHEPrs’] skills and to strengthen their work. We collect the reports
on a weekly basis, then submit it monthly to the Health Bureau.
HEP supervisor, Assosa, Benishangul Gumuz
page-628
“
The FHT have their own way of recording. When the report is sent
from the Health Center, it is consolidated by the HMIS system.
Some participants also mentioned that HMIS did not incorporate all
activities found in the 16 packages, like the WaSH components, which in turn
“
underestimated the UHEPrs’ performance reporting.
The activities done by UHEPrs were not included in HMIS. For

example, there are 16 packages, and 16 of these packages should be
included in HMIS. We saw that with WaSH. There are 9 packages
(previously 7), but only 2 or 3 packages were included in the HMIS
while the others were not counted in the report. So, how are the
packages going to be measured?
RHB Head
On the other hand, some participants indicated that the reporting format
lacked the process indicators and measured only outcome or output. For
example, while it counted the number of model households, it did not contain
information on their process or progress.
“
There are no indicators on knowledge, practice, or attitude. Almost
the only things there are output indicators. It only asks whether there
is a model household or WDA…how many are functional. But it does
not show the preceding processes that show continuity, including their
training, recruitment, and selection. There may be dropouts. In what
way they have graduated? At what level are they now, etc.? are not
shown.
page- 629
Additionally, some participants said that the Family Health Recording form was
not comprehensive and only allowed them to register the number of individuals
who received the intervention rather than the type of intervention and time
“
spent to provide the intervention.
I personally don’t like the Family Health form. We work beyond the
form. For example, we record and take pictures with our mobile phones
and file a report showing them what we have captured. For instance,
if there is a person with Diabetes, whether you give [the intervention]
for 1 or 3 individuals, you write 3, but you might have spent 3 hours
on counseling…and write a referral. It does not state that.
FGD participant, UHEPr
The other issue raised related to the CHIS was its implementation in all places.
Some reported that the CHIS was not implemented everywhere, and, as a
result, there was no standardized, comprehensive format to register their daily
“
activities. One participant described this:
There is no standardized registry for daily activities.…As I said, there

is a register prepared by JSI, and the Health Extension professionals
register their daily tasks in this registration book and give us what
they registered by separating it from the pad sometimes, because
compiling tasks are not strong yet. So, we accept the rough reports.
For example, let’s say the number of waste-disposal holes dug is
mentioned. The name and address of the household members is not
registered along with it to evaluate the work by actually going to the
site.
The implementation of Urban Community Health Information System (U-CHIS)

was assessed in this study. In this regard, the assessment showed that about
44.6% of respondents indicated that the U-CHIS was used for the urban
program. All respondents from Dire Dawa stated that the U-CHIS was not
implemented at all.
page-630
More than half of the study participants (50.8%) supported the availability of
the system for monitoring UHEPrs’ timely reporting and completeness at the
HC level. On the other hand, the availability of the monitoring system in Dire
Dawa was low (14.3%).
The assessment revealed that a high proportion (84.1%) of HCs had a system to
monitor the timeliness and completeness of UHEPrs’ reporting. Similarly, most
(80%) respondents thought that there were encouraging practices of UHEPrs
updating their catchment population profiles; among them, 75% updated every
year. These profiles include information like the number of pregnant women,
children under 1 year of age, children under 5, and household latrines (Table
5-12).
Table 5-12: UHEP Health Information System implementation among HC in

urban centers
Addis Ababa Dire Dawa Other Towns Total

Characteristic
N % N % N % N %
Urban community health information system
Being implemented 47 51.1 0 0.0 11 35.5 58 44.6
Not implemented 45 48.9 7 100.0 20 64.5 72 55.4
Catchment population profile updating by UHEPrs
Yes 78 84.8 3 42.9 23 74.2 104 80.0
No 14 15.2 4 57.1 8 25.8 26 20.0
Frequency of catchment population profile updating by UHEPrs
Semi-annually 15 16.3 0 0.0 3 13.0 18 17.3
Annually 59 64.1 3 100.0 16 69.6 78 75.0
Every 2 years 2 2.2 0 0.0 2 8.7 4 3.8
Every 3 years 2 2.2 0 0.0 2 8.7 4 3.8
5.8.2 UHEP Reporting System
Data reporting is the process of collecting and submitting data on time and
using an appropriate format to help make appropriate decisions and plan,
allocate, and receive the budget. Regarding the frequency of reporting,
page- 631
participants stated that the UHEP’s activities varied in terms of their nature
(multi-purpose and multi-level) and type of report, such as telephone or written
report. Participants stated that there were vertical and horizontal reporting
chains: one from the UHEPrs or kebele to the HC and the other from the
UHEPrs or kebele to the WorHO and then to the sub-city. Some WorHOs
report to their respective Zonal Health Departments.
Participants reported that UHEPrs submitted their activities and monthly

reports to the HC, then the HC reviewed and gave feedback to the UHEPrs
weekly. In the case of the FHT, the leaders gave a signed report to the
coordinators. In both cases, the coordinator or the disease prevention process
unit organized the report and sent it to the sub-city quarterly.
“
The Health Extension Workers have a daily reporting format, and
they report to the Center.…We [the HC] evaluate who has performed
better and who has performed worse. Depending on the evaluation,
we write feedback for them.
“
UHEP supervisor
The Family Health Team has a coordinator, and the Family Health
Team leader gives signed reports to the coordinator weekly.…The
coordinator integrates the reports of the 5 Family Health Teams and
sends them to the sub-city.
FHT coordinator
Some participants also mentioned that there were UHEPr supervisors at the
woreda level who were responsible for compiling data and sending it to the
“
sub-city, which then sent it to the Health Bureau quarterly.
In our case team—I have said there are 10 woredas—these 10 woredas

have monthly reports, and there are supervisors leading the Health
Extensions. They bring the reports, and we [the sub-city] compile that
in Excel and report to the Bureau monthly, quarterly…
Program Officer, Addis Ababa
page-632
Additionally, some participants mentioned that UHEPrs wrote for and reported
to different stakeholders and that there was a parallel reporting system. This
formed a challenge because UHEPrs have to write and report at the expense
of their focus on their assigned job.
“
Be it HMIS or CHIS, the programs run in parallel, and all they do is
write different reports, but they don’t have a permanent system that
could let them focus on their work. I think the system does not give
them the chance to focus on their jobs. When we go to the lower
level, work is not done only based on the Family Folder. There are
reports that will be sent to the kebeles, reports that go to different
programs in parallel. You will find ample reports when you go to the
Health Posts…
Regarding the content of the reporting, in addition to the CHIS, participants

specifically stated that the UHEPrs reported both CDs and NCDs—challenges
beyond their capacity—when they suspected epidemic like diseases like acute
“
watery diarrhea, measles, and scabies.
They also work with PHCUs and have reporting systems and meetings
too.…When there are epidemics of diarrhea affecting more than 10
people, they report it immediately. Scabies…or measles cases are also
reported. For instance, last time there was a situation that looked like
measles, and they reported it properly, and actions were taken.
RHB Head
5.8.3 Quality of reporting and verification
Participants described the quality of reporting in terms of trustworthiness,

completeness, and using the correct indicators to measure or calculate the
intended outcome. Most participants agreed that reporting using a system
like CHIS or other format was poor most of the time. They also mentioned
inadequate UHEPr skill in recording and reporting in CHIS, which resulted in
incomplete CHIS reports. Others said that the CHIS itself was not refined and
detailed enough.
page- 633
“
The data elements are not properly collected or written in the data
sheets.…The CHIS is not well coordinated. It is lacking when checking
monthly because it is not detailed and refined when it transfers from
the lowest level to the federal level…
“
RHB Head
Yes, I am saying that, you will show them the reporting and recording,
and, again, when they [UHEPrs] are back, they don’t fill it out the
way they saw.…They say that they have given health education, but
when you ask them for data, there is nothing to see….So, the skill in
recording data is very low. The other thing is that there are diseases
in the HMIS, so based on that, their [UHEPrs’] skill in collecting data
or entering it accurately is very low.
Program officer, Assosa, Benishangul Gumuz
Some other participants stated that, because UHEPrs were given an unrealistic
and unachievable plan or because of fear of criticism and punishment, the
UHEPrs over-, under-, or falsely report . The poor layout of the registry and
“
Family Folder were also reported as a reason for inaccurate reports.
You will find the data about how many graduate models there
are at the end of the report. This can be manipulated by Health
Extension Workers or the woreda. Because there is a public image,
mostly numbers are added.…The other thing is when they are told
that maternal deaths are auditable, they tend to under report them,
fearing the audit. We can correct this by raising their awareness.
page-634
On the other hand, some participants stated that the national standard had a
targeted plan. Those indicators were not measured properly at the lower level,
but rather reported at the end as though they measured and achieved the
target during reporting.
Regarding verification of the reporting, participants reported various
“
mechanisms, such as interviewing the community or auditing clinical treatment.
We monitor the service quality, including awareness creation by the

UHEPrs. We conducted interviews with households in the community.
We also do clinical auditing. For example, we checked the service
quality of the integrated management of childhood illness. We
checked whether they diagnosed it based on the national guidelines,
whether they classified it correctly, whether they provided medicine
correctly…
HC Head, Amhara
A few participants reported that working as a FHT helped them avoid false
“
reporting. One participant described this:
If the UHEPrs don’t report the correct and accurate work they do on
a daily basis, there’s no mechanism to check it…But now, since they
have been deployed with the FHT, we can identify health workers
who participated in the Family Health Team and those who didn’t.
Findings from the quantitative study also showed that there was a poor
monitoring mechanism to check the quality of UHEP reports at the HC level.
Only 50% of HCs had a documented mechanism for checking reports’ timeliness
and completeness, and about 16% of the respondents reported the absence of
a data-quality assessment system for the UHEP in the HCs (Table 5-13).
page- 635
Table 5-13: UHEP Reporting and Data Quality Assessment Practice in Ethiopia

Characteristics
N % N % N % N %
UHEPrs reporting a timeliness
and completeness monitoring
mechanism at Health Center
Yes, observed 50 54.3 1 14.3 16 48.5 67 50.8
Yes, not observed 39 42.4 5 71.4 10 30.3 54 40.9
No 3 3.3 1 14.3 7 21.2 11 8.3
Data quality assessment for UHEP
Yes 85 92.4 4 57.1 22 66.7 111 84.1
No 7 7.6 3 42.9 11 33.3 21 15.9
5.8.4 Monitoring and evaluation of UHEP
STAKEHOLDERS IN MONITORING AND EVALUATION
Monitoring and evaluation of the HEP is a key component that involves different
stakeholders, like the MoH, RHBs, WorHOs, HCs, and community members at
different levels.
Participants stated that they used a horizontal and vertical monitoring and
evaluation system. The level and domain of monitoring and evaluation differ
for different program staff, UHEPrs, and the community. The MoH and RHBs
play various roles, but most participants stated that monitoring the budget and
progress of the HEP were an area that needed improvement.
“
The Bureau’s role is to monitor and secure enough money, provide
meeting halls, etc. Furthermore, they should check the progress and
pinpoint areas that have setbacks so that they can discuss them
with the sub-city and other office managers and consider different
approaches to solve the problems.
RHB Head
page-636
Women’s Development Army (WDA) participants also mentioned that UHEPrs

evaluated their work and gave them feedback on activities that needed further
improvement, and that the WDA were also involved in evaluating the 1-to-5
“
groups.
I have told you that the UHEPrs give us grades. We like that part. We
also give these grades to the 1-to-5 groups. If a woman has a toilet-
use problem, we give her a B. When we report that she works on the
compound, which is graded as a B, people who were graded B will
try to come to the A places. We like this system. We have taken such
lessons from them and we are working the same way. If my neighbor
is better than me or she has a B grade, she is graded as a B because
there is another person better than her, so we try to bring her to the
A position together. So, we think this method is very good.
WDA, Assosa, Benishangul Gumuz
Another key informant from Addis Ababa also mentioned community

participation in monitoring and evaluation as follows:
“
There is a board established at the Health Center that reports to
the woreda administrator. There are evaluation criteria called a
Community Score Card (CSC) in which the evaluation of the HC
is conducted by members of the community, and the representative
comes to the HC. They evaluate it and give the results to the Health
Center.
Program Officer, Addis Ababa
FREQUENCY AND QUALITY OF MONITORING AND EVALUATION
Regarding the frequency and quality of monitoring and evaluation, most

participants from the WorHOs, RHBs, and MoH reported that they monitored
monthly and quarterly, while participants from the HCs in the regional and
woreda towns (i.e., not in Addis Ababa or Dire Dawa) sated that they gave
page- 637
supportive supervision and monitoring and evaluation on a daily and weekly

basis. In the case of the FHT, HEP coordinators in the HCs evaluated their work
“
every week.
There are different stakeholders who do monitoring and supervision

activities. For instance, the Zonal Health Department did supervision
quarterly. The Woreda Health Office also did regular supervision.
The HC–UHEPr linkage expert at the Health Center did a frequent
supervision every day at their sites. Every week, we discuss among our
team how things are going and moving forward. There are weekly
reports, and we have a meeting to discuss that.
Most participants, however, also mentioned that the schedule and quality of
monitoring and evaluation was inconsistent and differed by place, time, and
level of health care.
“
Monitoring means following the process of the work. For example, if
she gives vaccinations monthly, whether the vaccination is provided
continuously shall be evaluated. But it is not regular monitoring. There
are dropouts and gaps. Even though a monthly program is planned,
we sometime conduct monitoring every other month.
“
HC Head, Oromia
There is a monthly monitoring and evaluation at the Health Center

level. But it was not run regularly. There are interruptions…There is also
an expectation of support and supervision from the Health Center,
but, due to the high numbers of patients in the Health Center, it is
not enough, and there are interruptions.
HEP supervisor, Tigray
page-638
MECHANISMS OF MONITORING AND EVALUATING THE HEP
According to participants, multiple monitoring and evaluation system are

commonly reported. These include integrated supportive supervisions (ISS),
review meetings, supervisions, checking the report against the plan, direct
observation of the work done, and assessment of community satisfaction with
the HEP service.
Assessing service-user satisfaction is one mechanism used to evaluate the

effectiveness of the program. Participants also stated that they got the
information from their field supervisors and community members by preparing
“
a discussion forum.
The quality needs assessment. We collect service-satisfaction forms

from supervisors. We get some clues from that.
HC head, Addis Ababa
“
A key informant from one woreda also said that:
There are also events in which people from the regional and federal
levels will come for a monitoring visit and the collection of information
from the Woreda Health Office, including observing some physical
facilities at the community level (e.g., like latrine coverage). They will
come with their own plan and go into the field with the woreda staff
to cross-check those reported achievements and performances.
Program Officer, Harar
page- 639
Participants also stated other methods they were using for M & E were direct
observation of randomly selected households and evaluation of the work done
“
using a checklist.
We do research and assessment on whether the community or

society really benefits from the collective health program. When
we do supervision at the lower levels, we see 10 households based
on the lottery method for door-to-door services, and we do actual
observations in households. We have a checklist of what the patients
have and how they are handling it, whether the children have been
vaccinated, and agendas of this sort. We also have conferences, even
though they are small in number. We also check on the professionals
working at the lower levels, whether they are doing their jobs
effectively.
RHB Head
Participants also mentioned evaluating the reported activities done and their
outcomes against the baseline plan and conducting a review meeting with
different stakeholders. One UHEP team leader at the sub-city level said that:
“
We [the sub-city] prepare the plan together, and each department
presents it to the group. When we call a meeting, all the Health
Center workers and woreda health sector workers come together.
Then we evaluate both the plan and the report.
UHEP team leader, Addis Ababa
page-640
One participant from another sub-city in Addis Ababa described the review
“
meeting thus:
The standard suggests that we have quarterly review meetings. The

Addis Ababa Health Bureau organizes and calls us sometimes. We
also conduct quarterly review meetings in our sub-city. Before the
review meetings [organized by AAHB], we go out for mentorship
and evaluate our performance based on checklists and identify our
challenges and strengths in relation to the Program and the Family
Health Team. The tool has many measurements, like timeliness,
quality, and whether activities are implemented correctly. And we
invite the Health Centers, stakeholders, and other implementing
bodies to evaluate and assess the strengths and weaknesses.
Participants also reported various challenges and limitations related to

monitoring and evaluation. A lack of output-based evaluation, a failure to pay
equal attention to all HEP activities, and a failure to follow the set criteria are
mentioned as limitations of the monitoring and evaluation of the HEP.
“
Focus is not given to the tasks in as much detail as it should be. Your
evaluation and supervision of tasks should be results-based, should
be counted and assigned.…This is not how things are done here. The
only thing that is given enough emphasis is the graduation of model
families. That task also has its own trend, but when you are dealing
with people with limited knowledge of the program, you will be forced
to follow the path others created for you without proper feedback.
One key informant from a Regional Health Bureau also reported that the
lack of guidelines to accomplish the tasks of supervision and monitoring and
evaluation at different levels were a challenge:
page- 641
“
UHEPrs don’t follow the 1-to-5 networks leaders. As well, the leaders
don’t follow and support the regular members. HCs are not following
whether UHEPrs are working as planned or not. Other HC support
professionals don’t check which of the pregnant mothers give birth
and who are left…we do the same. Regarding FP, they [UHEPrs]
are not even properly filling new or repeat [clients]...If a woman is
recorded as a repeat, it will continue in that way. That is the reason
there is a problem during data verification.
The lack of sustainability and follow-up in monitoring and evaluation were also
“
concerns of the participants.
The major issue with the review meeting is the lack of follow-up.
For example, the implementation of the action plan set during the
meeting is not followed up on, mostly because the implementors will
be occupied with other activities. Therefore, we will keep on raising
the problems at the next meeting, and the trend continues.
Expert, HC, Addis Ababa
The findings of the quantitative study on the HC level indicated that a high
proportion (89.5%) of HCs reported that HCs conducted supportive supervision
of UHEPrs, and the majority (82.9%) of the city or town HOs and HCs used
standardized supportive supervision checklists and field supervision to ascertain
whether UHEPrs were truly supervised. Similarly, 92.1% of respondents replied
that they provided written feedback to the UHEPrs themselves to ensure their
supervision (Figure 5-7).
page-642
Figure 5-7: Supportive Supervision System and Practice by HCs on the

Implementation of UHEP
PERFORMANCE EVALUATION
Regarding the UHEPrs’ performance evaluation, most participants from HCs

reported that it was done daily by UHEPr supervisors & weekly at the HC.
Most participants from WorHOs, sub-cities, and RHBs mentioned monthly and
quarterly evaluation.
Regarding the method of evaluation, most participants from UHEPrs, HEP

coordinators, FHT coordinators and WorHO reported that the performanceand
the achievement of every package was checked against randomly selected
households using direct observation and interviews of community members and
specific service users at the clinic about the service they had received from
“
UHEPrs, and their satisfaction with it.
We get feedback about the quality of services from the community.

Moreover, we check the quality through the evaluation of services on
a weekly basis…Health Extension Professionals should meet pregnant
mother twice a week. We randomly ask pregnant mothers whether
the UHEPr has met them or not. If she has visited them, what kinds of
services were they given.…So, we will check them against the reports
the UHEPrs have made to the Center.
page- 643
In addition, participants stated that, at the HC level, there was performance

evaluation in the form of reviewing their plan, achievement, and existing gaps,
and that this evaluation was used to address the challenges through supportive
“
supervision.
There is a monthly meeting at the Health Center level to review

the performance of all UHEPrs in the presence of their supervisor.
They review each plan, performance, achievement, and existing gaps.
Finally, an action plan will be developed to address the identified
gaps through supportive supervision.
WorHO, Harari
Daily work registration and follow-up by supervisors, weekly meeting with

supervisors, and monthly performance evaluation with all the HC staff were
“
also reported.
Once a month, a group that consists of the Health Extension

supervisors, Health Office, medical director, unit leader, and
supervisors, evaluates the program. Monthly performance looks
like…when we evaluate monthly performance, and identify gaps.…
The second is that, weekly, there is a meeting held by the Health
Extension Workers and their supervisors in which they evaluate their
weekly performance.…The third is that supervisors made Extension
Workers record their daily tasks/performance.
The quantitative finding largely supports the above results, with almost all
(98.4%) respondents reporting that UHEP activities were integrated with the
HC annual work plan. The vast majority (96.2%) of respondents indicated the
presence of functional performance management or evaluation systems for the
performance evaluation of UHEPrs by the head of the HC. Most respondents
reported that supportive supervision, review meetings, and feedback were used
to discuss the performance of UHEPrs and their supervisors (Table 5-14).
page-644
Table 5-14: Planning and Performance Monitoring for UHEP at HC Level

N % N % N % N %
Integration of UHEP activities in HC plan
Yes 91 98.9 3 60.0 30 96.8 124 98.4
No 1 1.1 2 40.0 1 3.2 4 3.2
Regular performance review meeting with UHEPrs
Yes 88 95.7 7 100.0 30 93.8 125 95.4
No 4 4.3 0 0.0 2 6.3 6 4.6
Availability of a functional performance review team at HC
Available 90 97.8 7 100.0 30 90.9 127 96.2
Not available 2 2.2 0 0.0 3 9.1 5 3.8
Review of UHEP performance by the HC performance review team
Yes, verified 61 66.3 0 0.0 16 53.3 77 60.6
Yes, not verified 29 31.5 7 100.0 11 36.7 47 37.0
No 0.0 0 0.0 3 10.0 3 2.4
Communication of performance review feedback to UHEPrs
Yes, verified 48 52.2 1 14.3 11 36.7 60 47.2
Yes, not verified 36 39.1 6 85.7 12 40.0 54 42.5
No 6 6.5 0 0.0 7 23.3 13 10.2
Documentation of performance monitoring decisions involving UHEPrs
Yes, verified 27 29.3 0 0.0 9 30.0 36 28.3
Yes, not verified 21 22.8 6 85.7 10 33.3 37 29.1
No 42 45.7 1 14.3 11 36.7 54 42.5
USE OF THE FINDINGS/DATA USE
Use of the findings collected through supportive supervision, monitoring and

evaluation, performance evaluation, reports and resources, and lesson learned
from strengths and weaknesses and other means are crucial to making
reasonable decisions concerning planning, implementation, resource allocation,
the formulation of policy, and strategies to strengthen the HEP.
page- 645
In this study, participants stated the findings generated by different means

were used to take corrective measures for further improvement of HEP
services through preparing action plans, giving feedback to Health Extension
professionals, formulating solutions, and sharing the findings with stakeholders
and experiences with each other. One key informant from the Amhara region
“
described how the findings were used for corrective measure:
The [monitoring and evaluation] team meets at the end of the day
and provides them [UHEPrs] with feedback. They [UHEPrs and
the evaluation team] jointly plan how to address gaps and shares
responsibility for the actions to be taken, which is specific in terms
of what to do, when, and who. During the next monitoring visit, they
always start by checking whether the previously identified gaps are
corrected or not. If it has not been corrected, they discuss the problem,
and sometimes it comes to management for a decision.
HC Head, Amhara
UHEPrs from pastoralist areas (regional and woreda towns) raised concerns,
however, that they did not receive feedback based on the evaluation,
“
neither from the HO nor HC.
We have never heard feedback, either from the Office or the Health
Center. That is because of the problems I told you about earlier.
UHEPr, Semera Logia
Regarding the use of findings for planning, participants reported that they
“
began planning by reviewing their previous performance involving stakeholders:
When we prepare the plan, our starting point is the previous year’s
performance report. It shows us the setbacks in the main indicators,
and we can pinpoint exactly where more support is needed.…The
planning involves people from the sub-city, administration heads, and
the UHEPrs. We also invite sectors that are directly related to us….
UHEP coordinator, Sub-City Health Office, Addis Ababa
page-646
Findings are also used as opportunities for learning and experience sharing
“
and to plan accordingly for both the short and long term.
So we will do this in every quarter, depending on the achievements.

We will review the achievements 4 times a year. After the review,
we will say that this cluster achievement is good or poor.…So, we will
show the kebele who had the better performance….So we took it as a
good experience, and we shared it with others.…We will also classify
the gaps by planning to solve them in the short term and long term.
Program Officer, WorHO, Benishangul Gumuz)
In addition, these participants reported that the findings were also used for
“
resource allocation and hiring UHEPrs:-
The Health Extension professionals register the households in their

catchments to categorize them and give service. We got the number
of the households based on those registrations, then, when we
calculated the 1-to-500 ratios, we found the gap in the number of
UHEPrs and then requested additional UHEPrs from the civil service
to feel the gap. We requested 60 UHEPrs, but we were allowed
to hire 31. So we used the information for decision-making on the
Bureau level.
Some participants, however, indicated that there was a problem using the data
for decision-making because of poor reporting, not implementing findings after
“
a report, or a lack of commitment and capability from the concerned bodies:
We do fieldwork to identify problems. Then we report to the decision-

makers at each level for councils and the committees. Some of the
reports may not be sent to the top-level structures. Sometimes they
may be reported, but the job is here, not implemented….
UHEP team leader, Sub-City Health Office, Addis Ababa
page- 647
“
We do data analyses and send them to the concerned bodies, but the
use of these data for implementation is very weak. It may be due to
a lack of commitment, capability, or other factors.
RHB Head
5.9 Challenges to the Implementation of the UHEP
In general, the study assessed challenges related to the implementation of the

UHEP. Among the challenges repeatedly mentioned were: human-resources
development (motivation, satisfaction, incentive, and career development;
80%), community engagement, awareness, and acceptance (80.9%), supplies,
equipment and other infrastructure (72.2%), and collaboration with other
stakeholders and urban development (unplanned and rapid urbanization and
urban dynamics; 69.2%; Figure 5-8)
Figure 5-8: Major Challenges of UHEP Implementation Reported by Heads of

HCs
The qualitative study also assessed the major challenges that hamper the
implementation of the program. Several factors were reported to be challenging
the implementation of the UHEP. These challenges can be broadly categorized
into limited commitment by and demotivation of the UHEPrs, workload, limited
community acceptability of the program, inadequacy of the resources needed
for the program, limited stakeholder collaboration, package comprehensiveness,
and weak leadership and management problems.
page-648
5.9.1 Limited UHEPr commitment and motivation
Participants frequently mentioned low commitment and motivation among the

UHEPrs. The participants reported that the UHEPrs were less motivated to
discharge their duties. Several factors were mentioned as possibly contributing
to the UHEPrs’ lack of commitment, including those related to educational
and career development opportunities and UHEPrs’ dissatisfaction with their
current salaries.
5.9.2 High Workload of UHEPrs
The UHEPrs were reported to spend much of their time working on non-
health activities, which was believed to negatively affect the UHEP. Due to
the UHEPrs’ limited number, high workload, and large geographic area that
they are intended to serve, the UHEPrs fail to provide UHEP activities through
home visits. The participants reported that the UHEPrs have difficulty reaching
all the households in their catchments.
Furthermore, the UHEPrs reported that they were involved in and assigned
several activities other than their routine activities. This working condition was
reported to have made it difficult to work as a UHEPr and forced them to
leave.
On the other hand, there were also reports that there were UHEPrs who were
usually absent from their workplace. As a result, clients cannot get services. The
UHEPrs were reported to be less committed to providing health education. A
few participants also reported that there were incompetent UHEPrs as a result
“
of the weak selection process.
Currently, there are places where HPs are closed, there are places
where UHEPrs go to their jobs at 10 AM, there are places where
UHEPrs don’t work in the afternoons, but there are also places where
they work properly. In short, there is variability in this regard.
page- 649
5.9.3 Limited awareness and community acceptability of the UHEP

and UHEPrs
Limited community acceptance of the UHEP and the UHEPrs were also
mentioned as challenges to UHEP implementation. One important challenge
to UHEP implementation is the politicization of the program. There is a
perception that the UHEPrs are deployed to undertake political activities. The
1-to-5 structure and WDA are considered political tools. Thus, the community
fails to participate in meetings regarding the UHEP. This finding indicates that
the awareness of the community regarding the UHEP and the packages is
limited and needs action to be taken.
“
Sometimes our society is really hard to deal with. When you are on
your duties, you knock on the door, and they open the door and say,
“It is you again? You came to nag us more?” They close the door in
your face. This has happened to me. It really made me hate my job.
“
HEP supervisor, Amhara
The community has a perception that the HEP is related to political

issues. It is even challenging to organize discussion with the Health
Development Army groups. There is a perception that the HDA
structure has been dismantled due to the recent political reforms in
the country.
HC Head, Amhara
There is also a perception that the UHEPrs are less educated and that there
is not much to learn from them. The roles and responsibilities of the UHEPrs
are not clear to the community. There is a perception that the UHEPrs are
responsible just for collecting waste.
The level of community awareness, cultural beliefs, religious beliefs, and stigmas
related to some diseases were reported to be among the bottlenecks in the
uptake of the UHEP, including FP, institutional delivery, and HIV testing. For
page-650
instance, some women prefer to give birth at home due to cultural beliefs and
pressure from traditional birth attendants.
5.9.4 Limited inter-sectoral collaboration
The limited communication and integration among the Health Office, Water
and Sewage Office, municipalities, schools, and other relevant offices were
“
reported to negatively influence the implementation of the UHEP.
We said it earlier. It needs an integrated approach of working. For

example, in the community health control, the responsibility for
the water and drainage should be with Water and Drainage. For
garbage disposal, there should be an office for that. People who are
disposing garbage in places they are not allowed to should be asked
by the controllers. In creating awareness, the Health Extension should
take the responsibility. There should be coordination; the work should
not be left only to UHEPrs.
HC Head, Assosa, Benishangul Gumuz
5.9.5 Limited or shortage of resources

Most HPs and HCs were commonly reported to have an inadequate budget,
a shortage of drugs and other supplies, and limited facility infrastructure. The
participants mentioned the shortage of ambulances for patient referral and
linkage. More importantly, some HCs were reported to be physically inaccessible
and lack the equipment needed to provide health services. Transportation
was mentioned as one of the most important challenges to the
“
implementation of UHEP activities.
There are many problems. We can’t say there aren’t. The basic
problem is that there is no budget set for the Program. Because
there are no vehicles to allow HEPrs move from place to place as
they want. There was a bicycle given, but when it first arrived in the
region, it was nonfunctional. Because it had no quality.
Program officer, Assosa, Benishangul Gumuz)
page- 651
5.9.6 Incomprehensiveness of the UHEP package

The participants argued that the existing UHEP did not take the heterogeneity of
the urban population into consideration. The urban population is heterogeneous
and has different health needs. The UHEPs, however, were reported to be
designed in a way that does not accommodate the different community health
problems and needs in urban settings. The participants argued that the HEP
“
was not responsive to the changing community needs and health problems.
I personally don’t believe that the strategies are contextualized

in the case of implementation. Because the population dynamism
itself is a deciding factor in the success of the program. Contextual
understanding of the site is necessary due to differences in lifestyles
and culture even within a region. What we have now is a uniform and
highly centralized system, which is not so good at getting the needed
results.
“
RHB Head
The community does not get what they want to get because of the
shortages of the health package. Thus, if there is a shortage, there
is inaccessibility to the community. That is to say, the community
members will not obtain enough health education. Even if the
community members get education, what they get may not equate
to what they actually want.
“
The same is true in the UHEP; the UHEPrs came with agendas below
the expectations of the residents. Hypertension and diabetic cases
are higher in towns, and they expect to have their blood pressure or
blood sugar measured, while the UHEPrs are not ready in this regard.
FGD participant, other program officer, Tigray)
page-652
WEAK SUPERVISION AND MONITORING/LEADERSHIP AND

MANAGEMENT
The HEP’s supervision, monitoring, and support mechanisms were reported

to be low. Kebele administrations, Regional Health Bureaus, District Health
Offices, and HCs usually fail to provide technical support to the UHEPrs. The
supervisors were described as less competent and lacking capacity-building
training. Moreover, it was reported that supervisors were not given additional
incentives or compensations and that they faced transportation challenges.
Health officials’ commitment to the implementation of the HEP was also
reported as declining over time.
“
For the effectiveness of the M&E implementation, the transportation
and meal expenses of the monitoring team should be covered. Even
if we gave motorcycles for transportation, most of the staff does not
have a driver’s license. When you tell people to pay out of pocket,
they are not motivated to work. If we get budget support, the M&E
will be stronger.
HC Head, Amhara
page- 653
CHAPTER 6
Conclusions and
Recommendations
page-654
CHAPTER 6
6 Conclusion and Recommendations
6.1 Conclusions
The relevance of the Urban Program varies among its packages, despite all
UHEP packages’ having contributed to reducing the burden of CDs, maternal
and child morbidity, and mortality. The health education provided at the
household and community levels was found to be relevant in improving health
literacy, which in turn was reported to influence the health-seeking behavior
and health service use by the urban poor.
The interventions of the WaSH package were not adequate in addressing the
underlying causes of the water-, hygiene-, and sanitation-related problems of
the urban community. A lack of latrines, shortage of water treatment supplies,
shortage of a clean water supply, and poor waste-disposal systems were the
major bottlenecks and problems of the urban community. Those bottlenecks,
which hindered the implementation of WaSH through the UHEP, were beyond
the scope of the MoH or UHEP.
There is inadequate multi-sectoral collaboration among the governmental

organizations working on WaSH. As a result, although the UHEP was relevant
in creating awareness through the health education model families and home
visits, WaSH-related problems remained challenges to the urban community.
Moreover, UHEPrs lack the capacity and skills to handle sectoral collaboration,
particularly for the implementation of WaSH interventions.
Though the family health packages were found to be relevant to tackling the
priority health problems of the urban community, the interventions could not
meet the needs and expectations of the community. To meet the existing needs,
the family health packages should include some clinical services in addition
to the health education and referral services. Similarly, the disease prevention
and control packages, particularly NCD screening and follow-up services, were
found to be relevant, but poorly implemented, packages of the UHEP.
Inadequate implementation of the packages at the community level, shortage

of resources (both human and material) and the multifaceted nature of health
page- 655
Conclusion
problems in urban settings greatly limited the relevance of UHEP to addressing

the health needs of the urban community. More importantly, the absence of
strong integration, harmonization, and alignment among the various sectors
working on urban health-related activities greatly affected the implementation
of the program, particularly WaSH interventions.
The UHEP is being implemented through a combination of modalities and

approaches: mainly the model family, Women’s Development Army, home-to-
home visits, and the Family Health Team approach. Implementation of the
program using the above modalities greatly varies across the study areas,
where Dire Dawa and Addis Ababa were predominantly using the Family
Health Team approach.
The health information or education disseminated by UHEPrs through home-to-

home visits was inefficient and less relevant for the urban employed community
(including the self-employed, those working in small businesses, and government
employees), due to the overlap in working hours. Moreover, the current study
revealed that about half of the study participants, who are the urban poor and
the primary targets of the UHEP, had not been visited in the past year or never
visited at all. The findings of this study also indicated that the frequency and
coverage of home-to-home visits by UHEPrs have been declining over time,
mainly due to their reduced acceptance by the community.
The model family training is still one of the most effective strategies to implement
the UHEP packages and bring the desired healthy behaviors and outcomes to
households. A small proportion (less than a quarter) of study participants in
Addis Ababa and Dire Dawa, however, reported that at least one of their
household members had received and completed the model family training.
The collaboration of UHEPrs with HDA leaders was a major facilitator of the
implementation of the UHEP. In most cases, the HDAs play a significant role
by serving as bridges between the UHEPrs and households. Nevertheless, the
non-functionality of the HDAs in most urban areas and the lack of community
acceptance due to their perceived political affiliation limited their contribution
to the implementation of UHEP packages at the household level.
page-656
Conclusion
The FHT approach was the most effective UHEP implementation modality
through which the clinical service needs of the urban poor were being
addressed, which was the major challenge to the acceptance of the UHEP by
the community. The approach improved the integration of the UHEPrs and
health professionals working at the HC level and resulted in improved
motivation and confidence among UHEPrs. Implementation of the approach is
suffering from challenges related to the inadequacy of the health workforce to
organize the FHTs as per the guidelines and the shortages of drugs, medical
supplies, and equipment.
The UHEP workforce analysis showed an unacceptably higher level of job

dissatisfaction, demotivation, and intention to leave among UHEPrs. Their
high workload, low community acceptance, and low salary, the absence of a
recognition, motivation and incentive scheme, the absence of career development
opportunities, and the inadequate opportunity to advance their education were
major causes of UHEPrs’ dissatisfaction and demotivation.
The inadequate number of UHEPrs was the major reason for insufficient
implementation of the program, mainly in the low proportion of household
visits. Almost half of UHEPrs were assigned to serve more than the maximum
number of households (500) specified in the guidelines.
The shortage of medical supplies and equipment is another major bottleneck in

the implementation of the UHEP, which results in poor implementation of the
program, demotivation of the UHEPrs, and decreased community acceptance
of the program and UHEPrs. A majority of UHEPrs lack a weight and height
scale to use in conducting nutrition screenings and growth monitoring of children.
The UHEP faces multiple challenges related to the inadequacy of resources,

poor intersectoral collaboration, lower community engagement and acceptance,
and demotivated workforce (i.e., UHEPrs). The above challenges and the
complex and dynamic nature of urban health problems have greatly affected
the implementation of the program in urban areas.
page- 657
page-658
6.2 Recommendations
• All UHEP packages should be strengthened and continued, and
additional clinical (screening, examination, diagnosis, and follow-up)
services should be added to the maternal and child health and NCD
packages:
• Packages should be revised to meet the dynamic needs and

expectations of the urban community;
• Contextualized and tailored interventions should be designed by
disaggregating urban areas into different categories (i.e., city
administrations and regional capitals, zonal towns, and small
towns).
• Community midwifery needs to be introduced to advance and enhance

the maternal health service package as a backstop in the provision of
clinical services that require midwifery skills.
• Implementation of the FHT approach should be strengthened by
addressing the challenges of the availability of supplies (drugs, medical
supplies, and equipment) and the commitment and motivation of team
members:
• The supply chain management system should be strengthened;

• Performance-based motivational schemes for FHTs should be
designed and implemented;
• The FHT monitoring and evaluation system should be strengthened
to ensure its accountability.
• The FHT approach shall involve other non-health professionals, such as

social workers, psychiatrists, economists (to work on income-generating
activities), and sanitation engineers to address the root causes affecting
the health and well-being of the urban poor.
• Multi-sectoral collaboration needs to be strengthened for successful
implementation of UHEP through the integrated planning,
implementation, and monitoring and evaluation of WaSH interventions
at all levels of the system.
page- 659
Recommendation
• A new structure should be established at the woreda, sub-city, or town

level to effectively coordinate and liaise WaSH interventions at the
woreda and higher levels by deploying experts with sanitary engineering
competency.
• It is essential to establish a platform to link the WaSH interventions
with TVETs and institutions of higher learning to promote innovative
technologies that can solve the bottleneck of the operation with locally
available resources and technology options:
• There should be innovation regarding locally applicable

technologies for WaSH, particularly to make latrines available to
the poor community living in slum areas (linking with TVET and
other technology institutes);
• The construction of WaSH infrastructures at the household level
should be subsidized (mainly for latrine construction).
• A high-level integration of urban health policy should be considered,

and a clear roadmap with monitoring, evaluation and accountability
system should be put in place.
• Guidelines and manuals should be developed for the effective
implementation of UHEP service delivery modalities and approaches:
home-to-home visits, HDAs, model families, and the model woreda
initiative.
• It is also recommended that human resources (i.e., UHEPrs and FHT
members) be fulfilled as per the standards specified in the UHEP
implementation manual and FHT implementation guide. The number of
UHEPrs should be increased based on the existing number of households
in the catchment areas, and all households should be covered by the
program:
• Proper human resource planning at all levels should be designed

and implemented;
• A retention mechanism to reduce the turnover of UHEPrs should
be designed and implemented.
• It is thought to be important to enhance UHEPrs’ motivation by designing

and implementing performance-based motivation schemes with financial
and non-financial incentive mechanisms.
page-660
Recommendation
• Pre-service and in-service trainings need to be strengthened to enhance

the knowledge and skill of UHEPrs, particularly of new recruits.
• It is vital to promote the UHEP through different electronic and print
media to create awareness of and improve the acceptance of the
program by the urban community.:
• There should be a collaboration with mass media to broadcast

audio-visual programs focusing on the UHEP packages;
• IEC/BCC materials should be produced to improve community
access to health information and education;
• A UHEP call center should be established from which the urban
community can get information on any health and related issues
at their convenience, which can also serve as a reference or source
of information for UHEPrs;
• Influential people who are considered popular and role models by
the community should be collaborated with to help promote the
packages.
• There is a need to ensure the availability of all necessary materials and

equipment for the implementation of the UHEP:
• A national-level inventory should be conducted, and the required

materials and equipment should be quantified;
• The supply chain management system should be strengthened to
ensure the availability and periodic refill of commodities for the
UHEP;
• Periodic assessments should be conducted to check the functionality
of the supply chain management system and the proper use of
commodities supplied for UHEP.
• Last but not least, it is deemed necessary to revise and strengthen the
community networks, which can serve as primary collaborators and
contact points for UHEPrs using locally acceptable and valued social
structures like Idir and mahiber.
page- 661
page-662
References
1. OECD-DAC Network on Development Evaluation, Better Criteria for Better

Evaluation: Revised Evaluation Criteria Definitions and Principles for Use.
2019. p. 12.
2. Awoke Misganaw, Tilahun N. Haregu, Kebede Deribe, et al. National

mortality burden due to communicable, non-communicable, and other
diseases in Ethiopia, 1990–2015: findings from the Global Burden of Disease
Study 2015. Population Health Metrics. 2017;15(29):1-16.
3. Fassil Shiferaw, Mekitew Letebo, Yeweyenhareg Feleke, Terefe Gelibo. Non-

communicable Diseases in Ethiopia: Disease burden, gaps in health care
delivery and strategic directions. Ethiopian Journal of Health Development.
2018;32(3).
4. Hirbo Shore Roba , Addisu Shunu Beyene, Melkamu Merid Mengesha , and
Behailu Hawulte Ayele. Prevalence of Hypertension and Associated Factors
in Dire Dawa City, Eastern Ethiopia: A Community-Based Cross-Sectional
Study. International journal of hypertension. 2019;volume 2019:https://doi.
org/10.1155/2019/9878437.
5. World Health Organization. Global health risks: mortality and burden

of disease attributable to selected major risks. Geneva: World Health
Organization 2019. In:2009.
6. AI Mohammeda, and Li Zungu. Environmental health factors associated
with diarrhoeal diseases among under-five children in the Sebeta town of
Ethiopia. Southern African Journal of Infectious Diseases. 2016;31(4):122–
129.
7. World Bank Group, Cities Alliance. Ethiopia Urbanization Review: Urban

Institutions for a Middle-Income Ethiopia In.
8. Abebe Beyene, Tamene Hailu, Kebede Faris, and Helmut Kloos. Current
state and trends of access to sanitation in Ethiopia and the need to revise
indicators to monitor progress in the Post-2015 era. BMC public health.
2015;15:451.
page- 663
References
9. WHO. Global Tuberculosis Report 2017. World Health Organization

(WHO), Geneva. In:2017.
10. WHO. From MDG to SDG. World Health Organization(WHO). (http://apps.

who.int/iris/bitstream/10665/200009/1/9789241565110_eng.pdf?ua=1). In:2015.
11. Nebiyou Tafesse, Aregawi Gesessew, and Ergataw Kidane, Urban health
extension program model housing and household visits improved the
utilization of health Services in Urban Ethiopia: a community-based cross-
sectional study. BMC Health Services Research, 2019. 19(31): p. https://doi.
org/10.1186/s12913-019-3868-9.
12. R Burns, L O Nichols, M J Graney, and W B Applegate. Mortality in a

public and a private hospital compared: the severity of antecedent disorders
in Medicare patients. Am J Public Health. 1993;83(7):966-971.
13. World Health Organization. African Health Observatory: Analytical

summary - HIV/AIDS. In.
14. Mirgissa Kaba, Girma Taye, Muluken Gizaw, Israel Mitiku, Zelalem Adugna,
Addis Tesfaye. A qualitative study of vulnerability to HIV infection: Places
and persons in urban settings of Ethiopia. Ethiop J Health Dev. 2016;30(3).
page-664
page- 665
PART 4
In-depth Exploration of Specific
Topics
Study I: Assessment of the quality of Health Extension

Workers’ Training institutions
Study II: Attrition among Health Extension Workers
Study III: The role of the Health Extension Program in

Public Health Emergency Management
Study IV: Cost-Effectiveness Analysis of the Health

Extension Program
page-666
Specific Study 1:
Assessment of the Quality of
HEWs’ Training Programs
page- 667
CONTENTS
Contents --------------------------------------------------------668
List of Tables -------------------------------------------------------------------671
List of Figures ----------------------------------------------------------------672
1 INTRODUCTION ---------------------------------------675
1.1 Background -------------------------------------------------------675
1.2 Objectives of the Assessment --------------------------677
1.3 Conceptual Framework --------------------------------------677
2 METHODS -------------------------------------------------------681
2.1 Study Design --------------------------------------------------681
2.2 Study Sites 681
2.3 Sample and Sampling Techniques ----------------------------681
2.3.1 Quantitative sampling --------------------------------------681
2.3.2 Qualitative sampling ---------------------------------------------681
2.4 Data-Collection Tools ----------------------------------------------682
2.4.1 Questionnaires -----------------------------------------------------683
2.4.2 Interviews --------------------------------------------------------------683
2.4.3 Institutional Profile -------------------------------------------------684
2.5 Procedure ---------------------------------------------------------684
2.6 Data Analysis ---------------------------------------------------684
page-668
2.6.1 Quantitative analysis ----------------------------------------684
2.6.2 Qualitative analysis -----------------------------------------------685
2.7 Ethical Considerations --------------------------------------------685
3 RESULTS AND DISCUSSION -------------------------------------------687
3.1 Relevance of the HEP curricula --------------------------------688
3.1.1 Relationship of the HEP curricula to the country’s

broader health policy ------------------------------------------------688
3.1.2 Relationship of HEP curricula to expectations for HEWs’

community-based tasks after graduation ---------------------689
3.1.3 Consistency between HEP curricula
and task implementation -------------------------------------------692
3.1.4 Balance between theoretical and practical HEP

curricula components -------------------------------------------695
3.2 Adequacy of Resources in HEW Training Institutions ----------697
3.2.1 Instructors/trainers --------------------------------------------697
3.2.2 Facilities -----------------------------------------------700
3.2.3 Teaching modules -----------------------------------------------------706
3.3 Quality of Course Materials ---------------------------------708
3.4 Quality of Course Delivery and Assessment Methods --709
3.4.1 Teaching Modality ----------------------------------------------------711
3.4.2 Professional Competence of Course Instructors ------------711
3.4.3 Student Recruitment -----------------------------------------------714
page- 669
3.4.4 Teaching Methods ---------------------------------------------717
3.4.5 Effectiveness of the Teaching Methods Employed ----------719
3.4.6 English as the Language of Instruction -------------------------720
3.4.7 Assessment of Learning and its Effectiveness --------------------722
3.5 Competence of HEP Trainees -----------------------------------724
3.5.1 Instructors’ Evaluation of Trainees’ Competence ----------724
3.6 The Existence and Adequacy of Professional Development

Activities for HEWs -------------------------------------------729
3.6.1 In-Service Training ---------------------------------------------729
3.6.2 Continuous Professional Development (CPD) -----------------733
3.6.3 Concerns about the Professional Development

of HEP Instructors ----------------------------------------------------734
3.7 Challenges to and Opportunities in the HEW

Training Program ------------------------------------------------------735
3.7.1 Opportunities ---------------------------------------------------736
3.7.2 Challenges ------------------------------------------------------737
4 SUMMARY AND CONCLUSION ---------------------------------741
4.1 Summary of Major Findings -------------------------------------741
4.1.1 Relevance of the curricula -------------------------------------------741
4.1.2 Adequacy and Availability of Resources -----------------742
4.1.3 Quality of course materials (curricula, modules, references)

for HEWs training ----------------------------------------------------742
page-670
4.1.4 Quality of course delivery and student
assessment methods -----------------------------------------743
4.1.5 Competence of graduates from HEW training institutions ---743
4.1.6 Professional development activities -------------------------744
4.2 Conclusions --------------------------------------------------744
4.3 Recommendations ---------------------------------------------------745
APPENDIX: List of data collection tools --------------------------------747
References ----------------------------------------------748
LIST OF TABLES
Table 1: Reliability of Survey Measures -------------------------------683
Table 2: Demographic Characteristics of Quantitative

Survey Respondents ----------------------------------------------------------687
Table 3: Trainees’ and Instructors’ Ratings of HEP Curricula Relevance --689
Table 4: Instructors’ Educational Qualification and

Specialization (N=192) ----------------------------------------------------697
Table 5: Number of Instructors by College and Field of Specialization

(2018/19 Academic Year) ---------------------------------------------699
Table 6: HEP Training Facility Resources, 2019 --------------------------701
Table 7: Instructors’ and Trainees’ Evaluation of

Course Material Quality -----------------------------------------708
Table 8: HEP Trainees’ Ratings of the Instructors’

Professional Competence -------------------------------------712
page- 671
Table 9: Reasons for Joining the HEW Training Program ----------------715
Table 10: Effectiveness of Assessment Methods ----------------------723
Table 11: Trainee Competence by Program Ownership and Type ------726

Table 12: Number of attempts necessary to pass the COC exam ----------726
Table 13: HEW Participation in Integrated Refresher

Training (2007-2011 EFY) ---------------------------------731
Table 14: Frequency of HEW Participation in In-Service

Training (2007-2011 EFY) ------------------------------------------732
Table 15: Progress in Key Health Indicators, Ethiopia,

2005 and 2016 EFY ------------------------------------------------735
LIST OF FIGURES
Figure 1: Conceptual framework of the study. -------------------------678
Figure 2: Student Preference for Course Delivery Approaches ------------710
Figure 3: Educational Status of Trainees -------------------------------714
Figure 4: Frequency of Teaching Methods Used in the

HEW Training Program --------------------------------------------717
Figure 5: Student Satisfaction with Course -----------------------------719
Figure 6: Frequency of Assessment Method Use in

HEW Training Program ---------------------------------------------722
Figure 7: HEW Trainee Competence -------------------------------------725
Figure 8: HEW training program challenges ------------------------------------739
page-672
page- 673
1
page-674
1 Introduction
1.1 Background
Ethiopia launched the Health Extension Program (HEP) in 2003 to expand
the national health program to reach the poor and deliver preventive and
basic curative high-impact interventions to all of its citizens.1 The program is a
flagship of the Ethiopian Health Sector Development Program (HSDP) and
was developed in a context where health outcomes and coverage of essential
services were very poor and where there was a large disparity between rural and
urban populations, and between better-educated and less-educated people.2
HEP is a package of basic and essential promotive, preventive, and curative

health services targeting households in a community, based on the principle of
providing primary healthcare to improve families’ health status through their
full participation.1 The overall goal of the HEP is to create a healthy society
and reduce maternal and child morbidity and mortality rates.3
Presently, it is evident that Ethiopia has achieved substantial reductions in

maternal, neonatal, infant, and child mortality. This accomplishment has been
attributed in part to the country’s expansive investment in the HEP, which has
trained and deployed nearly 35 000 Health Extension Workers (HEWs) and
established 15 000 local primary health service facilities, referred to as health
posts (HPs), and approximately 2 500 health centers (HCs) since 2003.4
The core of the HEP is the production and deployment of HEWs to provide
health-related services to the community at large at the grassroots level. The
HEWs are posted to rural communities across Ethiopia, where they provide
more accessible and more equitable access to health services for the poor,
women, and children in a sustainable manner.5 The program focuses on 4
major areas and provides 16 different packages to reach the poor and address
inequities.6
The training and preparation of HEWs for deployment are carried out by
various Technical and Vocational Education and Training Institutes (TVETIs)
and health colleges under the Regional Health Bureaus (RHBs) in collaboration
page- 675
Assessment of the Quality of HEWs’ Training Program
with other concerned bodies. When possible, HEWs are recruited from the
communities in which they will work, according to specific criteria: they are
female (except in pastoralist areas), are at least 18 years old, have at least
a 10th grade education, and speak the local language. The selection is made
by a committee comprising members nominated by the local community and
representatives from the Woreda Health Office (WorHO) and the District
Education Office.7
The selected HEWs commonly go through a year-long training, which includes

both theoretical training in training institutions and practical training in HCs2.
One quarter of the training period is allocated to theoretical teaching at training
institutions, and three quarters is spent in a practicum in the community. HEWs
are trained to manage the operations of HPs, conduct home visits and outreach
services to promote preventive health activities, refer cases to HCs and follow
up on referrals, identify, train, and collaborate with voluntary community health
workers, and provide reports to DHOs.6 Upon graduation, HEWs are usually
assigned to their home villages to provide health services.
In Ethiopia, a low-income country, this innovative practice has made impressive

progress in improving health outcomes with limited resources. Concerns linger,
however, regarding HEWs’ skills, knowledge, and competence to deliver quality
services. An assessment by the Center for National Health Development in
Ethiopia revealed that some HEWs lack the critical knowledge and skills
needed to effectively discharge their professional duties in the community.1
Yayehyirad and colleagues have also reported similar findings and identified
poor procedures for recruiting HEWs and inadequate resources in training
institutions as factors that compromise the quality of HEWs’ training8.
Similarly, a number of studies have consistently revealed concerns about the
competence of HEWs in delivering health services effectively, mainly due to a
lack of knowledge and skills.2,9,10 These suggest the need for the assessment and
evaluation of HEW training programs so as to identify gaps for improvement
and modification.
page-676
1.2 Objectives of the Assessment

The objectives of the assessment of HEWs training programs are to determine:
1. The relevance of HEW training curricula to the needs of the HEP

2. The adequacy of resources (e.g., faculty, facilities, labs, supplies) in
training institutions
3. The quality of course materials (e.g., curriculum, modules, references) for
HEW trainings
4. The quality of course delivery (or teaching method) and student
assessment methods in HEW training programs
5. The competence of graduates emerging from HEW training institutions
6. The existence and adequacy of HEWs’ professional development
activities
7. The challenges to and facilitators of training HEWs in Ethiopia
1.3 Conceptual Framework
This assessment is informed by a conceptual framework prepared in the form

of a logic model showing the cause–effect relationship among 4 domains of
measurement related to HEW training programs: 1) situation, 2) inputs, 3)
process, and 4) outcomes.
As indicated in Figure 1, the situational issues of low coverage of health

intervention, low access to essential health services, and shortages of skilled
health workers prior to 2003 created the need to launch the HEP in Ethiopia as a
strategy to realize the HSDP. The recruitment and production of trained human
resources are at the heart of HEP; this has been carried out by various health
colleges and TVETIs with the help of trained instructors and the provision of
resources as inputs. The teaching–learning processes, teaching methodologies,
and assessment processes and feedback were the main activities in the training
processes. This training was intended to produce outcomes of knowledgeable,
skilled, and competent health workers who could deliver quality primary-
care services to improve families’ health status in 4 core themes: hygiene and
environmental sanitation, disease prevention and control, family health services,
and health education and communication.
page- 677
Figure 1: Conceptual framework of the study.
page-678
page- 679
page-680
2 Methods
2.1 Study Design
We used quantitative and qualitative methods to compare and triangulate

data, identify trends and similarities, and identify areas of divergence both
within and across research sites. In particular, we used a convergent parallel
mixed-methods design11 in which the qualitative and quantitative data were
gathered in parallel (or simultaneously).
2.2 Study Sites
Among the 23 colleges in the country that offer HEW training, we were able to
collect data from 21. Data were not collected from Hawassa (as students and
instructors had left the college early for unknown reasons) and Pawe Health
Science Colleges (not visited for security reasons).
2.3 Sample and Sampling Techniques
The study participants included trainees, instructors, department heads, deans,

preceptors (who supervise trainees during their internships), COC assessors,
and focal persons (officials responsible for HEP matters within organizations)
from Regional Health Bureaus (RHBs), regional TVET agencies, and the MoH,
as well as the Federal TVET Agency.
2.3.1 Sampling for Quantitative study
Trainees were selected using a stratified random sampling technique (in

which the colleges and the 2 programs, generic and upgrade, are the strata).
Attempts were made to make the sample a proportionate stratified sample. In
page- 681
some colleges with relatively few trainees, however, we selected a minimum of

32 trainees in order to make between-group comparisons meaningful, although
the sample size was not proportional to the total number of trainees in those
colleges. According to unpublished data obtained from the MoH, there were
9 478 trainees in the 22 training colleges in the country at the beginning of
academic year 2018/19. Thus, 10 000 was estimated as the population of
HEW trainees because 1 additional college was included for which we had no
enrollment data.
Assuming a 2.5% margin of error, a 95% confidence level, and a 10 000

population size, the minimum recommended sample size was 1 333 12,13. Adding
10% for non-response (n=133.3) and another 10% for subgroup analysis
(n=133.3) yielded a total sample size of 1 600. This was then divided across the
regions, colleges, and generic and upgrade programs. Even though we planned
for a sample size of 1 600, our response rate was 77.8%, resulting in a total
trainee sample of 1 245.
2.3.2 Participants of Qualitative Study
Qualitative interviews were conducted with 43 key informants (3 deans, 7

department heads, 7 instructors, 9 trainees, 7 preceptors, 6 COC assessors,
and 4 focal persons). Other than the trainees, all interviewees were selected
because their positions made them particularly knowledgeable about our key
areas of inquiry. One level IV trainee from each college was selected based
on her communication skills as assessed by the department head or instructors.
Level IV trainees were specifically selected because of their experience as level
III trainees and HEWs in the past and their current status as level IV trainees.
Unlike the other trainees, we assumed that level-IV trainees could provide data
about their previous training programs, their current training programs, and
their practical experiences as HEWs.
2.4 Data-Collection Tools
Data were collected using questionnaires, key informant interviews, observations,

report reviews, and document reviews (e.g., curriculum, teaching materials,
standards, and policies). Each tool or method is discussed in brief below.
page-682
2.4.1 Questionnaires
We employed 2 questionnaires to gather the relevant data for this assessment:

1 was filled in by HEP trainees and 1 by their instructors. Comprising mainly
closed-ended items, the surveys addressed: the relevance of the HEW training
curricula to the needs of the Health Extension Program (HEP), the adequacy
of resources, the quality of the course materials, the quality of course delivery
and student assessment methods, the competence of graduates, professional
development activities, and the challenges and opportunities of the program
(see Appendices A and B). As shown in Table 1, the reliability of the survey
measures for the trainees ranged from 0.74 to 0.95 (median=0.86) and for
instructors ranged from 0.63 to 0.96 (median=0.80). Overall, the reliability
coefficients were in the acceptable range.
Table 1: Reliability of Survey Measures
HEP Trainees (N=1245) HEP Instructors (N=192)
No. of No. of
Measure/Scale Alpha Measure/Scale Alpha
Items Items
Perceived effectiveness of Perceived effectiveness of
10 0.90 10 0.86
training program training program
Perceived quality of
Perceived quality of instructors 12 0.92 13 0.89
trainees
Satisfaction with courses’
5 0.89 Teaching methods 10 0.72
contribution
Satisfaction with teaching-
6 0.86 Assessment methods 7 0.63
learning process
Satisfaction with course delivery 7 0.79
Perceived competence of Satisfaction with teaching– 4 0.74
21 0.95 learning process
instructors
Attitude toward Health Perceived competence of
10 0.74 17 0.96
Extension Workers trainees
2.4.2 Interviews
Semi-structured interviews were conducted with key informants, including

deans of the institutions, department heads, COC assessors, preceptors, level-
page- 683
IV trainees, and focal persons at different levels. The interview items included
questions about the relevance of the curriculum, course delivery, student
assessment methods, graduates’ competence, professional development
activities, and challenges surrounding the HEP (see Appendices C–H).
2.4.3 Institutional Profile
Institution-level data were collected through a form prepared to record

information about each college. The form was completed partly by the
enumerators through observation and partly by the dean of the college or
his or her representative. We used the institutional profile form to assess the
availability and adequacy of human and material resources (faculty and
facilities in particular) in the training institutions (see Appendix I).
2.5 Procedure
Data collection at each site was performed by data collection assistants or

enumerators. The researchers offered training to all enumerators so that they
could conduct the fieldwork in an ethically acceptable manner. Accordingly,
the training focused on (1) ethical standards and how the enumerators could
meet those standards, (2) data-gathering tools, and (3) how to probe for more
information when conducting interviews, particularly in situations where the
interviewees provided only short and general responses. After securing consent
from the key informants, the interview sessions were recorded digitally. The
recordings were transcribed and translated into English by health professionals
with experience in transcription and translation. The data entry, transcription,
and translation tasks were carried out under the supervision and random checks
of senior consultants.
2.6 Data Analysis

2.6.1 Quantitative analysis
The quantitative data were analyzed using: (1) descriptive statistics (frequency,
percentage, means, and standard deviations, as applicable), (2) an independent
page-684
sample t-test and a 1-sample t test to examine between-group differences and

the attitudes of HEP trainees regarding their future jobs as HEWs, and (3)
a coefficient alpha to examine the reliability (i.e., internal consistency) of the
survey measures.
2.6.2 Qualitative analysis
The audio-recorded qualitative data were transcribed verbatim and translated

into English. After making a thorough quality check of all transcripts, data
coding and synthesis were done using NVivo 7 software. Using participants’
responses and interview guides, 346 codes were generated. After matching
codes and transcripts, redundant codes were removed, and the final codes
were reduced to 306. These nodes were then reduced to 45 treed nodes (i.e.,
categories) from which themes were developed.
The main findings were synthesized under the following themes: (1) the relevance
and scope of the HEP training curricula, (2) the student screening and enrollment
process, (3) the management and leadership of the training program, (4)
resources and infrastructure, (5) teaching and assessment methods, (6) the
competence of HEWs, and (7) the overall challenges and opportunities of the
training program.
2.7 Ethical Considerations
The assessment was conducted in a way that maintained all ethical standards
pertaining to research. The research assistants or enumerators clearly informed
all respondents about the purpose of the assessment. The assistants also told
the respondents that (1) their participation would be based on their full consent,
(2) the data would only be used for the purpose of the assessment, and (3)
the data would be used only without the name of the respondents attached to
them (i.e., under confidentiality). Overall, the respondents were informed about
their right not to participate in the study and to withdraw at any point in the
data-collection process even if they initially consented to participate. Finally, the
assistants confirmed that nobody had participated in the assessment without
his or her consent.
page- 685
page-686
3 Results and Discussion
Twenty-three training institutions offer HEP training in Ethiopia, from 21 of
which both quantitative and qualitative data were collected. For a detailed
list of the 21 training institutions, see Appendix J. Table 2 presents the survey
respondents’ demographic characteristics.
Table 2: Demographic Characteristics of Quantitative Survey Respondents
Trainees
Variable Level Frequency Percent

Male 55 4.4
Sex
Female 1190 94.6
Single 666 53.5
Married 568 45.6
Marital Status
Widowed/Divorced/
11 0.9
Separated
Upgrade 562 45.1
Program Type Generic 683 54.9
Total 1245 100.0
Instructors
Male 149 77.6
Sex
Female 43 22.4
Graduate Assistant 38 19.8
Assistant Lecturer 62 32.3
Academic Rank
Lecturer 85 44.3
Technical Assistant 7 3.6
Total 192 100.0
Variable
N Min-Max Mean SD
(Respondent)
Age (Trainees) 1245 15-45 23.19 4.53
Age (Instructors) 192 21-54 30.72 5.86
Years of Experience
192 0-25 4.58 3.91
as Instructor
page- 687
3.1 Relevance of the HEP curricula
We assessed the relevance of the HEW training curricula from 4 perspectives:

(1) the relationship of the curricula to the country’s health policy, (2) the
relationship of the curricula to expectations for HEWs’ community-based
tasks after graduation, (3) the relationship between the curricula and task
implementation, and (4) the balance between theoretical and practical
components of the HEP curricula. Each is examined below.
3.1.1 Relationship of the HEP curricula to the country’s broader

health policy
In 1993, the government published the country’s first health policy in 50 years,
articulating a vision for the development of the health sector.14 The policy’s
major themes focused on (i) democratization and decentralization of the
health system, (ii) expanding the primary healthcare system and emphasizing
preventive, promotional, and basic curative health services, and (iii) encouraging
partnerships with and the participation of the community and nongovernmental
actors.
Since 1997/1998, the country has developed an HSDP every 5 years. HSDP
II, enacted in 2003, included a strategy called the Health Extension Program
(HEP), whose major aim is to improve primary healthcare (PHC) services in
rural areas through an innovative, community-based approach that focuses
on prevention, healthy living, and basic curative care.15 The design of the HEP
packages was based on an analysis of the major disease burdens and health
crises for most of the country’s population.
It is clear from the above facts that HEP is consistent with the country’s health
policy. More importantly, in light of the fact that the HEP is part of HSDP II,
one can conclude that the HEP is relevant to the health needs of the country’s
population at large.
page-688
3.1.2 Relationship of HEP curricula to expectations for HEWs’

community-based tasks after graduation
For HEWs to discharge their responsibilities in the community after graduation,

it is critical that they develop relevant subject matter knowledge and skills
during their training. Accordingly, we assessed the relevance of the curricula
in terms of its contribution to the development of the trainees’ (1) subject
matter competence, (2) skills, and (3) ability to meet the health needs of the
community.
Using 3 survey items, we asked trainees and instructors to rate the relevance of
the curricula. Their ratings are summarized in Table 3 below. The data indicate
that the majority of both trainees and instructors favorably rated the relevance
of the curriculum in all 3 areas. With few exceptions, both trainees and instructors
perceived that the HEW training programs contributed to developing the
trainees’ subject matter knowledge and skills, as well as producing professionals
who could meet the health-related needs of the community.
Table 3: Trainees’ and Instructors’ Ratings of HEP Curricula Relevance
Trainees Instructors
Aspect of Relevance Rated Rating
N Percent N Percent
Relevance of the content/courses Good/very good 1056 84.8 136 70.8
to development of HEP trainees’ Fair 149 12.0 38 19.8
subject-matter competence Poor/very poor 40 3.2 18 9.4
Good/very good 1046 84.0 142 74.0
Relevance of the content/courses to
Fair 168 13.5 34 17.7
development of HEP trainees skills
Poor/very poor 31 2.5 16 8.3
Relevance of the courses to Good/very good 1009 81.0 146 76.0
producing professionals who can Fair 178 14.3 28 14.6
meet the needs of the community Poor/very poor 58 4.7 18 9.4
Total 1245 192
The information we obtained through key informant interviews generally

suggest that the HEP curricula are relevant, but there was disagreement
among informants. Several indicated that there were gaps or problems with
the curricula. One department head of a college had the following comment:
page- 689
“
We can’t deduce that it is fully relevant because there are some
parts that are not relevant, and there are essential parts that are not
included. For instance, if Health Extension personnel are to become
health professionals, they should learn about anatomy, physiology,
medicine administration, and injection. The curriculum is not perfect,
and its content needs modification.
Another department head stated that the trainees graduate as HEWs “with no
idea about anatomy and physiology. It would be good if these were included.”
By contrast, a key informant from the MoH argued that the training module
was competence-based and integrated; thus, the topics relevant to a given
competence are integrated into 1 module rather than presented independently.
For example, this informant indicated that the ANC module contained the
anatomy and physiology relevant to that competency. According to the key
informant, the comment that the curriculum lacks relevance is not valid in view
of the reality.
Some department heads argued that topics were redundant. For instance, a
department head explained the point as follows:
“
There are redundant parts; for instance, at level III they learn about
the prevention of communicable diseases. When they become
level IV, they learn about managing communicable diseases and
tropical diseases. So, it is repetitive. It would be nice to address such
redundancies.
Other criticisms surfaced regarding gaps, redundancy, and datedness. Some

suggested that certain courses should be added and that others should be
removed from the current curricula; one respondent suggested that “Kaizen”
should be a topic in an in-service training, while “Introduction to Computers”
and “Pharmacology” courses need to be added where they are currently not
offered. One college dean also recommended that courses in English language
and ethical education be included in the curricula. Furthermore, criticisms were
made that the curriculum was redundant for level III and level IV students. The
page-690
most frequently mentioned problem with the curricula was that the modules
are outdated.
A closer examination of the modules of the curricula for level III and level IV
does not support the criticism that the curricula are redundant. Examination of
the 2016 curriculum suggests that the modules are rather different. The level
III modules deal with health promotion (i.e., preventing disease by educating
the community), while in level IV the modules focus on managing diseases (i.e.,
rather than preventing them). Thus, there appears to be no evidence to support
the claim that some of the modules presented in level III and level IV are
redundant. There seems, to be confusion among instructors and trainees alike,
however, perhaps because the topics are the same except for the difference in
the 2 key terms, “promotion”—2 level III modules, for example, are (1) prevention
and control of common communicable diseases and (2) prevention and control of
common non-communicable diseases—and “management” —2 level IV modules,
for examples, are (1) management of common communicable diseases and (2)
management of common non-communicable disease.
It should be noted that the above argument pertains to the curricula on

paper—but what about implementation? Are the 2 curricula implemented in
a way that makes their distinction difficult? The main problem appears to be
the implementation of the curricula. Several key informants (e.g., instructors,
department heads, and trainees) were of the view that, even though the
curricula for the 2 levels are different on paper, in practice the theoretical
component dominates the training. Because the teaching of the practical
component suffers for several reasons, it is difficult to distinguish between the
theoretical components of “health promotion” and “health management”: in the
absence of good coverage of the practical component of the training, the
modules of the level III and level IV curricula would generally be similar. This
appears to be the main reason that several key informants believed that the
modules were redundant.
This finding has important implications for improving the training. If the modules
of the level III and level IV curricula are almost the same simply because the
theoretical component dominates the training, then there is an urgent need to
provide all the necessary resources (both human and material) and to cover
the practical component of the training, as stipulated in the curricula. The
curricula allocate more time to the practical component than the theoretical
page- 691
component, and this should be observed when implementing them. A further

implication of the finding concerns the HEWs’ competence. In a training where
the theoretical component receives more attention and coverage than does the
practical component, it would be naïve to expect the graduates to possess the
requisite knowledge and skills. Furthermore, if the graduates do not have the
required knowledge and skills, then cannot do their jobs effectively. Again, this
calls for urgent action that improves the training in general and the practical
component of the training in particular.
What makes HEP particularly relevant to the communities it serves is the fact
that it is designed to fit into rural communities and the cultural norms and
practices in those communities. For example, HEWs are all female except
in pastoralist areas; rural men spend most of their time farming, while rural
mothers and children are available during house-to-house visits, and female
HEWs are more accepted than males in family health-related interactions.
Another point that considered the diversities of the Ethiopian population is

the design and implementation of the 3 versions of the HEP: the rural HEP
(both agrarian and pastoralist) and the urban HEP.16,17 Taken together, there is
evidence to support the HEP’s relevance to the needs of the target communities.
This does not mean, however, that key informants’ concerns about the curricula
regarding gaps and redundancies in the level III and level IV curricula are
irrelevant. These concerns need to be addressed, and those who participated
in the development of the curricula are better positioned to address the issues,
whether in the form of modification, revision, or clarification.
3.1.3 Consistency between HEP curricula and task implementation
While most key informants reported that the curricula were well designed and
relevant to addressing the community’s health issues, some observed that there
were implementation problems, which could be attributable to several factors.
One relates to imbalance between the breadth of the curricular content and
the time allotted to cover the content, as stated by a key informant:
page-692
“
The training curriculum is broad. The breadth of the content and
the time allocated to cover the content do not correspond. Quite a
number of courses are required to be covered within a short period
of time, and that will overload the trainees. In that respect, the
curriculum needs revision and improvement.
If the curriculum is too broad and the allocated time is insufficient, instructors
may choose either to cover all the content of the curriculum without addressing
the topics in depth or to cover some of the topics—usually those that come
earlier in the list or that they find more important—and leave out other topics.
In both cases, proper coverage of the content suffers, which is a disadvantage
to the trainees.
Many key informants agreed that the current curricula were broad—nearly
identical to a bachelor’s degree program—and beyond the capacity of the HEP
candidates: that is, the curricula and the student capacity do not match. Due
to their advanced content, the present modules serve as reference material
for instructors and clinicians. It is not only the content of the module, however,
that is judged difficult for the trainees. According to many key informants,
the medium of instruction (English) is also a serious problem. The curriculum
is prepared in English, and courses are expected to be delivered in English,
even though most students do not understand English well. Indicating that the
curriculum was beyond the HEP candidates’ abilities, one key informant from
“
the Gambela region described the problem as follows:
For example, they are expected to know about malaria up to the level
of an expert, from prevention up to treatment. They need to know
about prevention, epidemiology, and clinical and prevention methods
of treating malaria. The same is true with vaccinations and diseases
like typhoid. There are 7 sections in the environmental sanitation
package, and the HEWs are expected to know all of them like an
expert.
COC Assessor, Gambella
page- 693
Joining the upgrading training program after a prolonged stay with the
community was also reported to be a challenge in terms of the students’ language
skills, since many have forgotten English during their stay in the community
working in local languages. One upgrading trainee from the Amhara region
“
stated the problem as follows.:
As we are HEWs, we live with the community and rural farmers, and
we mostly use Amharic, and we tend to forget English. So, it would
be better if this was corrected.
Other key informants noted that the curricula were comprehensive but did
not indicate that this was a problem. Instead, they indicated that the trainees
learned about many topics in college but translated only a few into practice
when working at HPs. In addition, according to these informants, even if the
instructors could cover the content broadly, the trainees had difficulty following
the instructor and understanding the content because of their poor educational
backgrounds. For example, even though the generic curriculum is designed for
students who have completed grade 10, students who completed grade 8 or
even lower are also reported to have been recruited as HEP candidates in
Somali.
In brief, the design of the program and the teaching–learning process adopted
in the program are generally considered suitable for addressing the health
needs of the community if they are translated into action. Many key informants
strongly believed that the curricula were beyond the capacity of HEWs and
beyond the scope of actual performance in the community or at HPs.
In sum, if there are problems in the training (as a result of the students’
backgrounds, the coverage of topics, the poorly equipped skill labs, or the
absence of skill labs), this is reflected in the competence of the HEP candidates
and their ability to deal with health problems in the community after graduation:
if the HEP candidates are not equipped with the requisite knowledge and
skills during their training, then their competence to deal with actual health
problems in the community will be questionable.
page-694
3.1.4 Balance between theoretical and practical HEP curricula

components
In principle, HEWs are expected to promote health in the communities

through educational and sensitization programs. They are also expected to
manage health problems in the community or treat patients. Thus, to meet
both expectations, the training needs to address both theoretical and practical
components, and this calls for balance between them.
According to the model HEP curriculum (for level III), which is based on
occupational standards, “the program will have a duration of 1 769 hours
including the on-job practice or cooperative learning time and civic education.”16,18
According to a key informant from the MoH, the theory–practice ratio should be
30:70, which generally gives much more emphasis to the practical component
than to the theoretical component. The practical component is further divided
into practical work in the colleges (model, group assignment, demonstration
and role play) and apprenticeship, which takes place in health institutions.
Among the trainees who participated in this assessment, the great majority
were (moderately, highly, or very highly) satisfied with the effectiveness of
the theoretical component (95%) and the practical component (91.4%). Even
though the trainees’ levels of satisfaction with the theoretical and practical
components were similar, several trainees complained specifically about the
practical components, including the lack of skill labs (or demonstration rooms)
in some colleges (Dire Dawa, Hosana and Arba Minch). Skill labs have still not
been put in place in many of the colleges since the beginning of the program,8
and this remains a challenge in some colleges.
Additionally, in colleges where a skill lab is available, not all of the necessary
equipment is available, and the lack or shortage of equipment hampers the
normal implementation of the practical component of the training. As well,
some HEP department heads spoke about the shortage or absence of skill lab
assistants as an impediment to the smooth implementation of the practical
component. According to some HEP department heads, the shortage of lab
assistants would force the instructors to do the demonstrations on their own,
which in some cases overburdened them; this can in turn affect the quality of
the practical component of the training.
page- 695
3.1.5 Summary on Relevance of Curriculum

This section examined the relevance of the HEP curricula using 4 criteria:
the curricula’s relationship with (1) the country’s health policy, (2) the actual
tasks that the HEWs are expected to carry out in the communities, (3) their
implementation, and (4) the balance between the theoretical and practical
components of the curricula. First, the findings indicate that the HEP curricula
are in line with Ethiopia’s health policy and thus are relevant in this regard.
Second, the findings also disclosed that the HEP curricula are consistent with
the actual tasks that HEWs are expected to carry out in the communities after
graduation. Despite some concerns about the breadth of the curricula (that
they are broader than the actual tasks of the HEWs, which should generally be
the case), the curricula are found to be relevant. Third, the findings showed that
the implementation of the curricula had encountered several problems (e.g., the
breadth of the curricula does not match the time allotted to cover them; the
English-language proficiency of the trainees is poor and affects the teaching–
learning process), but there were positive aspects: e.g., the program recruits
female HEWs in all sites (except in pastoralist areas, where the HEWs are
male); this develops HEPs (urban, rural, and pastoralist HEPs) that take into
account the diversity of the communities. Thus, with regard to implementation,
the curricula are partly relevant, but the problems must be rectified to make it
fully relevant. Fourth, the findings further disclosed that, for the most part, the
balance between the theoretical and practical components of the curricula is
compromised: the practical component in particular has encountered several
problems, including the absence of skill labs, at least in some colleges, the
shortage or lack of equipment in the skill labs, the shortage of lab assistants,
and the poor supervision of the trainees during their apprenticeships. This
appears to be the main challenge that hampers the quality of the program in
general and the curricula in particular. If there is any doubt about the relevance
of the HEP curricula, it should therefore be about maintaining the balance
between their theoretical and practical components.
Finally, the curriculum-development process does not seem to be participatory.

We observed in the curricula that instructors from different colleges were
the ones who had developed the curricula. In principle, curricula should be
developed by a multidisciplinary team of experts composed of subject-area
specialists, subject-area instructors, curriculum experts, experts on teaching
methods, and educational psychologists. Viewed from this perspective, the
curriculum development process has had flaws in the past, and this needs to be
rectified in similar undertakings in the future.
page-696
3.2 Adequacy of Resources in HEW Training Institutions
The adequacy of resources in the HEW training institutions influences the quality
of the HEP training, either directly or indirectly. We assessed the adequacy of
both human and material resources in HEP training institutions or colleges
directly through the examination of the institutional profile of the colleges and
indirectly through the examination of the stakeholders’ satisfaction. The latter
is based on the assumption that stakeholders’ level of satisfaction is a reliable
indicator of the quality and/or adequacy of the material and human resources.
3.2.1 Instructors/trainers
One of the key stakeholders of the HEP training is the faculty, trainer, or
instructor. The quality of the training is directly affected by the instructors’ or
trainers’ qualifications, experience, fields of specialization, and whether there
are the number of instructors needed for the smooth implementation of the
HEP training. Table 4 shows the instructors’ educational qualifications and
specializations.
Table 4: Instructors’ Educational Qualification and Specialization (N=192)
Variable Level N Percent

Diploma 10 5.2
B.Sc. 130 67.7
Educational
qualification MPH 31 16.2
M.Sc. 21 10.9
Nursing 80 41.7
Environmental Health 14 7.3
Health Officer/Public Health 55 28.7
Midwife 30 15.6
Area of Lab Technician/Technologist 1 0.5
specialization Epidemiology 2 1.0
General MPH 3 1.6
Others *
7 3.6
Total 192 100.0
Others includes specializations in Child and Reproductive Health (1), Health Extension (2), Human
*
Nutrition (1), Pediatric and Child Health (1), Public Health Nutrition (1), and Monitoring and Evaluation
(1).
page- 697
In terms of educational qualification, slightly more than two thirds (67.7%) of

the instructors have a bachelor’s degree, while 27% have a master’s degree.
The mix of the instructors’ educational qualifications appears to be generally
good. Additionally, most instructors have extensive experience (up to 25 years).
Only 2 instructors began teaching in the 2018/2019 academic year, and a large
majority (87%) have served between 1 and 8 years.
With regard to specialization, most studied nursing (41.7%), followed by health

officer training (28.6%) and midwifery (15.6%). The data indicate that only 1
lab technician/ technologist participated in the study. In addition, some key
informants clearly stated that there was a shortage of lab technicians and
technologists in the HEP departments. The institutional profile of 15 HEP
training colleges, for its part, provides a good picture of the instructors’ mix of
fields of specialization. This is presented in Table 5. The data clearly show that
the mix or distribution of instructors in terms of their fields of specialization is
not uniform or even. The relationship between the required competence and
field of specialization is discussed in more detail below under the heading
“professional competence of course instructors” (see section 3.4.2).
Data from the institutional profile support these findings. For example, in
2018/2019, there was no lab technician or technologist in 5 of the colleges
(Minilik, Metu, Negele Borena, Dire Dawa, and Hosana), while the number of
lab technicians in other colleges ranges from 2 to 19 (median 8).
Table 5 shows a similar problem with the number of female instructors in

institutions. The number of female instructors is relatively small, and in view
of the fact that the trainees are predominantly female, this result is somewhat
surprising. The proportion of female instructors ranges from 0% in Gambela to
60% in Minilik.
page-698
Table 5: Number of Instructors by College and Field of Specialization (2018/19

Academic Year)
Field of Specialization Total

College (female
HO Nursing EH MW LT Other Total instructors)
Debre Tabor 0 22 4 11 15 18 70 NA
Minilik 0 2 1 3 0 4 10 6
Gondar 6 17 4 12 7 7 53 7
Metu 11 7 1 4 0 5 28 5
Negele Borena 11 10 1 7 0 4 33 6
Dire Dawa 3 7 2 5 0 0 17 5
Gambela 0 3 1 0 2 0 6 0
Bahir Dar 4 22 2 8 5 16 57 9
Hosana 20 9 5 5 0 0 39 6
Mekelle 0 24 3 9 9 17 62 14
Arba Minch 33 1 2 1 11 0 48 14
Debere Berhan 22 26 4 10 8 1 71 8
Semera 2 17 0 4 2 0 25 5
Mizan 28 22 4 10 19 10 93 18
Dessie 11 22 2 8 12 22 77 20
Abbreviations: HO, Health Officer; EH, Environmental Health; MW, Midwife; LT, Lab Technician/
Technologist; NA, not available.
Another issue concerns the number of instructors who work in the HEP
departments. At least in some colleges, very few or no staff belong to the
Health Extension Program department exclusively, which suggests that the
department may be transitory rather than permanent. A Health Extension
Program department head had the following response when asked about the
“
number of instructors now teaching in the department:
There are no staff for the Health Extension Program Department

alone. The instructors teach in different departments of the health
science college, and I cannot tell you how many staff the Health
Extension Program department has. But we haven’t had a lab assistant
up until now. We have planned to hire one, but so far everything has
been done by the instructor.
page- 699
Angered by the lack of value placed upon or respect accorded to HEWs and
the program, one key informant insisted that this had to change urgently and
that all stakeholders needed to exert effort to this effect. He stated his views
“
as follows:
Because Health Extension is a multi-disciplinary profession, it has a

combination of competencies from environmental health, midwifery,
and public health. Usually, if there are nursing students, there is a
nursing department, and if there are lab students, there is a lab
department. In most colleges, the number of HEP trainees is much
higher than in other departments, but the HEP department and the
trainees are usually neglected by people, including the management
of the colleges. It appears that many of the colleges are interested
in the financial support they get from the MoH because of the
HEP. They seem not to be interested either in the program or in the
students. This calls for a serious effort on the part of key stakeholders
to reverse the trend.
3.2.2 Facilities
The main facilities required for the HEP training include classrooms, a library,
and demonstration rooms or skill labs with equipment. These resources required
for the HEP are supplied by RHBs or the TVETIs themselves. The status of the
facilities in each of the 15 colleges for which data were obtained is summarized
in Table 6.
page-700
Table 6: HEP Training Facility Resources, 2019
Internet Toilets
Library Skill lab Computer access Practice for Water
College Classrooms
capacity capacity lab for sites males points
trainees (females)
Debre
14 250 + 25 +- ++ +- (+-) +-
Tabor
Minilik 1 100 + 30 +- ++ +- (+-) +-
Teda/
13 100 + 40 ++ ++ +- (+-) +-
Gondar
Metu 10 60 + 20 0 +- +- (+-) +-
Negele
15 30 + 40 +- +- +- (+-) +-
Borena
Dire
3 0 0 0 0 ++ +- (NS) ++
Dawa
Gambela 18 200 + 30 +- +- +- (+-) +-
Bahir Dar 4 150 + 35 0 ++ +- (+-) +-
Hosana 10 400 0 0 0 ++ +- (+-) +-
Mekelle 38 200 + NS 0 +- +- (++) +-
Arba
15 160 0 20 0 ++ +- (+-) +-
Minch
Debre
12 40 + 45 ++ +- +- (+-) +-
Berhan
Semera 6 50 + NS 0 +- +- (+-) +-
Mizan 21 500 + 30 +- +- ++ (+-) ++
Dessie 14 80 + 25 0 +- +- (+-) +-
Note: 0, not available; +, available; +-, available but not adequate; ++, available and adequate; NS,
not sure.
CLASS SIZE
Despite the number of classrooms reported here, many key informants reported
large class sizes as a problem. Among the required facilities for the HEP training
is a classroom, and, according to the Ethiopian TVET-System model curricula
for level III and IV HEWs16,17 the classroom: trainee ratio shall be 1:25 or 1:30;
that is, no more than 30 trainees should be taught in a single classroom.
page- 701
The institutional profile data we obtained from each training institution indicate
that the class sizes range from 35 to 60 (median 50). One can thus say that
the stipulated ratio was not observed in many of the colleges. In some colleges,
reports indicated that the HEP classrooms were crowded (e.g., Bahir Dar and
Hosana), while in others up to 60 trainees were taught in a single classroom
(e.g., Negele Borena). Many colleges were reported to have taught many more
students per classroom than are specified in the standard (e.g., Arba Minch
and Mekelle), and some face a shortage of classrooms given the number of
trainees they have accepted (e.g., Teda/Gondar). In a few colleges (e.g., Dire
Dawa), the available classrooms were perceived as adequate for teaching the
HEP trainees.
LIBRARY
As shown in Table 6, every college except Dire Dawa has a library. The capacity
of the libraries ranges from 30 to 500. Thus, at one time, the libraries can
serve from a minimum of 30 to a maximum of 500 trainees (median 125).
While the existence of a library is important, several other factors influence
its smooth functioning, including sufficient ventilation, the number of chairs
and tables, and the variety of books available. In this regard, the libraries in
some colleges (e.g., Mekelle) have limited capacity compared to the number of
trainees they serve. The library in Bahir Dar, in particular, is reported to have
limited seating and no ventilation and require maintenance. In other colleges
(e.g., Teda/Gondar), many books are outdated and old.
SKILL LAB
A skill lab is another essential facility for HEP training. Among the 15 colleges
(Table 6), 3 reported having no skill lab at all. The capacity of the skill labs
in the colleges that have them is between 20 and 45. While the availability of
the skill labs or demonstration rooms is positive, it is important to note that the
skill labs in almost all colleges serve a large number of HEP trainees at a time.
According to the Ethiopian TVET System model curriculum for the Health
Extension Service18,19 the ratio of skill lab to number of trainees should be 1:6. In
addition, the availability and adequacy of all necessary equipment for the skill
lab is another issue: not all necessary equipment is available in most colleges,
page-702
and this inhibits the skill labs’ functioning. Thus, even though available, most
skill labs do not have the necessary equipment, and not all skill labs are fully
functioning.
COMPUTER LAB
As shown in Table 6, 2 of the 15 colleges have no computer lab for HEP

trainees. While most colleges have computer labs, the labs’ capacity in terms of
the number of trainees they serve at a time ranges from 20 to 45 (median 30).
The capacity of the computer labs is slightly more than what is indicated in
the Ethiopian TVET System model curricula for level III and IV HEPs,17,18 which
is 1:25. What is more important, however, is whether the available computers
are in good condition, and in this respect there are no complaints. The main
complaint raised in a few colleges (e.g., Bahir Dar) is that HEP trainees are
not allowed to use the computer lab, although trainees and students from other
departments (e.g., nursing and midwifery) are well served.
INTERNET ACCESS FOR TRAINEES AND INSTRUCTORS
Internet access for trainees is another important requirement for colleges to

meet, but in 8 (Metu, Dire Dawa, Bahir Dar, Hosana, Mekelle, Arba Minch,
Semera, and Dessie) of the 15 colleges for which we obtained data (Table
6), HEP trainees have no internet access. In 5 colleges (Debre Tabor, Minilik,
Negele Borena, Gambela, and Mizan), trainees have internet access, but it is
considered inadequate. Only in 2 colleges (Teda/Gondar and Debre Berhan)
do students have very good (or adequate) access to the internet.
Internet access for instructors is better than that for trainees. It is only in 2
colleges (Dire Dawa and Semera) that instructors have no internet access.
In the remaining 13 colleges, instructors have access to computers. While in
10 colleges (Debre Tabor, Minilik, Metu, Negele Borena, Gambela, Bahir
Dar, Hosana, Mekelle, Arba Minch, and Mizan) the access is inadequate..
Inadequate internet access means that there is either no Wi-Fi or that there are
problems with access because of the limited number of computers compared to
the number of trainees who use the computers in a lab to access the internet.
page- 703
PRACTICE SITES
Practice sites are health institutions (mainly HCs and hospitals) where the
HEP trainees are placed during their apprenticeships. As shown in Table 6,
all colleges have practice sites, but, despite their availability, the practice
sites are perceived to be adequate in only 7 of the 15 colleges. Adequacy is
determined by the number of HEP trainees placed in a single health institution.
In the remaining 8 colleges, the practice sites are considered inadequate. This
simply means that, except in urban centers like Addis Ababa, where there are
several health institutions (hospitals and HCs) where HEP trainees are placed
for their apprenticeships, in small towns the number of health institutions is
correspondingly small, and HEP trainees are placed in these institutions. The
placement of many HEP trainees in 1 health institution for their apprenticeships
may affect the effectiveness of the apprenticeship in at least 2 ways. On the one
hand, the placement of a large number of trainees burdens the preceptors or
on-site supervisors, which in turn can affect the effectiveness of their supervision.
On the other hand, the trainees may not get the chance to practice as much
as necessary because of their limited interaction with and supervision from the
preceptors.
The level IV curriculum for the Health Extension Services, which is developed
based on occupational standards,17 notes the following in relation to the
apprenticeship of the HEP trainees and how the apprenticeship should be
“
implemented:
The mode of delivery is cooperative training. The TVETIs, identified

hospitals, and HCs have forged an agreement to cooperate with
regard to implementation of this program. The time spent by the
trainees in the health institutions will give them enough exposure to
the actual world of work and enable them to get hands-on experience.
The cooperative approach will be supported with lecture-discussion,
simulation, and actual practice. These modalities will be used before
the trainees are exposed to the health institutional environment.17
The curriculum describes the general features of the apprenticeship program for
the HEP trainees and how it should be implemented. The cooperation between
the HEP training colleges and the health institutions appears to be generally
page-704
good. There are, however, complaints regarding the instructors’ follow-up with
the HEP trainees when the latter are placed in the health institutions for their
apprenticeships. In other words, the instructors’ follow up with the HEP trainees
is judged “poor” by preceptors, who are on-site supervisors of the trainees. More
specifically, several preceptors reported that “once the supervisors attached
students to the head of the health facility, they visit them once a week, and this
poor follow-up will affect students’ performance.”
In sum, our data indicate that all HEP training colleges have practice sites.
Despite their current availability, the lack of practice sites was reported as a
problem encountered by HEP training colleges, particularly at the beginning
when the HEP training was launched.8 The present findings, therefore, show a
significant improvement in the availability of practice sites. The findings also
show, however, that the problems persist pertaining to practice sites. These
include the placement of a large number of trainees in 1 health institution and
the instructors’ relatively rare supervision of the trainees’ apprenticeships. There
is a strong belief among some key informants that these problems affect the
effectiveness of the apprenticeship in general and the practical aspects of the
training in particular.
TOILETS AND WATER POINTS
The data summarized in Table 6 show that toilets for both male and female
trainees are available in all colleges (except in Dire Dawa, where the availability
of toilets for female trainees was unknown). The toilets for both male and female
trainees in most colleges, however, are generally rated as inadequate given
the number of the trainees who use them. In this case, there are 2 exceptions
(Mekele and Mizan), where either the toilet for female trainees (Mekele) or the
one for male trainees (Mizan) is judged to be adequate.
A water point is available in each of the 15 colleges (Table 6). The water points
are considered inadequate, however, in terms of the number of trainees who
use them in all colleges except 2: Dire Dawa and Mizan.
page- 705
3.2.3 Teaching modules
Twenty-one teaching modules have been developed for the HEW training
program. The trainees (both level III and level IV) use the teaching modules as
a reference. The model curriculum for the HEP17 recommends that the student-
teaching/training module ratio be 1:1. Data from the institutional profile indicate
that in all 15 colleges, teaching modules are available. They are perceived to
be adequate, however, in only 3 colleges (Gondar, Dire Dawa, and Debre
Berhan), where the number of modules matches the number of students. In
12 colleges (Debre Tabor, Minilik, Metu, Negele Borena, Gambela, Bahir Dar,
Hosana, Mekelle, Arba Minch, Semera, Mizan, and Dessie), the number of
students is very large compared to the number of teaching modules available;
hence, not every student can have a module.
Another problem raised by key informants in relation to teaching modules is
the fact that the colleges still use the old module, even though the curriculum
they are using is new. A key informant from Gambela stated the problem as
“
follows:
The other problem is related to modules.…For example, there are

old curriculum modules. As you see, we are using 2016 modules, but
the curriculum we have is the 2018 curriculum. We don’t have any
modules for the new curriculum. It would be great if the students
could get the module as before.
Put differently, the concern of the key informant is that the curricula and the
accompanying modules do not match: while the curriculum has been revised,
the modules have not. This perception was contradicted by a key informant
from the MoH, who argued that the revision was only minimal and that only
some chapters were modified, so revisions do not prohibit the old models from
being used with the new curricula. This key informant also emphasized that the
instructors were essentially responsible for developing the modules and that the
MoH developed the modules centrally so that they would be used uniformly
by all colleges. In essence, the key informant stated that the colleges could
use the old modules, but, when they do this, they need to prepare some of the
chapters of the modules that correspond to the new chapters in the curricula.
Overall, there seems to be some confusion regarding the responsibilities of the
instructors, the college, the RHB, and the MoH. Clarifying the responsibility of
each party will help make the parties accountable and avoid the confusion that
may hinder the smooth functioning of the HEP training.
page-706
In brief, consistent with the quantitative data reported above, the qualitative
data obtained through key informant interviews showed several problems
pertaining to material and human resources. It is reported that, due to the
large number of students, there is no match between the number of students
and the available classrooms, which in turn leads to large class sizes, a factor
that could affect the quality of the training. According to the informants, the
narrow size and limited number of demonstration rooms and the shortage of
equipment or demonstration materials fueled by the large number of students
are factors that impede the learning of skills.
Primarily, however, the shortages of instructors, skill lab assistants, and course
modules were reported as critical gaps in the teaching–learning process. The
shortage of the required resources is mainly related to high staff turnover, the
difficulty of getting materials in the market, a delayed procurement system,
“
and budget shortfalls. One college dean from Oromia described this:
There are problems related to the absence of demonstration or skill

lab rooms and required materials. Even if we have the money to
buy them, some equipment is not available in the market, and there
is a delay in the procurement process. With regard to teachers, we
don’t have enough qualified teachers. We get teachers experiencing
a lot of difficulties….We don’t even have the structures for skill lab
assistants at the TVET, let alone professionals.
On the other hand, the MoH has been supporting the improvement of the HEP
trainings by filling the budget gap at the regional level. In particular, the MoH
has been supporting the health science colleges through the RHBs since 2005.
A report by the human resource development directorate of the MoH shows
that a total budget of 453,953,808 birr has been directly financed by the MoH
to health science colleges from 2005-2011 EFY. On top of the regional budget
allocation, the federal budget support is intended to support all necessary skill
lab materials and operational costs based on the identified gaps.
In summary, there has clearly been improvement over the years in the
availability and adequacy of resources8. Compared to the findings of past
studies conducted, the present assessment shows great improvement in many
respects: the number, qualifications, and specializations of instructors, as well
as the availability of libraries and skill labs, to mention just a few. It should
page- 707
be noted, however, that some problems persist: in particular, even though skill
labs are available at present in most colleges, most do not have the necessary
equipment, so the training suffers considerably.
3.3 Quality of Course Materials
The quality of course materials is an important determinant of the process and

output elements in the training program. Some key indicators were adapted
from Bradley’s effectiveness model for curriculum development; the quality of
the training materials were rated by course instructors and trainees.
Table 7: Instructors’ and Trainees’ Evaluation of Course Material Quality
Features of course Very good/good Fair Poor/very poor

material Frequency Percent Frequency Percent Frequency Percent
Clarity of the course
1239 86.2 166 11.6 32 2.3
objectives
Relevance of the content to
developing subject matter 1192 83 187 13 58 4.1
competence
Relevance of the content to
1188 82.6 202 14.1 47 3.2
the developing trainees’ skills
Depth of content to produce
qualified Health Extension 1145 79.7 230 16 62 4.3
Workers
Breadth of content to
produce qualified Health 1136 79.1 247 17.2 53 3.6
Extension Workers
Time allocated for each
course in the training 901 62.7 333 23.2 203 14.2
program
Overall quality of the course
621 43.2 594 41.3 222 15.4
material
Generally, both instructors and students rated the quality of the curriculum and
course materials positively and found them relevant to their intended purpose. A
majority reported that the curriculum and course materials had clear objectives,
were relevant in terms of subject matter mastery and skills development, and
had the proper depth and breadth for their intended purpose. Fewer rated the
page-708
time allocation for each course as very good or good, but only 43% rated the
overall quality of the course materials as very good or good, while 56% saw it
as fair, poor, or very poor.
According to key informants, the major issue with regard to the course material
was the lack of access to teaching materials. Students and department heads
reported that the instructors met the quality standards for course outlines and
handouts. There was also a quality teaching module centrally developed for
“
each course that could be used for both level III and level IV programs and
disseminated to each college:
The modules are developed by the Ministry of Health in collaboration

with the Open University. They are developed for level IV training
but could also be used for level III. They could be used as models
even for other TVET programs. It was developed in line with the
occupational standards set in the curriculum guide with good quality.
Most modules, however, comprise in-depth, advanced concepts beyond the

trainees’ capacity, and some modules need updating (e.g., immunization) to
include the latest advancements and changes in service provision.
3.4 Quality of Course Delivery and Assessment

Methods
Seven components of the course delivery approaches and assessment methods

were assessed: the relevance of teaching modalities, the professional competence
of the course instructors, student recruitment, teaching methods, the language
of instruction, the effectiveness of the teaching methods, and student assessment
methods.
3.4.1 Teaching Modality
The HEW training program uses a modular approach in which students take
courses in modules consecutively based on their logical order. There have been
debates over the appropriateness of this modality.
page- 709
Figure 2: Student Preference for Course Delivery Approaches

As indicated in Figure 2, 58.6% of students preferred the current block-course
approach, while 40.3% would prefer a semester course approach, and 1.1% were
not sure.
Trainees described the modular approach as advantageous because it: (1)

makes it easier to study and remember courses, (2) provides an opportunity to
complete each course quickly, (3) gives students more time to study, and (4)
provides the opportunity to complete courses individually with focused attention.
Both instructors and students preferred the modular over the semester modality
because it allowed them to concentrate on one learning outcome at a time.
Trainees who preferred the semester approach cited the ample time it gave for
both practical and theoretical components, and instructors described it as more
suitable for practical exercises. Some instructors raised concerns about modular
teaching: (1) since the courses are delivered intensively every day, completing
each one within a week or two, it would be difficult to accommodate students
who miss classes for various reasons, (2) students may be unable to cope with
the pace of the course delivery due to their poor academic backgrounds, (3)
there is a shortage or absence of modules to be given for students, and (4)
instructors are challenged to teach the same course for the same class every
day, morning and afternoon, for a whole week or more.
Furthermore, due to resource and time limitations, modular approaches are not
always applied. As stated by a key informant from the MoH:
page-710
“
In some situations, teachers cluster similar courses together and deliver
courses simultaneously. After completing the courses, the students go
out for the already completed clustered courses’ apprenticeships at
one time.
Our findings revealed that, although there are a number of factors that affect
its practice, the modular teaching approach was preferred by the majority of
key informants and survey participants.
3.4.2 Professional Competence of Course Instructors
The professional competence of the instructors is a vital element in a training

program that is presented here under 2 themes: the relevance of the instructors’
educational qualification and their professional competence.
RELEVANCE AND EDUCATIONAL QUALIFICATION OF TRAINERS
According to the model curriculum of the Health Extension Service, instructors

conducting this particular TVET program should be at B level (a minimum of
a BA or B.Sc. degree) and have satisfactory practical experience or equivalent
qualifications. Specializations include health officer and midwifery (appropriate
for delivering all training) or environmental health (appropriate for delivering
courses only in select modules). From this perspective, the most relevant
instructors are those with at least a B.Sc. degree in public health or
midwifery.
More than 90% of the participants included in this study had a B.Sc. or BA and
an MPH or M.Sc. degree and were eligible to teach courses in the program
according to their educational qualifications. The majority of instructorswho
were playing an active role in course delivery, however, had a nursing
specialization.
page- 711
“
According to one focal person from the MoH:
The professionals most appropriate for delivering courses in the

HEW training programs are health officers or midwives. Nurses
do not have the competence to provide most of the courses in the
training curriculum. Nurses are trained to provide mostly hospital
services and bedside care. There are cases in which nurses provide
training in various modules in different colleges, but this is completely
wrong. They could provide training in limited courses like emergency
nursing.… Otherwise, they don’t have the required competence to
provide the training.
Therefore, the analysis of educational qualifications shows that, although most

instructors have the required educational qualification to deliver the courses,
nurses’ delivery of different courses is not in line with the profile of the trainers
required by the curriculum guidelines.20
TRAINERS’ PROFESSIONAL COMPETENCE
To carry out the teaching process effectively, the course instructors’ professional
competence, motivation, and commitment are vital. Table 8 below presents
the trainees’ evaluation of the professional competence and related qualities
of their course instructors in the course of the teaching–learning process. Data
from key informants help explain the professional competence of the instructors
in the HEW training program.
Table 8: HEP Trainees’ Ratings of the Instructors’ Professional Competence
Features of instructors’ Very Good/good Fair Poor/very poor

professional competence N Percent N Percent N Percent
Teaching motivation 1043 83.8 166 13.3 36 2.9
Preparation for class 1048 84.2 163 13.1 34 2.8
Interest in the subject matter they teach 1006 80.8 202 16.2 37 3
Professional competence 1051 84.4 165 13.3 29 2.3
Mastery of the subject matter 1023 82.2 194 15.6 28 2.3
Effort to create a participatory
994 79.8 202 16.2 49 4
classroom atmosphere
Provision of useful feedback to students 864 69.4 264 21.2 117 9.4
page-712
Development of realistic and relevant

950 76.3 57 4.5
assessment tasks 238 19.1
Organizing course-related materials 934 75.1 242 19.4 69 5.5
Stimulating student interest in the
912 73.3 73 5.8
profession 260 20.9
Using class time effectively 886 71.2 265 21.3 94 7.5
Commitment to supporting students’
913 73.3 260 20.9 72 5.7
learning
As indicated in Table 8 above, a large percentage of the trainees perceived

their instructors as professionally competent in all areas, including motivating
students, preparation, providing feedback and support for student learning.
Consistent with the quantitative findings, students, instructors, department

heads, and deans indicated qualitatively that most instructors were competent
at training HEWs. They did indicate, however, that there were few opportunities
for course instructors to receive continuing education or skills development,
especially in emerging science advancements. This problem is particularly
serious among programs run by TVET agencies that prioritize instructors
in other programs. This compromises the opportunities for students to learn
current concepts and undermines instructors’ capacity. This was well articulated
“
by one key informant:
Modules are not updated in a timely way. Teachers are not updated
with new clinical management, like vaccinations, TB, or HIV.
Particularly those students who are at level IV have taken a lot of
trainings in their workplaces. They have more recent information in
some areas than their teachers do. As a result, there was a situation
in which the teacher was embarrassed [for being behind his students].
“
A similar issue was raised by a respondent from the MoH:
There are many changes in some courses like Vaccination that are
not updated in the teaching modules. A teacher who doesn’t have
information on these changes may teach the old concepts from the
modules, which are not currently in practice in HCs or HPs.
page- 713
Generally, the findings on the professional competence of course instructors

indicated that, although most instructors are professionally qualified to
handle the learning process, their opportunities to participate in professional
development activities are very limited.
3.4.3 Student Recruitment
Participants criticized the recruitment criteria and procedures for training

program candidates. According to the curriculum guidelines20, female students
having completed grade 10, living in their communities, and having achieved
the required GPA on the grade 10 National Examination, which is annually set
by the MoE for TVET programs, could be recruited. In some cases, entrance
exams are developed and administered by RHBs, particularly when the number
of applicants is much larger than the actual need in the region.
To compare participants’ entry characteristics against the standards, trainees

were asked to provide their own grade completion information and describe
the factors that motivated them to apply to the HEP.
Figure 3: Educational Status of Trainees

Figure 3 shows that the majority (78%) of students enrolled in the program
had completed only grade 10, and an additional 21% had completed grade 12.
Surprisingly, about 15 students (1%) who joined the program had not completed
grade 10, which is below the standard set in the curriculum guidelines and entry
requirements proposed by the Ministry of Education.
page-714
Table 9: Reasons for Joining the HEW Training Program
Reasons for joining the Health Extension

Frequency Percentage
Program
Keen interest in being an Health Extension
657 52.8
Worker
Job availability 364 29.2
Free scholarship 46 3.7
Family pressure/recommendation 65 5.2
Peer pressure/recommendation 41 3.3
No other options 72 5.8
Total 1245 100.0
Students were asked to report their primary reason for joining the HEW
training program. Over half reported that their keen interest in the profession
motivated them, while one third reported that they had joined the program
because of job availability. A minority were motivated by the free scholarship,
family pressure, or peer pressure, and 5.8% of respondents joined the program
due to a lack of other options.
Qualitative informants described problems in the program’s candidate

recruitment and selection procedures. Respondents mentioned the selection of
students with poor academic backgrounds, the recruitment of candidates unfit
for other TVET programs, and candidates uninterested in the health science
profession. Some reported that the selection process was vulnerable to abuse
and that some government officials intentionally abused the selection process
by recruiting their relatives. Additionally, the screening system does not allow
training college representatives to participate in selection committees; the
RHB alone selects and deploys candidates to their respective colleges. With
regard to the upgrading program, those who are interested and can fulfill the
criteria (notably by passing the COC exam) can join the training program. An
additional complication is that some trainees do not complete the program
in a timely manner, stopping and restarting after long delays, which prohibits
the successful mastery of the material due to their lack of English-language
competence and poor retention of the theory taught in their earlier training.
page- 715
Some trainers also identified the late selection and deployment of trainees as
a major hindrance because, unlike other programs with explicit registration
periods in the same colleges, the HEW training program has no fixed admission
period. This rolling admission procedure creates pressure on both candidates
and trainers to deal with every aspect of the course, both practically and
theoretically, within the remaining short period of time. Both the quantitative
and qualitative data show that the recruitment procedure is one of the most
problematic areas of the training process and needs urgent attention.
TRAINEES’ ATTITUDES
As an important factor in the effectiveness of the program, trainees’ attitudes

were measured using an attitude scale. The scale comprises 10 items, which are
presented to the trainees on a 5-point scale. The trainees’ mean attitude score
was 37.55, and this was compared with a score that represents the minimum
positive attitude value. The score we selected was 40. One sample t-test was
run to assess the existing differences, and the result disclosed a statistically
significant difference between the actual mean attitude score of the sample
trainees and the test value (t=-17.8, df 1244, p<0.05). The trainees’ mean
attitude score (37.55) was less than the expected mean score of 40, which
is considered here as a score that represents a moderately positive attitude
toward the HEW job. In general, the findings indicate that the attitude of
trainees toward the HEW position or career is marginally positive but not as
strong as it should be.
In fact, the above finding is also well supported by information from key
informants. Students claimed that their profession was highly undermined by
other health science students and the community in general and that serious
modification needed to be introduced, starting with something as basic as the
name. As one key informant explained, the term “HEP” should be changed to
something more appropriate since it is highly undermined and politicized in
the community in general and has come to be negatively perceived by the
community.
page-716
3.4.4 Teaching Methods
The HEW level III model curriculum is based on occupational standards2 and
identifies 3 components in the HEW training program: theory, practical work,
and apprenticeship. Since it follows the TVET system’s training approach,
70% of the courses should be delivered practically and 30% theoretically. The
practical training is to be given in both health institutions and training colleges
where students are required to have practical exposure in skill laboratories
after learning concepts in class.
To achieve the intended objectives, the curriculum urges instructors to employ a

variety of teaching methods to allow students to grasp concepts according to
their individual learning styles. Instructors were asked to rate the frequency of
the teaching methods they employed during training.
Figure 4: Frequency of Teaching Methods Used in the HEW Training Program
page- 717
Figure 4 shows the various types of teaching methods being used. A majority of
instructors reported frequently using lectures (82.8%), demonstration (65.6%),
guided practice (59.4%), and assignments (59.4%) in their respective courses.
The least frequently employed methods (by a considerable portion of the
respondents) were seminar (58.48%), individual learning (49.4%), and case
analysis (48.4%). The remaining methods were used at various rates that ranged
from rarely to always. Critical inspection of the results in Figure 4 indicates that
theoretical methodologies are used more frequently than practical ones.
Key informants also stated that the main teaching method in the HEP was
the lecture, followed by group discussion and assignment. In addition, both
instructors and students reported that more time was spent on theoretical
components than practical ones, demonstrations, or attachments. One focal
“
person from the MoH confirmed this:
As a part of the TVET system, HEWs are trained as skilled mid-level

professionals who meet the needs of the community. Indeed, 70%
of the training is expected to be practical and 30% theoretical. In
reality, this is not the case. Most teachers spend most of the time
allotted for practical training in class. Students have very limited
opportunity for practical experiences.
There is also poor monitoring of students during their apprenticeships. According

to the preceptors’ report, once supervisors assign students to the head of the
facility, they rarely visit them, which affects trainees’ performance. Additionally,
they reported that trainees were unwilling to be supervised or monitored by
preceptors in the HCs since preceptors do not take part in the evaluation
process.
Generally, it was found that the practical course components were compromised
by and resulted in limited practical skills. In fact, various studies recommend
the use of a more practical approach to the production of skilled workers
in the training of professionals in health sciences, like the HEP. The systemic
analysis of 37 articles by Bluestone and colleagues revealed that case-based
learning, clinical simulations, practice, and feedback were effective educational
techniques. Didactic techniques that involve passive instruction, such as readings
or lectures, have been found to have little or no effect on learning outcomes.
page-718
3.4.5 Effectiveness of the Teaching Methods Employed
The effectiveness of the teaching methodologies in the HEW training program

was also assessed in various dimensions ranging from scheduling flexibility to
the variety of learning activities.
Figure 5: Student Satisfaction with Course

Comparatively, a sizable segment of student participants expressed their
satisfaction with the variety of learning activities provided (58.5%), course
delivery (58.1%), and course coverage (50.1%). By comparison, a relatively large
segment indicated moderate satisfaction with the flexibility of course scheduling
(49.6%), the level of difficulty of the courses (48.1%), the quality of the course
materials (43.5%), and the adequacy of the course materials (39.4%). The
number of trainees who rated their satisfaction as low or very low should not
be discounted: it was more than 10% in the areas of course scheduling flexibility
and course materials.
page- 719
Key informants also reported gaps in the effectiveness of the methodologies,

particularly with respect to the shortage of time given to covering each course
“
in the modular approach and the modules crammed into an academic year:
The program has to cover both the theoretical and practical aspects
in 17 modules within almost a period of 10 months.…It is, practically
speaking, very difficult. More importantly, the competence level of
the trainees will not allow them to achieve this.
Health colleges’ prioritization of their regular diploma and B.Sc. health science
students over HEP trainees was an additional hindrance. One HEW trainer
“
explained:
The HEP trainees are the least prioritized in many cases. When
classes are scheduled, it is only after completing the schedules of
the other programs that the HEP trainees are considered. Because
there is a shortage of classrooms even for the other programs, in
many cases we teach our [HEP] trainees in the evenings and on
weekends. They are given almost no opportunity to practice in the
demonstration rooms that were established for other programs (e.g.,
the nursing department). All other departments except the HEP have
their own demonstration rooms.
Additionally, there seem to be no clear criteria for the time allocated for each
course. As described by the dean from Negele Borena HSC, the time allocated
for some modules was sufficient, and the instructors completed those modules
without any pressure, while very limited time was allocated to other courses
that put instructors and students under stress to finish within the timeframe.
3.4.6 English as the Language of Instruction
In general, English is the language of instruction in higher education. All HEP

teaching modules and instruction materials are developed in English, as are
many of the guides and reference materials used in HEWs’ professional work.
Only a few professional materials were also available in local languages.
page-720
English was selected as the medium of instruction with the aim that trainees
could pursue further education in other health disciplines as well. It was also
believed that trainees could benefit from the language experience since many
reference materials are available in English. Practically, however, it was found
more effective to use the local language as a medium of instruction for 2 main
reasons. First, students can easily understand and grasp the concepts when
they are presented in their respective languages. Second, since they support
the local community, familiarity with the services and procedures in the local
language will help them match what they have learned with actual practice.
In the current study, students reported that, despite the preparation of the
modules in English, most instructors teach them by translating into the local
language. Practically speaking, English language was not the language of
instruction in the program since instructors delivered the courses in local
languages. Health science colleges in Tigray, for example, deliver the courses
using the regional language as the medium of instruction. Some instructors
even questioned why English was the medium of instruction since HEWs are
expected to communicate using local languages. One COC assessor from
Oromia asked, “Everything they do at work is in Amharic. They do nothing in
English. So, what is the rationale for using the English language as a medium
of instruction?”
The relevance of English language as the language of instruction was also
“
questioned by a key informant from the MoH:
I don’t see the importance of English as a language of instruction

for HEP trainees. I believe that many of the problems in the HEP
emanated from the fact that students are learning courses in a
language they don’t really understand. It would have been better if
the courses were delivered in their respective regional languages.
The use of the English language as the medium of instruction was the factor
that most consistently believed to contribute to the complicated situation in
HEW training.
page- 721
3.4.7 Assessment of Learning and its Effectiveness
In the Ethiopian educational system, instructors are strongly encouraged to

employ continuous assessment approaches, regardless of the level they teach,
and are expected to use a variety of assessment techniques appropriate
for the courses they are delivering. The HEW training curriculum was also
developed with the expectation that instructors would use a variety of practical
and theoretical assessment techniques to effectively determine the learning
competencies of HEWs.
Figure 6: Frequency of Assessment Method Use in HEW Training Program

A majority of instructors (55.7%) reported that written exams were always
used as a major assessment technique. All other assessment techniques, except
project work and portfolio assessment, were frequently used.
Our qualitative data contradicted the quantitative data that indicated the high
prevalence of continuous assessment. An interview with a trainee from Debre
Tabor disclosed that instructors sometimes conducted classroom continuous
assessment but primarily relied on end-of-course final exams to evaluate
students’ performance. Others confirmed this:
page-722
“
Teachers only give 1 exam at the end of the course. If the course is
very broad, they could give us an exam at the end of each unit or
module.
Generally, teachers rely on midterm exams and final exams to

evaluate students.
End-of-course exams determine students’ eligibility for the next course. Students
who earn a total score below 50 are not allowed to proceed but will be given
the opportunity to sit for a re-exam after remedial support. They have 3 chances
to score the minimum passing mark; otherwise, they face academic dismissal.
Table 10: Effectiveness of Assessment Methods
Trainees
Indicators Rating
Frequency Percentage
Very high/high 741 59.5
Extent to which the assessment processes
Moderate 416 33.4
enable learners to demonstrate their ability
Low/very low 88 7.1
Extent to which learning is facilitated by
Moderate 399 32
feedback and follow-up examinations
Low/very low 170 13.7
Frequency of the feedback provided by
Moderate 438 35.2
instructors for student learning
Low/very low 210 16.8
The majority of trainees rated the assessment techniques used in the program
as highly effective in helping students demonstrate their learning and in
“
facilitating learning through the provision of feedback (Table 10).
Some trainees also praised the assessment methods:

These assessment methods were important in advancing our
knowledge and skills so far. The instructor frequently provides us
feedback based on our test results and classroom learning behavior.
page- 723
Others indicated that the assessment process had serious problems.The HEP
curriculum guides instructors to use formative techniques for thebenefit of
both students and instructors so as to enhance the teaching andlearning
processes.17,18 Based on the reports of trainees and deans, instructorsdo not
have time for formative assessment as a means of feedback for students’
learning.
The summative evaluation (i.e., written exams, assignments, presentations, and

seminars) administered at the end of each module or learning competence
determines the extent to which competence has been achieved. The results can
also help instructors label trainees as “competent” or “not yet competent.”17,18
Participants reported problems with the results’ dependability, however: (1)
college administrators pressure instructors to promote students despite their
actual competence because colleges want to maintain a low attrition rate,
(2) instructors give passing marks to every trainee to avoid repeat re-exams,
(3) instructors usually give passing marks to all trainees to get better student
evaluations, and (4) although exams are prepared based on learning outcomes,
most students do not do well on the exams because of their poor learning
competence. Respondents also complained about the dichotomous grading
system (“competent” or “not yet competent”), despite large variation in trainees’
actual scores. Awarding the same grade, “C” or “competent,” to students scoring
between 50% and 90% does not foster competition among students. Instructors
and students recommend changing the grading system to one of letter grades.
3.5 Competence of HEP Trainees
According to the HEP curriculum, upon the successful completion of the program,
trainees are expected to demonstrate their competence on the acquisition and
implementation of the HEP components.
3.5.1 Instructors’ Evaluation of Trainees’ Competence
Course instructors were asked to rate their students’ level of competency on the
courses they had already completed.
page-724
Figure 7: HEW Trainee Competence
Abbreviations: IMNCI: Integrated management of neonatal and childhood illnesses; PMTCT: prevention
of mother to child transmission of HIV.
Figure 7 shows the instructors’ assessment of student competence. The majority

of instructors did not rate the students from their courses as highly competent.
Students were perceived as relatively more competent in 4 components of
the package: managing ante-natal care (49.5%), performing advocacy and
community mobilization (44.2%), managing hygiene and environmental health
(42.2%), and intervening in nutritional problems (40.1%). Instructors rated
page- 725
students as least competent in applying computer and mobile technology

(34.3%), managing pharmaceuticals in HPs (18.2%), and managing delivery
practice (17.7%). From this it is possible to infer that instructors have concerns
about trainees’ competence.
Table 11: Competence Trainee by Program Ownership and Program Type
Ownership of the program N Mean SD df t-value p
Technical and Vocational Education

73 66.53 16.84
Training Agency 153 -3.6 0.000
Regional Health Bureau 82 76.96 18.82
Program type
Generic-level program 653 72.86 14.99
1210.73 -11.52 0.000
Upgrade-level program 562 81.90 12.35
A simple inspection of the scores shows differences in the ratings of students’

competence as a result of program ownership and type. To test the significance
of between-group differences, an independent sample t-test was computed.
Trainees’ competence at the colleges owned by the RHB were significantly
higher (t=-3.6, df=153, p<0.05) than those of trainees at colleges under
regional TVET ownership. On the other hand, according to the trainees’ self-
ratings, the upgrading students’ competence was significantly higher (t=-11.52,
df=1210, p<0.05) than that of the trainees in the generic program.
Most key informants also agreed that HEWs were not adequately competent
in learning outcomes or community-based tasks. Both instructors and COC
assessors confirmed that the majority of HEWs passed examinations only after
repeated attempts and support. In this regard, a dean from Amhara noted that
only 33% of trainees had passed the COC exam that year. Similarly, the deans
of the colleges identified the passing rates of trainees on the COC exams that
year as 40% in Somali, 30% in Afar, and 35% in Shashemene.
One key informant confirmed the quantitative findings when he explained that
level III trainees are poorer academically than level IV trainees for 2 reasons:
1) they are selected from grade 10 or 12 without prior experience and 2) they
usually join the HEP after being classified as unfit by other technical schools.
By comparing and contrasting the theoretical and practical aspects of the
page-726
training, participants illustrated that trainees were better in practical skills than
in theoretical knowledge because their English-language proficiency was a
barrier to their reading and understanding the modules prepared in English.
Key informant instructors evaluated most trainees as academically incompetent
both during admission and at the completion of their training. One instructor
expressed this view:
“
The main problem is the poor academic background of students.
Students are not academically capable of understanding the course.
They may pass the COC through many means, like cheating, but they
don’t have the required basic knowledge and skills. The competence
problem of HEWs is, therefore, directly related to the selection of
incompetent trainees during their entrance.
Preceptors (professionals at the HCs who supervise the HEWs’ apprenticeships)

made a similar claim that HEWs manifest a serious lack of skills and knowledge
during their actual practice at HPs. They argued that most trainees were not
even ready to learn under supervision and were negligent due to the preceptors’
lack of participation in the evaluation of trainees. Additionally, a key informant
“
from the MoH observed:
HEWs are trained to be general practitioners, to assume the roles

of multiple professionals, including nurses, health officers, midwives,
and health educators so as to provide inclusive quality services
independently in environments where getting other professionals
is impossible. It is based on this fundamental principle that the
curriculum is designed and the modules are prepared. The learning
competencies are also designed to reflect this principle. Practically,
however, the competence of HEWs falls very short of the standards.
The COC assessors also reported differences in performance between HEWs

who are trained in HSCs under the ownership of RHBs and those trained in
TVET agencies. HEWs trained in TVETIs were unlikely to become competent
due to the lack of necessary training materials and instructors. As illustrated
by a key informant from the MoH, colleges under RHB administration have a
relative advantage in getting resources and refresher trainings compared to
colleges under TVET-agency administration.
page- 727
A different scenario was also reported, however, in which HEWs were competent
“
enough to execute their tasks. One department head reported this as follows:
The competence level of the trainees who have graduated from

our college has been proven not only by our Health Extension
department but also by the regional and zonal organizations. They
are the primary choices of HCs throughout the region. They have
shown outstanding performance in terms of being disciplined in the
workplace and delivering quality service to the community in the
clinics. This shows that students who graduated from Teda College
are competent in all aspects.
In general, the findings show serious concerns about the competence level of
HEWs. The professional competence of HEWs is evaluated through COC
examinations, which are centrally developed by the MoE and disseminated to
regional COC centers for the evaluation of candidates’ competence. Findings
from the assessment of the procedure and the results of COC exams also
indicate problems in the candidates’ competence.
Table 12: Number of attempts necessary to pass the COC exam
Number of attempts Number Percent

1 526 66.8
2 213 29.3
3 36 4.6
>3 13 1.6
A total of 788 trainees who participated in this assessment across the 21

colleges in Ethiopia had already taken the COC exam at their previous level.
Of these, 66.8% reported having passed on their first attempt, while 29.3% and
4.6% indicated that they passed only during their second and third attempts,
respectively. One surprising finding was that 13 (2%) respondents had attempted
the exam more than 3 times, despite the COC exam rule limiting attempts to
a maximum of 3.
As explained above, the majority of the students scored a passing mark on the
COC examinations they took in their respective regions on their first attempt.
page-728
This does not, however, imply that HEWs are competent in general. The students
who apply for the upgrading program are relatively competent and confident
of passing the competence tests. If the assessment was taken from the general
population, the result would have been different. Furthermore, because the
assessment process is vulnerable to manipulation, passing may not necessarily
indicate that the candidate has competence.
3.6 The Existence and Adequacy of Professional

Development Activities for HEWs
Activities concerning HEWs’ professional development are generally classified
in 2 categories: in-service training and continuous professional development
(CPD). Integrated refresher training (IRT) is one variety of in-service training,
but in this assessment the terms “in-service training” and “IRT” are used
interchangeably.
3.6.1 In-Service Training
The main purpose of the in-service training is to offer training to HEWs with
a view to improving their knowledge and skills in order to improve the quality
of services at the HP and household levels. The IRT implementation guidelines
and plan explain the need for IRT:
page- 729
“
Following the initial rounds of pre-service trainings for HEWs (EFY
1996/97 EFY), gaps in HEWs’ skills and service delivery were
identified. It was seen that the pre-service training was not complete
in addressing the core competencies expected by the HEP. As a result,
HEWs were frequently pulled out of their posts for various trainings
by the RHB and its partners. The result was an uncoordinated effort
to address deficiencies in HEP performance.3
The IRT was therefore launched to coordinate the efforts of the
different parties and standardize the in-service training offered to
HEWs. To this end, the National In-Service Training Directive for
the Health Secto21 specifies the approach to follow when conducting
the in-service training. Among other points, it states that the “in-
service training courses should encompass integrated practical and
theoretical components in order to enhance the quality of health
services.” 5
According to the IRT implementation guidelines,8the modules covered in the IRT

are immunization, TB/HIV, malaria, integrated community case management/
integrated management of newborn and childhood illness (ICCM/IMNCI) and
family planning. The curriculum was developed by specialist programmers at
the MoH and reviewed and endorsed by the in-service/continuous professional
development (IST/CPD) team and Human Resource Development Directorate
at the MoH. The curriculum is accompanied by manuals for trainees and
trainers. The training is offered by public health professionals who have received
training of trainers on the training modules and are recruited mainly from HCs
and WorHOs or DHOs. Most trainings have been managed at the woreda or
district and HC levels.
A survey of 584 randomly selected HEWs from all regions of the country,
conducted along with this assessment as part of the general assessment of the
HEP, showed that the refresher training has been offered to HEWs every year
for 5 years (2007-2011 EFY). The survey asked the HEWs whether they had
received IRT thus far. In response, 511 (87.5%) of the 584 HEWs confirmed that
they had received IRT at least once.
The HEWs were also asked to confirm their participation on a yearly basis for
a period of 5 years (2007-2011 EFY). Their responses are summarized in Table
13. The data in the table show that every year some HEWs receive in-service
page-730
training. More specifically, 31.5%, 28.1%, 26.5%, 25.0%, and 19.7% of the 584
sample HEWs had received in-service training in 2007, 2008, 2009, 2010, or
2011 (EFY), respectively.
Table 13: HEW Participation in Integrated Refresher Training (2007-2011 EFY)
EFY Status of participation Frequency Percentage

Yes 184 31.5
2007 No 400 68.5
Total 584 100.0
Yes 164 28.1
2008 No 420 71.9
Total 584 100.0
Yes 155 26.5
2009 No 429 73.5
Total 584 100.0
Yes 146 25.0
2010 No 438 75.0
Total 584 100.0
Yes 115 19.7
2011 No 469 80.3
Total 584 100.0
Further analysis of the same survey data shows that, over this 5-year period,
some HEWs participated in the refresher training more than once. The data
pertaining to the HEWs’ frequency of participation over the 5-year period are
presented in Table 14.
Our data indicate that the frequency of participation ranges from not at all
(i.e., 0) to 5 times (median 1). The largest segment of HEWs (38.0%) had not
participated in refresher training at all, followed by those who had participated
either once (27.4%) or twice (15.1%). About 19.5% of the surveyed HEWs had
participated in the refresher training 3, 4, or 5 times.
page- 731
Table 14: Frequency of HEW Participation in In-Service Training (2007-2011

EFY)
Frequency of Participation N Percentage

0 222 38.0
1 160 27.4
2 88 15.1
3 54 9.2
4 34 5.8
5 26 4.5
Total 584 100.0
According to a key informant from the MoH, every HEW is expected to

participate in IRT once every 2 years. Accordingly, every HEW should have
participated at least twice between 2007 and 2011 (EFY). Only 24.3% of the
sample, however, had participated either 2 or 3 times. It is unclear why a
minority (19.5%) of the HEWs had the chance to participate in IRT 3 to 5 times
when the majority did not get this chance at all (38%) or participated in IRT
only once (27.4%). Admittedly, those HEWs in their first year of service must
work for 1 more year to become eligible for the IRT and thus cannot participate
in the IRT; this might explain why at least some HEWs in the sample did not
participate in the IRT, but the data indicate that only 16.6% of the surveyed
HEWs had worked for less than 2 years, so this cannot fully explain why the
majority of the HEWs did not participate in the IRT.
One explanation could be that IRT participation is based on a needs assessment,

but a key informant from the MoH argued that this would not be the case and
that a more likely reason would be whether the HEWs worked in remote villages;
HEWs who are placed in remote areas would not get the chance to take IRT
as frequently as those in closer towns and villages would. He suggested that,
while those in nearby villages can closely follow the training announcements
and messages sent to them, those in remote villages do not have this chance.
The latter do not usually get the chance to participate in the IRT because they
cannot get the information on time in their remote villages even if they are
selected by the WorHO, DHO, or RHB.
On the whole, there is some reason to believe that the IRT is not implemented
according to the guidelines for in-service training. For example, even though the
page-732
National In-Service Training Directive for the Health Sector21 clearly outlines
the roles and responsibilities of several stakeholders, it appears that the
stakeholders have not fully discharged their respective responsibilities. Among
other points, the directive indicates that the MoH shall “establish and maintain
a national IST database and monitor and evaluate the IST program. 1” In the
same manner, the RHB shall “conduct an annual training needs assessment at
the regional level and use the findings for the following annual planning and
establish and maintain a regional IST database.” 2,9,10 The directive further
states that “the national and regional IST database shall be updated every 6
month[s] and sent to training institutions by the MoH and/or RHBs.”8
In general, it appears that no party has played its role fully, and most of
the roles indicated had not been put into practice. Clear evidence for this
argument is the fact that there is no IST database whatsoever either at the
federal or regional level. The authors undertook extensive efforts to secure data
about the trainees and instructors of the HEW training colleges for 2011 EFY
and know firsthand that securing organized data from the official records of
colleges, RHBs, WorHOs, DHOs and even from the MoH is extremely difficult,
if not impossible. In other words, there is no database as envisioned. In addition,
with regard to monitoring and evaluation, it seems that the MoH can only
provide technical and financial support to RHBs and could not conduct robust
monitoring and evaluation even if it wanted to because of the decentralized
administrative structure, which limits the power of the MoH to carry out
monitoring and evaluation at the regional, woreda, or district levels. Thus, much
remains to be done in relation to both the IST database and monitoring and
evaluation at both the federal and regional levels.
3.6.2 Continuous Professional Development (CPD)
The second type of professional development is CPD. According to the Directive
“
on Continuing Professional Development for Health Professionals,
Continuing Professional Development” shall mean a range of

learning activities through which health professionals maintain and
develop throughout their careers to ensure that they retain their
capacity to practice safely, effectively, and legally within their scope
of practice.3,20
page- 733
The CPD directive also outlines important principles to which the implementation
“
of CPD should adhere. Most important among these are
1. No health professional shall be re-licensed without fulfilling the

required continuing education unit by this directive;
2. No continuing professional development course or activity shall
be counted unless accredited by Accreditors and endorsed by
Ministry; and
4. Health practitioners will commit themselves to meeting the
requirements for continuing education in the belief that the
practitioner and patient/client will reap the benefits of lifelong
learning.4,20
According to an interview with a key informant from the MoH, CPD has not yet
become functional. Thus, unlike with IST or IRT, we were unable to conduct a full
evaluation of the functionality of the CPD. While the IRT is almost exclusively
organized by the WorHO, DHO or RHB, the CPD directive shows that the
health practitioner (in this case, the HEW) shall take courses, participate in
different kinds of training and workshops organized or offered by accredited
organizations, and fulfill the required credits called a continuing education unit.
The HEW will then submit evidence that she has met the requirements and
applied for certification. Thus, the individual HEW is responsible for her own
CPD. Finally, while the plan and directive of the CPD appear promising, we
must wait to see whether this mechanism will turn out to be an effective way of
facilitating the CPD of HEWs.
3.6.3 Concerns about the Professional Development of HEP

Instructors
Some key informants confirmed that HEWs often participate in short-term

trainings during their professional practice, but there is a strong belief among
the key informants that the professional development of instructors—a significant
factor in HEP training—is very limited and, as a result, the instructors’ knowledge
is dated.
Additionally, instructors report that the RHB often excludes HEP instructors
from HEP-related training even though the instructors desperately need such
page-734
training to update themselves and become confident in their teaching. HEP

instructors reported that they often received training on other topics, such
as life skills and teaching methods, but because their major responsibility is
teaching HEP modules and because some of them have been teaching for a
long time without any chance for a refresher training on HEP-related topics,
several key informants argued that the training on health in general and HEWs
in particular would help HEP instructors in terms of not only updating their
knowledge but also improving their confidence in teaching. Key informants
believed that instructors were often excluded from HEW-related training simply
because they worked in colleges run by a TVET agency. Many recommended
that such activities on the part of the RHB be reconsidered and remedial
measures taken.
3.7 Challenges to and Opportunities in the HEW Training Program
There is sufficient evidence to support the idea that the HEP has registered
notable achievements in Ethiopia. One of the program’s achievements was
to train and deploy more than 35 000 HEWs in 15 000 HPs throughout the
country, including remote rural woredas and kebeles that have no access to
health services in hospitals or HCs 22. Additional evidence of these significant
achievements observed in the health sector is shown in Table 15 below and can
be attributed, at least in part, to the HEP because the indicators are directly
related to the HEP packages.
Table 15: Progress in Key Health Indicators, Ethiopia, 2005 and 2016 EFY
Heath Outcome/Output 2005 2016 Change

Under-5 mortality rate 123 67 -45.53%
Maternal mortality ratio 673 412 -38.78%
Infant mortality rate 77 48 -37.66%
Toilet facilities 38 41 7.89%
Total fertility rate 5.4 4.6 -24.81%
Contraceptive prevalence rate 15 36 140%
Source: DHS, 2005; 2016.
Despite the achievements and opportunities created by the program, there

were many challenges. Both opportunities and challenges are presented in brief
below.
page- 735
3.7.1 Opportunities
We identified existing facilitators of successful HEW training, including policies,

cost-effectiveness, infrastructure, career development, and partner involvement.
EXISTENCE OF A STRONG POLICY BASE
One facilitator that could provide momentum for the effective implementation
of the HEW training program is the existing strong policy base in which the HEP
is considered a vital strategy to effectively implement widespread, integrated
primary care in rural Ethiopia.
COST-EFFECTIVENESS OF THE PROGRAM
The contributions and achievements of HEWs are now well recognized by

stakeholders as a vital strategy to reach the majority of the population cost-
effectively. Key informants identified HEWs as effective in reaching the wider
community compared to other more expensive professionals, like nurses and
midwives. Achievements in various health indicators suggest that functioning
“
HEWs lead to cost savings. One COC assessor also made this connection:
There are changes brought at a national level due to this

program. For example, using toilets, handwashing, and timely child
vaccinations are now becoming a tradition at the household level as
a result of the influence of the program. If you compare the current
status of the malaria epidemic with the situation that existed some
5 or 6 years before, it is highly reduced. These changes are due
to the presence of the HEWs in all kebeles and their hard work in
health-promotion activities in the community.
EXISTENCE OF INFRASTRUCTURE
The existing infrastructure of physical facilities, human resources, and

administrative entities could also be used effectively to deliver training. There
page-736
are HPs, HCs, and hospitals throughout the country that could facilitate
trainings and colleges, as well as trainers and RHBs in every region that could
also facilitate the training effectively.
EXISTENCE OF CAREER DEVELOPMENT STRUCTURE
Recently, the MoH has taken the initiative to develop a career-development

structure that allows HEWs to be upgraded to a higher-level health service
provider. HEWs can pursue their education by studying family health at the
bachelor’s-degree level. (Currently, 8 public universities have launched the
program and admitted 240 HEWs as students in family health.) This initiative
could serve as a good opportunity to motivate current HEWs, candidates, and
potential HEW applicants.
PARTNER INVOLVEMENT AND SUPPORT
A number of partner organizations have been supporting the program and have
great interest in continuing their support. They have been playing a substantial
role in coordinating curriculum reviews, preparing training modules, providing
equipment for skill labs, and organizing in-service training for HEWs. Key
informants mentioned JSI (working on the urban HEP), AMREF (engaged in
capacity-building and module preparation), JHIPEGO (supporting pre-service
training), and the CDC (working on capacity-building in HIV). Together with
local organizations, these partners have contributed and are still contributing
to the successful implementation of the HEP.
3.7.2 Challenges
Briefly, the challenges fall into 4 categories:

1. curriculum, learning materials and facilities;
2. students and instructors;
3. teaching methods; and
4. the assessment and competence of trainees.
page- 737
CURRICULUM, TEACHING MATERIALS, AND FACILITIES
While materials and curricula are mostly relevant and appropriate for the
HEP’s objectives and goals, the scope of the curriculum, the time allocated for
each course, the lack of periodic revision of training materials, and the lack of
proper training facilities create inconveniences and ineffectiveness.
INSTRUCTORS AND TRAINEES
Instructors have the educational qualifications required to deliver courses, but

many do not have the specialization required to teach specific modules. Many
instructors were clinical nursing professionals despite only being able to properly
handle limited courses. In addition, staff shortages, the turnover of experienced
instructors, and the limited IST opportunities for HEP instructors were frequently
cited as challenges that could compromise the program’s quality.
Critical problems were observed from the student side as well. Some trainees
selected for the program were not officially eligible (i.e., the highest education
level they attained was below grade 10). Others who had completed grade 10
or grade 12 had limited educational competence to cope with the nature and
pace of instruction in the HEP. Still others had an English-language deficiency
and limited interest in the HEP. Moreover, organizing the HEW training
program under the TVET agency appeared to impede student competence
because the agency is thought to prioritize independent technical education
over HEW training.
TEACHING METHODS AND ASSESSMENT
The teaching methods and assessment procedure were affected by instructor

emphasis on: (1) theoretical over practical components, (2) limited teaching
methods due to time constraints; (3) disregard for importance of apprenticeships
conducted in cooperation with health institutions, and (4) unreliable assessment
procedures of student competence, including instructors arbitrarily increasing
student test scores.
page-738
COMPETENCE OF TRAINEES
In general, candidates were not perceived as meeting the competence

standards stated in the curriculum. This could be attributed to the selection of
the least competent candidates for the program, the lack of time allotted for
proper theoretical and practical training, trainees’ lack of English-language
comprehension, or ill-equipped facilities.
Figure 8: HEW training program challenges
page- 739
page-740
4 Summary and Conclusion
This study has assessed the quality of the HEW training programs with a focus
on 7 areas:
1. the relevance of the curricula;
2. the adequacy of the resources;
3. the quality of the course materials;
4. the quality of course delivery and student assessment methods;
5. the competence of HEW graduates;
6. the existence and adequacy of professional-development activities; and
7. the challenges to and facilitators of the training of HEWs in Ethiopia.
To achieve these objectives, we employed a mixed-methods approach. HEP

trainees (N=1245) and HEP instructors (N=192) completed quantitative
surveys, and 43 key informants were qualitatively interviewed. Data were
analyzed and are summarized below.
4.1 Summary of Major Findings

4.1.1 Relevance of the curricula
We used 4 criteria to assess the relevance of the HEP curricula:
1. Are the curricula in line with the country’s health policy?
2. Are the curricula consistent with the actual tasks HEWs are expected to
carry out in the communities?
3. Are the curricula implemented consistently with the guidelines?
4. Is the balance between the theoretical and practical components of the
curricula maintained in practice as proposed in the curricula?
First, the HEP curricula are in line with Ethiopia’s health policy in general and
the second HSDP in particular. Second, the HEP curricula are consistent with
the actual tasks that the HEWs are expected to carry out in the communities
when they begin their jobs after graduation. Third, the implementation of the
page- 741
curricula was problematic—the time allotment was not well-aligned with the
content expectations—but there were positive aspects as well (e.g., the program
recruits female HEWs in all sites except in pastoralist areas, where the HEWs are
male, and the urban, rural, and pastoralist HEWs were developed in response
to the diverse communities in the country). Thus, with regard to implementation,
the curricula are partly relevant. Fourth, maintaining the balance between the
theoretical and practical components of the curricula was found to be difficult.
The practical component in particular is problematic: there is an absence of
skill labs, a shortage or lack of equipment in skill labs, and poor supervision of
the trainees during their apprenticeships.
4.1.2 Adequacy and Availability of Resources
Human and material resources are among the essential requirements for
quality training. The colleges were found to have a relatively good number
of instructors with the required qualification (67.7% with a bachelor’s degree
and 27% with a master’s degree), but the distribution across specializations
is significantly skewed; e.g., there are a relatively large number of nurses,
midwives, and health officers across all colleges, but in 2018/2019 there was
no lab technician or technologist in 5 colleges (Minilik, Metu, Negele Borena,
Dire Dawa, and Hosana), while in other colleges the number of lab technicians
ranged from 2 to 19 (median 8).
While most college have some of the necessary facilities (classrooms, library,
computer lab, skill lab, practice sites, toilets, and water point), only a few had
enough of each. For example, class sizes in some colleges could be as large as
70 or 80, far higher than the standard of 25-30 per classroom.
4.1.3 Quality of course materials (curricula, modules, references)

for HEWs training
The teaching modules specifically developed for training are found to be up

to the standard and relevant, but they did have flaws, including an advanced
scope beyond their intended purpose, an absence of periodic revision in some
practices, and inadequate time allocation.
page-742
4.1.4 Quality of course delivery and student assessment methods
We assessed the relevance of the teaching modality, the professional competence

of the course instructors, student recruitment, methods of teaching, and methods
of assessment to the program objectives.
We found that the modular method of course delivery is relevant to HEW

training, but the effectiveness of this modality is hampered by a shortage of
training time and an inadequacy of training materials. While most instructors
are qualified and relevant, in some cases they deliver courses outside their areas
of specialization, and shortages or turnover of faculty limit effective instruction.
Large class sizes and the high workload of both instructors and students were
also identified as hindrances to student learning.
The major problem observed in this assessment was a poor candidate-selection

procedure, resulting in trainees who were academically less qualified enrolled
at the colleges. These students were challenged by learning the material in
English through a primarily theoretical, didactic teaching approach with limited
practical opportunities at either colleges or health institutions. They were then
evaluated primarily through summative written tests that gave inadequate
focus to practical performance evaluation or continuous assessment.
4.1.5 Competence of graduates from HEW training institutions
Although many instructors had positive comments about trainee competence in

some areas, they were concerned about other areas, and almost all recognized
that the HEP training program was challenged to produce competent health
professionals in all areas. They attributed the problem to a number of factors,
including biased and problematic recruitment procedures, English-language
incompetence, controversy over the ownership of the program, and the late
admission of candidates. Trainees who study in colleges under the administration
of the RHB are found to have better competence than those from colleges
under the auspices of TVET agencies. Similarly, upgrade trainees demonstrated
better competence than generic trainees.
page- 743
4.1.6 Professional development activities
HEW professional development can be either IST/IRT or CPD. The main

purpose of IST is to improve the knowledge and skills of existing HEWs and the
HP and household service quality. IST has reached many HEWs, though not
all, but our findings indicate that CPD is not yet functional in Ethiopia. Another
concern is that college instructors do not receive training as a part of their own
professional development; this may limit their competence in and commitment
to the program.
4.2 Conclusions
Based on the findings discussed in the preceding section, the following conclusions
are drawn.
• The HEW training curricula were found to be relevant to addressing
the major problems and gaps identified in the provision of health
service. The teaching modules were found to be relevant to and
appropriate for delivering the expected learning competencies. Still,
the broad scope of the curricula, the absence of periodic revision in
some key areas, and the poor module distribution among the trainees
in which the student-module ratio is not 1:1 are major problems in
relation to the curriculum.
• Although improvements have been made since the launch of the
program, institutions providing training suffer from a lack of the
necessary facilities and equipment critical to delivering quality
training for HEP trainees.
• Training modules are workable but include outdated concepts in
some courses and modules. The main challenge in some modules is
the imbalance between the module scope and the time allotted to
cover all content.
• Modular teaching is relevant to and appropriate for delivering
courses to HEWs, but the scope and depth of some modules are
beyond the students’ learning capabilities and pace.
• Although some trainees positively described teaching methods and
the HEW training program promotes a practice-based teaching
approach, in many cases, instructors deliver the courses using a
page-744
lecture method. Additionally, instructors are using the time allotted

for practical instruction to teach theoretical topics in class with a
limited focus on projects and collaborative learning.
• Instructors are educationally and professionally qualified to deliver
HEP courses, but some instructors deliver courses outside their areas
of specialization.
• Poor HEP trainee recruitment procedures (late or delayed and
biased procedures), trainees’ poor academic background, trainees’
poor English-language skills, and the deployment of under-qualified
candidates are primary challenges associated with the trainees.
These serious problems need urgent action.
• Instructors have little time to continuously assess or evaluate trainees’
practical skills and instead rely on summative paper-based exams
for trainee evaluation.
4.3 Recommendations
Based on the findings, we recommend the following measures, in order of

priority:
1. Reconsider recruitment procedures and criteria to attract better-
qualified candidates for HEW training programs. This can be done
through early recruitment and the administration of an entrance
exam to identify those candidates with better performance and
commitment or motivation. The recruitment procedure can be
improved by making the procedure participatory and involving the
colleges.
2. Train in local languages or provide English classes to solve the
language barriers inherent in HEW training programs. We strongly
recommend an in-depth study investigating the pros and cons of
each option, considering its effect on the teaching–learning process
on the one hand and the career or professional development and
progression of the HEWs on the other.
3. Maintain a balance between practical and theoretical training
components. This would require well-equipped demonstration
laboratories and laboratory assistants. Additionally, trainees’
apprenticeships need close monitoring by college administrators.
Supervisory instructors must be fully engaged in coaching and
supervising trainees’ day-to-day activities.
page- 745
4. Revise the curriculum, focusing on scope and time allotment for

modules to balance the core competencies of the curriculum and
students’ academic potential or abilities. While the curriculum
review and related policy decisions need further study and expert
evaluation, we recommend that the current core competencies be
divided into at least 2 HEP areas: (1) health education or promotion
and (2) clinical services to be provided at the HPs. This requires a
separate training program in each of these 2 HEP areas.
5. Make the necessary facilities and equipment available to the
colleges.
page-746
page- 747
References
1. Center for National Health Development in Ethiopia. Ethiopia Health

Extension Program Evaluation Study, 2005-2010. Addis Ababa:
Columbia University;2010.
2. Workneh MH, Bjune GA, Yimer SA. Assessment of health system
challenges and opportunities for possible integration of diabetes mellitus
and tuberculosis services in South-Eastern Amhara Region, Ethiopia: a
qualitative study. BMC health services research. 2016;16(1):1-11.
3. FMOH. Health Sector Transformation Plan. 2015.
4. FMOH. Quarterly Health Bulletin. Policy and Practice: Information for
Action. Addis Ababa2013.
5. Ghebreyesus TA. Achieving the health MDGs: country ownership in four
steps. The Lancet. 2010;9747(376):1127-1128.
6. Bilal NK, Herbst CH, Zhao F, Soucat A, Lemiere C. Health extension
workers in Ethiopia: improved access and coverage for the rural poor.
Yes Africa Can: Success Stiroes from a Dynamic Continent. 2011:433-
443.
7. FMOH. Health Extension Program in Ethiopia Profile. Addis Ababa2007.
8. Kitaw Y, Ye-Ebiyo Y, Said A, Desta H, Teklehaimanot A. Assessment of
the training of the first intake of health extension workers. The Ethiopian
Journal of Health Development (EJHD). 2007;21(3).
9. Medhanyie A, Spigt M, Dinant G, Blanco R. Knowledge and performance
of the Ethiopian health extension workers on antenatal and delivery
care: a cross-sectional study. Human resources for health. 2012;10(1):44.
10. Assefa Y, Gelaw YA, Hill PS, Taye BW, Van Damme W. Community health
extension program of Ethiopia, 2003–2018: successes and challenges
toward universal coverage for primary healthcare services. Globalization
and health. 2019;15(1):24.
page-748
11. Creswell JW. Educational research: planning. Conducting, and Evaluating.

2012.
12. Krejcie RV, Morgan DW. Determining sample size for research activities.
Educational and psychological measurement. 1970;30(3):607-610.
13. Cohen L, Monion L, Morris K. Research methods in education 5th ed.
London UK and New York. USA. Routeledge Falmer; 2000.
14. FMOH. Health Policy of the Transitional Government of Ethiopia. Addis
Ababa1993.
15. United States Agency for International Development. Health Extension
Program: An Innovative Solution to Public Health Challenges of Ethiopia:
A Case Study. 2012. Washington, DC: USAID2012.
16. FMoH. Model curriculum of the Health Extension Service. In:2006.
17. Ministry of Education, Ministry of Health. Ethiopian TVET System
Model Curriculum for Health Extension Services Level IV Based on
Occupational Standard. In:2018.
18. Ministry of Education, Federal Ministry of Health. Ethiopian TVET-
System Model Curriculum, Health Extension Service Level III based on
Occupational Standard. In:2013.
19. Ministry of Education. TVET Curriculum Development Manual Revision
4. In:2012.
20. FMOH. Directive on Continuing Professional Development for Health
Professionals. Addis Ababa2018.
21. FMOH. National In-Service Training Directive for the Health Sector.
Addis Ababa2014.
22. FMOH & UNICEF. HEWs’ Competence Evaluation Report for Rural
Ethiopia. Addis Ababa2010.
page- 749
page-750
Specific Study 2:
Attrition Rate of Health
Extension Workers in Ethiopia
page- 751
CONTENTS
Contents ---------------------------------------------------------------------------752
Executive Summary ---------------------------------------------------------------------758
1 INTRODUCTION -----------------------------------------------------------763
2 Objectives -------------------------------------------------------------------765
2.1 General Objective ----------------------------------------------765
2.2 Specific Objectives ---------------------------------------------------765
3 METHODS -------------------------------------------------------------767
3.1 Study setting and period ------------------------------------767
3.2 Study design --------------------------------------------------768
3.3 Study population -------------------------------------------------768
3.4 Sample size determination --------------------------------------769
3.5 Sampling procedure ------------------------------------------770
3.6 Operational definition ------------------------------------------771
3.7 Data collection procedure ----------------------------------------772
3.8 Study variables --------------------------------------------------772
3.8.1 Dependent variables ------------------------------------773
3.8.2 Independent variables -----------------------------773
3.9 Data management and analysis ---------------------------------773
page-752
4 RESULTS ------------------------------------------------------------------775
4.1 Characteristics of the study sites and data completeness ---775
4.2 Socio-demographic characteristics of the

Health Extension Workers --------------------------------------776
4.3 Magnitude of HEW attrition – Reports of WorHOs -----------782
4.4 Magnitude of HEW attrition – Findings from

review of personnel files ---------------------------------783
4.5 Trend of attrition among Health Extension

Workers in Ethiopia ------------------------------------887
4.6 Median time to attrition and incidence rate of attrition for

HEWs in Ethiopia ---------------------------------------892
4.7 Factors associated with Health Extension

Workers’ time to attrition --------------------------894
4.8 Causes of HEW attrition in Ethiopia ------------------------899
4.8.1 Psychosocial factors -------------------------------800
4.8.2 Administrative and/or structural factors ---------804
4.8.3 Salary and incentive packages -------------------819
4.8.4 Work environment related factors -------------------820
5 DISCUSSION -----------------------------------823
6 CONCLUSION AND RECOMMENDATIONS --------------829
7 REFERENCES ------------------------------------------------833
page- 753
LIST OF TABLES
Table 1: Distribution of study woredas and number of

files reviewed, by region ------------------------------------------775
Table 2: Socio-demographic and other characteristics of

Health Extension Workers ---------------------------------------------777

Health Extension Workers ---------------------------------------------778
Table 4: Recognition letter of Health Extension Workers -------------780

Health Extension Workers ----------------------------------------------780
Table 6: Description of HEW who have left the job in terms of certification
and current job -------------------------------------782
Table 7: Magnitude of attrition, by region and woreda,

with type of attrition of HEWs -----------------------------------784
Table 8: Magnitude of attrition, by different s

ocio-demographic variables ----------------------------------------------------785
Table 9: Magnitude of attrition, by certification status -------------786
Table 10: Magnitude of attrition by different background variables ---787
Table 11: Weighted number of Health Extension Workers attrition

per 10,000 by region from 2005 to 2018 -------------------------------------789
page-754
Table 12: Attrition of Health Extension Workers, by
woreda type, 2005-2018 ------------------------------------------------------790
Table 13: Incidence rate of attrition of HEWs at specified

time intervals [n=712] -------------------------------------------------792
Table 14: Probability of attrition by HEWs, by year ------------------794
Table 15: Variables selected using bivariate analysis for inclusion in

to the survival model -----------------------------------------------------------795
Table 16: Variables with a statistically significant association

with the time to attrition of HEWs -------------------------------------796
Table 17: Predictors of time to HEWs’ attrition ----------------------798
LIST OF FIGURES
Figure 1: Trends in weighted number of HEWs per 10000 who have been
upgraded their certification level at least one step up in Ethiopia,
2019[unweighted total number =859] ---------------------------------793
Figure 2: Attrition rate of HEWs, by region, as reported by

Woreda Health Offices ---------------------------------------------797
Figure 3: Attrition per 10 000 since the implementation of

HEP, 2005-2018 ----------------------------------------------------------802
page- 755
Figure 4: Weighted number of attrition of HEWs by livelihood
from 2005 to 2018 --------------------------------------------------------------805
Figure 5: Cumulative percentages of attritions of HEWs by

their year of employment as HEW [N=3476] --------------------805
Figure 6: HEW attrition incidence rate and service years, 2005-2018 --807
page-756
page- 757
Executive Summary
Background: The Health Extension Program (HEP) is a flagship program

launched by the Ethiopian government in 2004, with the aim of expanding
health care services nearer to the community. The HEP was first started in
agrarian settings, then expanded to pastoralist and urban areas. The program
encompasses promotive, preventive and basic curative services, which are
provided by trained and salaried Health Extension Workers (HEWs). In recent
years, there has been a sharp increase in the number of HEWs leaving their
jobs, which could strain the program and could potentially roll back the gains
made by the program so far. This survey was made with the objectives of
determining the prevalence, incidence rates, median time of attrition among
HEWs. Moreover trend of attrition and identifying factors associated with
attrition were the main objectives.
Methods: The study was conducted in nine regions and two city administrations.
A retrospective cohort study design was employed using the records of HEWs
(2004-2019). In addition, in-depth interviews with resigned HEWs were
conducted to explore the reasons for attrition. All HEWs who were registered
and deployed by Woreda Health Offices (WorHOs) were eligible for inclusion.
A random sample of the implementing woredas was taken proportionally
from all three types of woredas (agrarian, pastoralist, and urban). The cluster
sampling method was used to include all HEWs in the selected woredas. A
total of 85 woredas and 3 486 HEWs were included in the quantitative study,
while the sample size for the qualitative study was 18 HEWs who have left their
jobs. A data-extraction tool was developed by the study team, and trained
data collectors used tablet computers to extract data from the personnel files
of HEWs. The data’s quality was continuously monitored through daily calls
with each team and on-site supervision. Before data analysis, rigorous data
cleaning was performed to ensure the effective treatment of implausible or
inaccurate data. Descriptive statistics, such as measures of central tendencies
and variation, were used to describe the study population and estimate the
prevalence and trend of attrition. Additionally, analytic studies were conducted
to estimate the incidence, time to attrition and various factors associated
with attrition. The analytic study was mainly conducted using Kaplan Meier
estimation and Cox regression analysis. The qualitative data were analyzed
using thematic analysis.
page-758
Attrition Rate of Health Extension Workers in Ethiopia
Results: Data collection was conducted in 68 woredas, and a total of 3 476

HEWs’ personnel files were reviewed. The overall attrition rate was found to be
21.1% for the years 2004-2019. The lowest overall attrition rate was observed
in Harari (6.1%), followed by Dire Dawa (8.4%). The highest attrition rate was
recorded in Addis Ababa (38.5%), followed by Afar (36.1%). The HEWs left
their jobs in different ways. About 27.32% formally submitted a resignation
letter, while about 40% simply disappeared without notice. In addition, 3% of
attritions were due to death, while 4.7% were discharged from employment
for various reasons, including absenteeism and poor work performance. The
attrition rate for agrarian woredas was 18.7% and higher in pastoralist and
urban woredas (31.9% and 21.2%, respectively). Attrition began in 2005, just
one year after the HEP’s launch, and attrition increased every year except
2014, when it showed a slight drop. The increasing trend peaked in 2013 at
975 per 10 000 HEWs. The median time to attrition of the HEWs in this study
was 5.9 years, with an inter-quartile range of 2.9 to 8 years. The overall median
time of service of the HEWs included in the study, however, was 7 years [IQR
= 3 – 11.5]. The overall incidence rate of attrition is 288 per 10 000 person-
years [95% CI 267–309]. It was 182 per 10 00 person-years [95% CI 142-235]
in the first year. The highest incidence rate was seen during the 8th year of
implementation: 520 per 10 000 person-years [95% CI 421-642]. The factors
significantly associated with attrition were the HEWs’ age, place of birth, type
of deployment, current and deployment certification level, perceived distance of
the HP from the WorHO, having a child, and certificate of competence (COC)
status. Personal and work-related reasons were also identified as thematic
causes of attrition.
Conclusion: In general, in this study, an acceptable level of attrition was observed

among HEWs. Relatively higher level of attrition was observed in urban and
pastoralist settings. Ensuring job competitiveness in terms of payment and other
benefits is important to retain experienced and high-performing workers in the
program. At the moment, HEWs who leave the program tend to serve for
several years before leaving, but this may not be the case in the coming years,
as the trend of attrition has been increasing in last five years compared to the
early years of the HEP’s implementation. Factors like the HEWs’ age, marital
status, number of children, level of certification and others could help identify
those who will serve longer. Reasons for HEWs’s attrition include personal,
administrative, incentives, and workplace hardship related factors.
page- 759
Recommendation: Even though attrition rate among HEWs is within acceptable

ranges given the potential for replacement, retention actions should be in place
to ensure that better performing HEWs stay in the system. A comprehensive
package of benefits, linked to objectively measured performance, will help
in the retention of better performing HEWs. Recruitment of HEWs should
involve provision of adequate information about the nature of work of HEWs
to candidates.
page-760
page- 761
page-762
1 Introduction
The Ethiopian Health Extension Program (HEP) is a community-based health-

service delivery system aimed at improving access to the health care system,
with a strong focus on preventive and basic curative interventions. The main
strategy of the program is to increase health awareness at the level of the
community, thereby enabling households to help produce and protect their own
health. The program was launched in 2003 in the country’s four large agrarian
regions: Amhara, Oromia, the Southern Nations, Nationalities and Peoples
Region (SNNPR) and Tigray. After tailoring the program to the particular
requirements of the country’s other regions, the HEP was later expanded to
pastoral and semi-pastoral communities in 2006 and to urban areas in 2010.
The HEP is a primary health care service provided mostly free of charge with four
main components: disease prevention and control, family health, hygiene and
environmental sanitation, and health education and communication. It delivers
a package of basic and essential promotive, preventive, and curative health
services targeting households to improve the health status of families with the
active participation of particular households and the community. The goal of the
program is to improve household behaviors and provide basic health services
that have high impact and are cost-effective, such as improving sanitation
and personal hygiene, childhood vaccinations, family planning, prevention and
treatment of infections such as malaria, diarrhea, and pneumonia in under-five
children.
The approach of using female salaried community health workers called

Health Extension Workers (HEWs) to deliver the services of the program was
introduced nationally. HEWs are secondary-school graduates who receive a
one-year training in basic community-based health service delivery at technical
and vocational training institutions (TVETIs) or regional health science colleges
(RHSCs). They are selected based on a set of criteria from the communities
that they will serve. HEWs are expected to work at the health post (HP)
level for a quarter of their time and in the community through house-to-house
visits for the remaining three quarters of their time. Over 40 000 HEWs are
employed in Ethiopia, a fact that has significantly contributed to increasing the
coverage in health services.
page- 763
Ethiopia’s HEP is generally given great credit for improving access to health
care services at the grassroots level, an essential component of the efforts
made toward the goal of achieving universal health coverage. This, in turn, has
contributed to an improvement in the health outcomes of communities with
limited resources. The establishment of the health development army (HDA) has
empowered communities through participatory learning and action meetings to
discuss, decide on and take care of their health. Many households throughout
the country have graduated after getting and implementing mandatory training
on various aspects of health care promotion and prevention. The HEWs were
pivotal in all of these achievements. Due to the various challenges they face,
however, several studies and reports indicate that the attrition of HEWs has
increased in recent years. This could affect the implementation of the program
and may even roll back the progress made so far.
Hence, this study aims to identify the magnitude, attrition rate, and reasons
behind the mass exodus of HEWs. The study also tracks the whereabouts of
the HEWs after dropping out of the HEP. The findings of this study may help
policymakers identify recommendations at various levels of the health sector.
page-764
2 Objectives
2.1 General Objective
The general objective of this study was to describe the pattern of attrition,
associated factors, and reasons for attrition among HEWs in Ethiopia.
2.2 Specific Objectives

The specific objectives of this study were:
1. To assess the magnitude of attrition among HEWs in Ethiopia.
2. To determine the incidence rate of attrition among HEWs in Ethiopia.
3. To estimate the median time of attrition among HEWs in Ethiopia.
4. To assess the trend of attrition among HEWs in Ethiopia.
5. To assess factors associated with attrition among HEWs in Ethiopia .
6. To assess HEWs’ reasons for attrition.
page- 765
page-766
3 Methods
3.1 Study setting and period

This study covered all nine regions and the two city administrations of Ethiopia.
The regions are administratively divided in to zones, and zones into woredas.
The four regions (Amhara, Tigray, Benishangul-Gumuz and Gambella) are
agrarian regions. Afar and Somali regions are pastoralists regions. Oromia
and SNNP regions have mixed livelihoods (agrarian and pastoralist). In Addis
Ababa, the city administration is divided into sub-cities and then further into
woredas. In Dire Dawa the city administration is divided into kebeles. The way
HEWs deployed are different in different regions. In Afar, Oromia, SNNPR,
Amhara, Tigray, Benishangul-Gumuz, and Gambella the deployment is made at
the woreda level (i.e. their personnel file is found in the WorHO). In Somali and
Dire Dawa, deployment is conducted by the Regional Health Bureau (RHBs).
In Addis Ababa the HEWs are deployed at the Health Center (HC) level.
A recent estimate shows that Ethiopia has a total population of around 100
million. Addis Ababa is the capital city of Ethiopia, with a projected total
population of more than 3.6 million. According to the reports of the Ministry of
Health, HEP started in 2004 in four regions (Amhara, Oromia, Tigray and the
SNNPR). The program was later launched in pastoralist and urban settings,
after the original program was customized to suit these settings. To facilitate
the implementation of the program, more than 17, 000 Health Posts (HPs) were
built by the government, local people and occasionally by non-governmental
organizations (NGOs). The program has trained and deployed close to 40,
000 HEWs. Average of two HEWs per HP in rural areas and more than two in
urban HPs were assigned. In most of the agrarian areas of implementation, the
HEWs are female, as recommended by the program implementation guideline,
but in parts of the pastoralist and semi-agrarian areas, some HEWs are male
due to a shortage of eligible female candidates. Data were collected from June
to July 2019.
page- 767
3.2 Study design

A retrospective cohort study was employed to review the records of HEWs from
2004 to 2019. To supplement the quantitative data a qualitative exploration
was conducted.
3.3 Study population

All HEWs ever deployed to the study woreda were included in this study. All
HEWs who have been registered by the study woredas’ human resources units,
RHBs and health centers were eligible for this study.
3.4 Sample size determination

The main objective of this study was to determine the proportion of attrition
among HEWs. Hence, the sample size was calculated using a single population
proportion formula with a 95% confidence interval (95% CI), proportion of
attrition 7.2% among HEWs and a 0.9% margin of error:
n= (Zα⁄2) pq
2
d2
Where: = proportion of attrition among HEWs in Ethiopia;

q= proportion of HEWs on the job; and
d2= desired degree of precision (margin of error).

Using this formula, the sample size, including a 10% non-response rate, was
3 486.
The study woredas were sampled from the list of woredas from the three strata
based on the type of HEP being implemented (i.e. Agrarian, Pastoralist or
Urban). This was aimed to ensure the proper representation of all types of
woredas. It was assumed that, on average, a woreda has a population of 60 000-
100 000. Given a 3 000-5 000 population per kebele, the number of kebeles in
a woreda was estimated as 20, each with 2 HEWs. Based on this assumption,
on average a single woreda may have deployed and retained an average of
page-768
45 HEWs, excluding woredas where additional HPs were constructed or where

HEWs were added due to a larger population. Taking this into account HEWs
from 64 woredas, 27 HCs and one RHB were planned to be included in the
study. In Addis Ababa (AA) HEPrs from 20 HCs and in Dire Dawa from
7HCs were selected. A total of 85 employers (64 weredas, 20 HCs & I City
administration) of HEW were included. Thirty-six resigned HEWs were
purposely identified for an in-depth interview with the aim of identifying
their reasons for resignation.
3.5 Sampling procedure

A cluster sampling procedure was used to select the study subjects. WorHOs
were selected within each regional or city administration. The detail selection
of weredas were presented under population and sampling in part 2 section 1.
The sources of data for review were human resources and financial documents,
such as hiring, transfer files, and payroll documents. A unique identification
number was assigned to each file, and a list of all HEWs were developed.
page- 769
3.6 Operational definition
The following operational definitions were applied in this study during the data-
collection process, analysis and the interpretation of results.
1. Attrition: Health Extension Workers/Professionals who have ceased to

provide the HEP service, starting from the day of deployment, for any
reason including their mode of separation by resignation, dismissal,
disappearance, death, retirement, change in qualification or transfer out.
2. Change in qualification: Health Extension Workers/Professionals who are

assigned to any higher position at an institution or facility based on their
educational achievement or high performance.
3. Separation by transfer out: Health Extension Workers/Professionals who

are transferred to another Health Post from their original Health Post
assignment due to family, health or related issues.
4. Separation by resignation: Health Extension Workers/Professionals who

have submitted an official letter for leaving their jobs as Health Extension
Workers/Professionals.
5. Separation by disappearance: Health Extension Workers/Professionals

who have left their jobs as Health Extension Workers/Professionals
without informing or submitting a resignation letter to the concerned
office or body.
6. Separation by dismissal: Health Extension Workers/Professionals who

have been fired from their jobs due to poor performance, poor discipline
or any other reason the officials consider inappropriate.
7. Separation by death: Health Extension Workers/Professionals who

have died from any cause while working as Health Extension Workers/
Professionals.
page-770
3.7 Data collection procedure

The main data source was the HEWs files located in the WorHO, HCs and
RHB. To ensure the validity and reliability of the data-collection tool, structured
questionnaires were developed by the study team. The questionnaire included
questions on socio-demographic variables, the remuneration and recognition
process, employment and salary history, career opportunities, factors that
contributed to their separation from the program and other questions. The
content and constructs of the structured questionnaire was validated by experts:
biostatisticians, epidemiologists and other public health professionals. In
addition, the structured questionnaires were pre-tested in one of the WorHOs,
and the issues identified during pre-testing were considered an input to improve
the questionnaire by rewording and/or rephrasing as necessary.
Trained research assistants/data-collectors were used to extract information from

HEWs’ files using a structured questionnaire. The human resource employees
from WorHOs also assisted in providing information. During the data-collection
process quality were ensured by field supervisors and the study team. A daily
conference call was made between investigators and field supervisors, and
immediate decisions made about any errors happening in the field.
Data were collected using Open Data Kit (ODK) software on tablets.
Supervisors monitored the data-collection and ensures the quality on a daily
basis. Moreover to improve the quality there were sessions of feedback among
the supervisors and data collectors. The collected data were uploaded to a
central server on a daily bases. There were also a central data quality team
that oversight the overall quality of data. Three forms of data-collection tools—
module one, module two and module three—were used to collect data from
the study sites. Qualitative interviews with resigned HEWs from their jobs were
interviewed.
page- 771
3.8 Study variables

3.8.1 Dependent variables
The primary outcome variable for this study was the time to attrition of the
HEWs. The event of interest is the presence or absence of attrition/dropout of
the HEWs from the HEP. This was a binary variable (Attrition) coded as 0 =
No and 1 = Yes.
3.8.1 Independent variables

Various variables were considered independent variables. These variables were
assumed to explain the variability in the attrition of HEWs in the country. Some
variables fall into the categories of individual-level socio-demographic indicators,
career and motivation or job performance. This study also considered factors
that could act as confounders affecting other variables and the interaction of
the various independent variables as effect-modifying variables.
3.9 Data management and analysis

Data completeness and consistency were checked during the data-collection
process. This helped to detect errors at an early stage and correct any missing
or incorrect information. Data-collection was performed using ODK and stored
on a central server. The final data were exported into MS Excel (Microsoft
Corporation, WA, USA) for further data cleaning and thorough review. STATA
version 14 (STATA Corporation, College Station, TX, USA) was used for most
of the statistical analyses. Descriptive analyses, including measures of central
tendency and variations, like mean, median, interquartile range, standard
deviation and proportion, were used to describe the study population. Marginal
and stratified frequencies were also used for the distribution of attrition across
different risk factors.
The overall magnitude of attrition was calculated by taking the proportion

HEWs who dropped out of the program from those who were deployed during
the whole period of the implementation. The trend of attrition was estimated to
understand the rate of attrition over the years. Survival analysis was employed
using the median time to attrition A p-value of 0.20 was used to select the
potential variables to be included in the multivariable model. Some critical
page-772
variables of priority were included in the model irrespective the p-value. Using
the selected variables, a multivariable Cox regression was used to identify
significant predictors of the outcome variable. A p-value of less than 0.05 was
regarded as statistically significant. The hazard ratio with 95% CI was reported.
A Kaplan Meier table was used to estimate the probability of attrition in
different periods. Log-rank tests compared the survival probability among
various groups of categorical variables.
Interactions between significant covariates were checked in the final model.

Various model diagnostics were employed to check for the goodness of fit of
the model. To see the presence of outliers and influential variables, DFBETAs
were generated and checked for values greater than one. The proportionality
assumption of Cox regression was checked using Schoenfeld residuals. An
overall global test and scaled Schoenfeld residual plots were computed for
each statistically significant predictor variables. Cox Snell residuals were used
to check the goodness of fit of the final model.
page- 773
page-774
4 Results
4.1 Characteristics of the study sites and data

completeness
Of the total 85 HEW deployers (64 woredas, 20 HC and 1 city administration)

77 responded to module one (the woreda-level questionnaire), 68 to module
two (file review) and 16 to module three (qualitative interview). Data in module
two or the file review were not collected from Benishangul-Gumuz or two
woredas from Gambella due to security issues, nor from Somali due to a lack
of personnel documents. Two woredas, one from Tigray and one from Addis
Ababa, declined to provide HEW files for review.
Table 1 illustrates the number of woredas included per region and the total
number of files reviewed.
Table 1: Distribution of study woredas and number of files reviewed, by region
Files reviewed
Response Response Files reviewed
Region by region
Sources to module to module (Unweighted
name (Weighted
one two frequency)
percentage)
Tigray 6 6 5 289 11.27
Afar 4 3 3 61 7.08
Amhara 10 10 10 762 15.17
Oromia 14 14 14 873 12.94
SNNPR 10 10 10 738 11.95
Gambella 4 2 2 105 4.28
Harari 4 4 4 49 11.05
Addis
20 19 19 408 16.66
Ababa
Dire Dawa 1 1 1 191 9.61
Somali 8 8 0 0 0
Benishangul- 4 0 0 0 0
Gimuz
Total 85 77 68 3 476 100%
page- 775
For module two, a total of 3 476 HEWs’ files were reviewed, giving a response
rate of 91%. Addis Ababa represented the largest share of files, 16.66%, and
Gambella the lowest, 4.28% (Table 1).
4.2 Socio-demographic characteristics of the Health

Extension Workers
A total of 3476 HEWs’ data were found. Among them, 96.01% were female.
More than half of the HEWs (1954, 56.42%) were recruited when the HEWs
are 20 - 24 years old. Most (95.64%) were newly deployed, while the rest
are transfers. Of the total HEWs, two thirds (2444, 63.79%) were born in
rural areas. At the time of their deployment, most (2697, 76.95%) were single,
and more than half (1611, 56.27%) got married after their deployment. Of the
HEWs, 2030 (57.31%) had at least one child (Table 2).
In terms of their education level before joining colleges for HEW certification, 66
(3.93%) of them had attended education below grade 10, while 3 399 (96.07%)
had an educational level of 10th grade and above. Most (2 008, 57.5%) were
able to speak three or more languages, while 1 360, 39.12%) spoke only two. A
majority (2 535, 63.76%) had a level III certification during their deployment, but
463 (19.16%) had a diploma-level certification. Of all HEWs reviewed, 46.98%,
11% and 4.80% had a level III, diploma and degree certification, respectively, at
the time of survey (Table 2).
page-776
Table 2: Socio-demographic and other characteristics of Health Extension

Workers
Unweighted
Variable Category Weighted percentage
frequency
Sex Male 85 3.99
Female 3 391 96.01
18-19 881 23.36
Age category 20-24 1 954 56.42
25-29 436 14.63
30 and above 205 5.58
Marital status at Single 2 697 76.95
deployment Married 734 22.01
Divorced/separated/widowed 29 1.04
Marital status after Married 1 611 56.27
deployment
Still single 1 092 43.73
Number of biological No children 1 267 22.01
children
One or more children 2 030 1.04
Education level of Below grade 10 66 56.27
HEWs
Grade 10 and above 3 399 43.73
Number of languages Speak one language 89 42.69

spoken Speak two languages 1 360 57.31
Speak three or more languages 2 008 3.39
Level I 184 96.07
Level II 65 3.38
Certification level at Level III 2 535 39.12
deployment
Level IV 221 57.5
Diploma nurse 463 6.58
Degree nurse/midwife 8 0.32
Furthermore, 159 (4.34%) of them had education in addition to HEW training.

A few (69, 2.61%) had two or three years of previous work experience before
joining the HEP. Only 14 (0.32%) had some kind of disability (visual, hearing or
physical). Seven HEWs (0.26%) had a driving license (Table 3).
page- 777

Workers
Unweighted
frequency
Level I 96 4.21
Level II 55 3.24
Current Certification Level III 1 935 46.98
level
Level IV 879 24.81
Diploma nurse 384 15.98
Degree nurse/midwife 111 4.79
Yes 159 4.34
Additional education
No 3 301 95.66
Yes 14 0.32
Disability
No 3 446 99.68
Yes 7 0.26
Driving license
No 3 453 99.74
Yes 69 2.61
Previous work experience
No 3 391 97.39
1-3 2 235 72.06
Family size
4 or more 1 224 27.94
Out of all HEWs, 2571 (73.67%) did not upgrade their education and still
working at the certificate level. The rest upgraded at least one level from their
initial level. Of those HEWs who were upgraded, 726 (82.98%) were upgraded
one level, 67 (8.6%) upgraded two levels, and 65 (8.2%) upgraded three levels.
The median time to upgrading was 8.23 years, with an IQR of 5.59-10.2 years
Number of HEWs that goes to upgrade training have been continuously

increasing starting from year 2011 with some degree of downward trend in
2018. (Figure 1).
page-778
Figure 1: Trends in weighted number of HEWs per 10000 who have

been upgraded their certification level at least one step up in Ethiopia,
2019[unweighted total number =859]
The median performance report was 80%, with an IQR of 76% to 86.2%. For
most HEWs (969, 59.39%), their performance was above average. For each
HEW who had been reviewed, the median salary was about 1 916 ETB, with an
IQR of 1 545.4-2 298 ETB.
The findings showed that there was a mechanism to encourage good performers
and innovators. Among the HEWs who received recognition, the majority (104,
96.5%) were awarded for their best work performance, 3 (1.8%) were awarded
for their innovative work. Regarding the type of recognition, most (81, 74.2%)
received a certificate, whereas 19 (15.1%) received monetary awards (Table 4).
The finding also showed that 916 (22.67%) had some history of administrative
reprimands or penalties due to absence from work (88.45%) or poor performance
(11.6%). Of those who received an administrative reprimand, 424 (50.06%)
had a salary reduction, and about 22 (0.49%) had a history of legal penalties
(Table 5).
page- 779
Table 4: Recognition letter of Health Extension Workers
Unweighted
frequency
No recognition 3 351 96.12
Recognition letter One or more
109 3.88
recognitions
Best performance 104 96.5
Reason for recognition Innovation 3 1.8
Other reasons 2 1.7
Certificate 81 74.2
Type of recognition In-cash 19 15.1
In-kind 9 10.7
The median distance of HPs from their respective WorHOs is 15 km, with an
IQR of 5-27 km. Some HPs, however, are more than 131 km far away from their
WorHO. Of the total participants, about 1 862 (64.74%) worked in HPs with an
average distance of less than 15 km from the WorHO (Table 5).
The largest share of HEWs (1 379, 49.16%) perceived that the HPs were near
the WorHO, 1 187 (30.57%) perceived that they were within a medium distance
and the remaining reported that their HPs were located in remote areas, far
from the WorHOs (Table 5).

Workers
Unweighted
frequency
No reprimand 2 545 77.33
Reprimand
One or more
915 22.67
reprimands
Low performance 103 9.15
Reason for reprimand Absenteeism from

785 88.45
work
Other 27 2.41
Reprimand includes salary Yes 424 49.94

reduction No 491 50.06
page-780
Unweighted
frequency
Reprimand includes Yes 76 6.3
dismissal No 839 93.7
Yes 22 0.49
Legal reprimand
No 3 438 99.51
Distance of HP to WorHO 15 km or less 1 862 64.74
>15 km 1 598 35.26
Perceived distance of HP Remote 894 20.27
from WorHO Medium 1 187 30.57
Near 1 379 49.16
COC status COC certified 2 347 67.02
Not certified 1 113 32.98
No annual leave 1 274 41.9
Received annual leave 1-3 annual leaves 1 312 34.94
More than three
873 23.16
annual leaves
Agrarian 2 505 51.36
Type of woreda
Pastoralist 323 11.32
Urban 648 37.32
The finding from this study concerning the competency of HEWs shows that
2 347 (67.02%) were COC-certified, while the rest (32.98%) were not certified.
Regarding annual leave, 1 274 (41.9%) received no annual leave throughout
their service as HEWs. Among the 58.1% who got annual leave, 34.94% had
taken one to three annual leaves, while 873 (23.16%) had taken four or more
annual leaves. Regarding the livelihood of the population in the areas where
HEWs were working, 2 505 (51.36%) of the HEWs worked in agrarian woredas,
while 323 (11.32%) and 648 (37.32%) worked in pastoralist and urban woredas,
respectively (Table 5).
Most of the HEWs 337 (38.00%) were at level III at time of exit from HEP, while
108 (19.73%) were on diploma level. Only one obtained a master’s degree. Of
those who left their jobs for various reasons, 292 (42.01%) remained within the
woreda. HEWs have joined different work streams after leaving their jobs. Of
the total attrition, 184 (29.44%) HEWs were promoted to health professional,
while 93 (11.93%) were engaged in their own private businesses (Table 6).
page- 781
Table 6: Description of HEW who have left the job in terms of certification and
current job
Weighted
Variable Category Unweighted frequency
percentage
Private business 93 11.93
Government employee in
184 29.44
health sector
Government employee not in
70 12.01
health sector
Job type
after exit Homemaker 61 7.32
NGO 18 1.98
Went abroad 32 4.26
Unknown 227 30.82
Other 21 3.33
Level I 37 7.68
Level II 14 3.82
Level III 337 38
Certifications Level IV 76 7.26
at the time of
attrition Diploma nurse/midwife 108 17.97
Degree 83 16.64
Master’s degree 1 0.087
Unknown 71 8.54
Whereabouts Within the woreda 292 42.01
of the Elsewhere in the region 212 29.75
HEWs after
Elsewhere in Ethiopia 25 4.58
resignation
from their Abroad 32 4.26
work Unknown 143 1.94
4.3 Magnitude of HEW attrition – Reports of

WorHOs
WorHO report showed that a total of 3 800 HEWs were ever deployed in the
study woredas between 2004 and 2019; an average of 49 HEWs per woreda.
Of these 66.9% were actively working during the study, 10.65% were attending
further education and 22.44% were left the system. Attrition various across
regions which is highest in Addis Ababa, Afar and Amhara in a respective
order and smallest in Dire Dawa (Figure 2).
page-782
Figure 2: Attrition rate of HEWs, by region, as reported by Woreda Health

Offices
4.4 Magnitude of HEW attrition – Findings from

review of personnel files
In addition to reports from WorHOs, data on the current status of HEWs
was obtained by reviewing personnel files of ever deployed HEWs. Over the
15 years period a total of 3476 HEWs’ were ever deployed in the study sites.
The weighted overall attrition rate of HEWs within the 15 years was 21.1%. The
attrition rate varies across regions, with the lowest rate 6.12% in Harari and
the highest 38.48% in Addis Ababa. Among the HEWs who left their jobs,
27.32% formally submitted a resignation letter to their WorHOs, while 39.81%
simply disappeared without notice. A large share of attrition (31.99%) was from
pastoralist woredas, followed by urban woredas (21.15%)(Table 7).
page- 783
Table 7: Magnitude of attrition, by region and woreda, with type of attrition of

HEWs
(Weighted
Variable Category Unweighted frequency
percentage)
No 2 749 78.86
Attrition
Yes 727 21.14
Tigray 47 16.26
Afar 22 36.07
Amhara 193 25.33
Oromia 131 15.01
Attrition by region SNNPR 137 18.56
Gambella 21 20
Harari 3 6.12
Addis Ababa 157 38.48
Dire Dawa 16 8.38
National 727 21.14
Agrarian 463 18.73
Attrition by woreda Pastoralist 88 31.99
type
Urban 176 21.15
Separation by
191 27.32
resignation
Separation by
304 39.81
disappearance
Separation by death 23 2.86
Type of attrition Separation by
49 4.74
dismissal
Separation by
43 6.6
transfer
Changed
117 18.67
qualification
Of the total 727 HEWs who left their jobs, 23 (6.94%) were males. Most of the
HEWs who left their jobs (710, 98.02%) had been deployed as new HEWs, while
only 13 (1.98%) of them were transferred in from other posts. Among HEWs
who left their jobs, 416 (50.69%) were born in rural areas, and the remaining
295 (49.31%) were born in urban areas. In terms of the age categories of those
who left their jobs, 398 (54.95%) were aged 20 to 24 years, and 201 (25.67%)
were between 18 and 19 years. HEWs 30 years and older were the least likely
to leave their jobs.
page-784
Male HEWs (36.7%) resigns more than females (20.5%). Attrition decreases
as the age group increases. HEWs deployed as a new are more likely to resign
than transferred in HEWs. More HEWs who were married at deployment left
their jobs (21.7%) than did divorced HEWs (3.4%), while more of those HEWs
who remained single after deployment left their jobs (22%) than did those who
got married after their deployment (19.4%). Attrition is higher among HEWs
who were able to speak two languages (28%) who able to speak one or three
or more languages (Table 8).
Table 8: Magnitude of attrition, by different socio-demographic variables
Attrition Total
Variable Category No Yes Unweighted
Weighted % Weighted % number
Male 63.3 36.7 85
Sex
Female 79.5 20.5 3391
18 to19 76.7 23.3 881
20 to 24 79.4 20.6 1954
Age category
25 to29 77.5 22.5 436
30 and above 85.7 14.3 205
Transferred in 90.4 9.6 161
Deployment type
Deployed as new 78.3 21.7 3297
One language 87.4 12.6 89
Languages
Two languages 71.9 28 1360
spoken
More than two languages 83.8 16.2 2008
Urban 72 28 1016
Birth place
Rural 83.6 16.4 2444
Single 79.4 20.6 2697
Marital status at
Married 78.3 21.7 734
deployment
Divorce/separated/widowed 96.6 3.4 29
Marital status Married 80.6 19.4 1611
after deployment Still single 78 22 1092
Biological No child 75.9 24 1267
children One or more children 81.7 18.3 2030
One to three 78.8 21.1 2235
Family size
Four or more 80.7 19.2 1224
page- 785
HEWs with level III and IV certification at time of deployment resigned from
the HEP less than HEWs with below level III, diploma and degree HEWs.
Attrition is high among diploma holder HEWs at time of deployment which
is 34%. When attrition is observed with their current certification level HEWs
with level I and degree certification left their job greater than other levels of
status. The most lower attrition is found among level IV HEWs which is 8.6%.
The attrition in general decreases as we go from level I to level IV certification
level, however it increases as we go from level IV to degree level. Non certified
HEWs (32.4%) leave their job as two times of COC certified HEWs (14.9%)
(Table 9).
Table 9: Magnitude of attrition, by certification status
Attrition
Total Unweighted
Variable Category No Yes
number
Weighted % Weighted %
Highschool Below grade 10 70.0 30.0 66
education Grade 10 and above 79.5 20.5 3399
Level one 72.3 27.8 184
Level two 72.9 27.0 65
Certification level Level three 83.3 16.7 2535
at deployment Level four 83.9 16.0 221
Diploma nurse 66.0 34.0 463
Degree nurse/MW 67.7 32.2 8
Level one 60.4 39.6 96
Level two 70.6 29.4 55
Current certification Level three 81.0 18.9 1935
level Level four 91.4 8.6 879
Diploma Nurse/MW 71.3 28.7 384
Degree 50.4 49.6 111
COC status COC certified 85.1 14.9 2347
Not certified 67.6 32.4 1113
There were no major differences in in attrition rates among HEWs with different
workplace related experiences and perceptions. Previous work experience,
recognition, administrative reprimand and other workplace related experiences
seem to have almost no effect on the magnitude of attrition among HEWs
(Table 10)
page-786
Table 10: Magnitude of attrition by different background variables
Attrition
Total Unweighted
Variable Category No Yes
total
Weighted % Weighted %
Previous work Yes 77.1 22.9 69
experience No 79.4 20.6 3391
No recognition 79.4 20.6 3351
Recognition
One or more recognitions 79.1 20.9 109
Administrative No reprimand 80.0 19.9 2545
reprimand At least one reprimand 77.1 22.9 915
No annual leave 78.5 21.5 1274
Annual leave One to three annual leaves 81.9 19.0 1312
Four or more annual leaves 77.0 23.0 873
Distance of HP 1 – 15 km 79.1 20.9 1862
from WorHO More than 15KM 79.8 20.2 1598
Near 79.0 20.9 894
Perceived
Medium 81.5 18.5 1187
distance
Remote 77.0 23.0 1380
Legal Yes 63.6 36.4 22
punishment No 79.4 20.5 3438
4.5 Trend of attrition among Health Extension

Workers in Ethiopia
The HEP was started in the year 2004 in Ethiopia’s agrarian regions. According
to the data, the earliest attrition occurred the following year, after the initiation
of the program’s implementation. There was no attrition in the year 2006.
Nevertheless, after 2006, the trend of attrition kept steadily increasing except
in 2014, when it showed a slight drop. The increasing trend of attrition peaked
in 2013, at 975 per 10 000 HEWs, and slightly decreased to 688 HEWs per
10 000. After 2014, the trend of attrition rose consistently until it peaked at
almost 2 000 HEWs per 10 000 in 2018 (Figure 3). The data also showed a
slight decrease in attrition in 2019, but this wasn’t reflected in the analysis, as
the data were incomplete for this year. The data were collected in June 2019,
so complete figures on attrition cannot be found.
page- 787
Figure 3: Attrition per 10 000 since the implementation of HEP, 2005-2018

Table 11 shows the trend in the attrition rate of HEWs, disaggregated by regional
state. The earliest attrition occurred in Amhara in 2005, with a magnitude of
19 per 10 000 HEWS (2 HEWs). This means that the first cases of attrition
occurred within the first year of deployment. The second earliest attrition
occurred in the SNNPR and Dire Dawa in 2007, with 7.8 and 24 per 10 000
HEWs, respectively. Amhara and SNNP are regions affected by longest and
high strike of attrition. The shortest strike and lowest attrition were seen in
Harari, which is first registered in 2017 (Table 11)
The highest rates of attrition were seen in Addis Ababa, Amhara and Afar.
In 2016 in Addis Ababa there were highest number of attrition as compared
to other regions and years. There is an increasing trend in attrition which may
coincided with unrest in the country, which was peak in 2016 and 2017. Overall,
the attrition of HEWs was seen in all regions of the country (Table 11).
page-788
Table 11: Weighted number of Health Extension Workers attrition per 10,000 by region from 2005 to 2018
2011
2017
2013
2015
2018
2014
2012
2016
2010
Total
2007
2005
2008
2009
2006
Regional
attrition
Number 0 0 0 0 0 8 6 3 7 2 3 1 3 10 46
TIGRAY
Per 10 000 0 0 0 0 0 151 113 57 132 38 57 19 57 189 868
Number 0 0 0 0 2 2 2 1 2 1 0 2 3 5 22
AFAR
Per 10 000 0 0 0 0 112 112 112 56 112 56 0 112 169 281 1236
Number 2 0 2 5 13 11 6 18 32 21 23 13 16 24 189
AMHARA
Per 10 000 19 0 19 48 125 106 58 173 308 202 221 125 154 231 1820
Number 0 0 0 0 2 3 7 12 21 15 24 13 11 16 131
OROMIA
Per 10 000 0 0 0 0 14 22 50 86 151 108 172 93 79 115 939
Number 0 0 1 1 6 7 16 10 14 13 14 12 22 10 137
SNNPR
Per 10 000 0 0 7.8 7.8 47 55 125 78 110 102 110 94 172 78 1073
Number 0 0 0 0 0 0 0 0 1 0 3 6 2 7 19
GAMBELLA
Per 10 000 0 0 0 0 0 0 0 0 20 0 59 118 39 138 374
Number 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2
HARARI
Per 10 000 0 0 0 0 0 0 0 0 0 0 0 0 109 109 218
ADDIS Number 0 0 0 0 0 0 1 5 6 8 16 25 39 41 156

ABABA Per 10 000 0 0 0 0 0 0 20 99 119 158 316 494 77 81 3082
Number 0 0 0 1 1 2 1 4 1 1 1 1 1 2 16
DIRE DAWA
Per 10 000 0 0 0 24 24 49 24 97 24 24 24 24 24 49 389
Number 2 0 6 9 24 35 39 53 84 62 85 73 98 117 688
Total
Per 10 000 19 27 80 323 494 503 647 647 975 688 959 1080 1574 1999
page- 789

page-790
When the trend is disaggregated by type of woreda, the findings suggest that attrition occurred continuously
and in an increasing trend until 2018, except in 2006 where no attrition occurred. The overall magnitude
of attrition was 4 558 per 10 000 HEWs in the agrarian woredas. Consistent with the national trend, the
highest peaks have been seen since 2013, with the highest level, 743 per 10 000, seen in 2018. Attrition
in pastoralist woredas first occurred in 2007, a year after it occurred in agrarian regions, and steadily
increased until 2018, when it peaked with 289 per 10 000 HEWs. Attrition among urban woredas began
in 2010, with a magnitude of 49 per 10 000 HEWs. The highest estimated magnitude of attrition in 2018
was seen among urban woredas, with 968 per 10 000, but the overall highest magnitude was seen in the

agrarian woredas, followed by urban and pastoralist woredas, which showed the lowest overall HEW
attrition rates. Table 12 shows the trends in HEW attrition per 10 000 HEWs disaggregated by type of
woreda (Table 12).
Table 12: Attrition of Health Extension Workers, by woreda type, 2005-2018
Woreda
type
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Total
Number 2 0 2 6 19 26 27 41 64 41 61 38 48 66 456
Agrarian
Per 10 000 19 0 19 56 171 310 284 379 634 371 572 395 454 743 4 558
Number 0 0 1 0 4 5 10 3 13 11 6 9 9 6 88
Pastoralist
Per 10 000 0 0 7.8 0 128 136 175 72 199 135 47 167 216 289 1 753
Number 0 0 0 0 0 2 2 9 7 9 17 26 41 44 174
Urban
Per 10 000 0 0 0 0 0 049 44 196 143 182 340 518 904 968 3 689
Number 2 0 3 6 23 33 39 53 84 61 84 73 98 116 716
Total
Per 10 000 19 0 27 80 323 494 503 647 975 688 959 1 080 1 574 1 999
There was an increasing trend of attrition throughout the implementation years

of the HEP. Until 2014, agrarian woredas had the highest attrition every year.
But since 2015 attrition among urban woredas is greater than attrition (Figure
4).
Figure 4: Weighted number of attrition of HEWs by livelihood from

2005 to 2018
Cumulative percent of attrition of HEW by year of employment was the highest

among employees of 2005, 2009 and 2010. The trend of cumulative percent of
attrition is decreasing among recent employees, specifically stating from 2013
(Figure 5).
Figure 5: Cumulative percentages of attritions of HEWs by their year of

employment as HEW [N=3476]
page- 791
4.6 Median time to attrition and incidence rate of

attrition for HEWs in Ethiopia
The median time of attrition of HEWs in this study was 5.9 years, with an IQR
of 2.9-8 years. Among HEWs who left their jobs, 50% of them served six years
before leaving. The overall median time of service by HEWs was 7 years [IQR
= 3 – 11.5].
The overall incidence rate of attrition was found to be 288 per 10,000 person-
years [95% CI 267.23 – 309.52]. The rate of attrition, however, differs during
different time intervals during the implementation. For instance, the incidence
for the first year was 182 per 10,000 person-years [95% CI 142.97 – 234.52].
The highest incidence rate was seen during the interval of the 5th and 10th
years of implementation, which is 248.35 and 412.29 per 10,000 person-years.
The following table shows the incidence rate of attrition during the different
years of interval (Table 13).
Table 13: Incidence rate of attrition of HEWs at specified time intervals [n=712]
Cohort Number of Incidence rate

Person years [95% CI]
time interval attritions per 10,000 years
(0-1] 3 343 61 182.47 141.97, 234.52
(1-2] 3 052.9 66 216.19 169.85, 275.17
(2-3] 2 731.8 55 201.33 154.57, 262.23
(3-5] 4 751.3 118 248.35 207.35, 297.46
(5-10] 8 270.8 341 412.29 370.78, 458.46
(10-15] 2 606.6 71 272.39 215.86, 343.72
Overall 24 757 712 287.60 267.23, 309.52
The incidence of attrition among HEWs also varies among different categories
of HEWs. Incidence varies by HEWs’ COC status. The incidence rate of attrition
among HEWs who had no COC was 488 per 10 000 person-years [95% CI
442 - 539] compared to only 182 per 10000 person-years [95% 163 - 204]
among who are COC-certified. Similarly, the overall incidence rate differs by
type of woreda. The highest incidence, at 518 per 10 000 person-years was
page-792
seen among urban woredas, while the lowest, 230 per 10 000 person-years
was seen among agrarian woredas. The incidence rate of attrition among
HEWs who had no children (480 per 10 000 person-years [95% 427 - 539])
was more than twice that of those who had at least one child (217 per 10 000
person-years [95% CI 196.6 - 240]). In terms of recognition, the incidence rate
among those who received no recognition at all was 286 per 10 000 person
years [95% CI 265 - 308] compared to 197 per 10 000 person-years [95% CI
127 - 306] (Figure 6).
Figure 6: HEW attrition incidence rate and service years, 2005-2018
According to the results of this study, the probability of a HEW quitting his
or her assigned job in the first year after deployment is 1.81%; this probability
increases to 3.91% and 5.82% during the second and third years after
deployment, respectively. In the fifth year after deployment, the probability of
leaving increases to 10.4%, meaning one in ten HEWs leaves within the first
five years of deployment. By 10 years, about 27% of HEWs had quit their jobs,
and the probability reaches 36.5% in the 15th year from the initial deployment.
The probability of leaving early is higher among HEWs who have no COC
status, are unmarried during their initial deployment, have no children and who
have received administrative reprimands, including dismissal. Table 15 shows
the probabilities of attrition during each year of the HEP’s implementation.
page- 793
Table 14: Probability of attrition of HEWs by length of service
Years of service Probability of attrition [95% Confidence Interval]

1 0.02 0.01, 0.02
2 0.04 0.03, 0.05
3 0.06 0.05, 0.07
4 0.08 0.07, 0.09
5 0.10 0.09, 0.12
6 0.14 0.13, 0.15
7 0.18 0.16, 0.19
8 0.22 0.20, 0.24
9 0.25 0.23, 0.27
10 0.27 0.25, 0.29
11 0.29 0.27, 0.31
12 0.31 0.29, 0.33
13 0.33 0.31, 0.35
14 0.37 0.33, 0.40
15 0.37 0.33, 0.40
4.7 Factors associated with Health Extension

Workers’ time to attrition
Both bivariate and multiple regression analyses were performed to identify the
predictors of time to attrition of the HEWs. Bivariate analyses were conducted
using Kaplan-Meier estimates and a log-rank test for categorical variables and
a univariate Cox proportional hazards regression for continuous variables to
identify potential predictors that could be included in the model. The criterion
for selection was having cutoff point of a p-value less than or equal to 0.25.
Those variables having a p-value of less than 0.25 were selected for inclusion in
the multiple regression models. Based on this, a total of twenty-six independent
variables were considered, of which twenty were selected for inclusion, provided
that their P-values were less than 0.25. Table 15 shows the variables selected
with their corresponding bivariate analysis and p-values.
page-794
Table 15: Variables selected using bivariate analysis for inclusion in to the
survival model
Variable Chi2 [d.f.] P-value

Age category 6.38 [3] 0.0945
Place of birth 87.56 [1] < 0.0001
Sex 0.01 0.9308
Education before joining HEP 0.08 0.7770
Current certification level 343.05 [5] < 0.0001
Certification level at deployment 189.62 < 0.0001
Distance of HP to WorHO, in km 1.22 0.2699
Perceived distance of HP from woreda 13.61 [2] 0.0011
Woreda type 129.79 [2] < 0.0001
Recognition status 3.26 [1] 0.0710
Annual leave 47.50 [1] < 0.0001
Children 138.12 [1] < 0.0001
Family size 48.35 [2] < 0.0001
Type of deployment 22.58 [1] < 0.0001
Marital status at deployment 1.62 0.4452
Marital status after deployment 80.52 [1] < 0.0001
Number of languages spoken 37.24 [2] < 0.0001
Previous work experience 2.09 [1] 0.1480
Additional education 1.83 [1] 0.1762
Ever received reprimand 0.00 0.9510
Legal penalty 1.81 [1] 0.1779
COC status 181.74 < 0.0001
Upgrading status 0.04 0.8363
Year of deployment 77.18 < 0.0001
Variables considered for inclusion but not selected were sex, distance in km
from the HP to the WorHO, receiving an administrative reprimand, marital
status during deployment, education level, and being upgraded. These were not
significant enough at crude analysis. A multivariable Cox proportional hazards
regression was carried out using a stepwise selection method.
page- 795
Table 16: Variables with a statistically significant association with the time to
attrition of HEWs
Variables Categories Crude HR [95% CI] Adjusted HR [95% CI] P-value
18 -19 1.00 1.00 1.00
20 - 24 1.04[0.87 - 1.23] 0.82[0.68 - 0.99] <0.03
Age category
25 - 29 1.25[0.97 - 1.60] 0.71[0.53 - 0.93] <0.02
30 and above 0.79[0.53 - 1.11] 0.56[0.38 - 0.82] <0.01
Type of Deployed as new 1.00 1.00 1.00
deployment Promotion 2.51[1.71 - 3.68] 1.74[1.17 - 2.59] <0.01
Rural 1.00 1.00 1.00
Place of birth
Urban 0.49[0.43 - 0.57] 1.08[1.02 - 1.13]
Level one 1.00 1.00 1.00
Level two 1.14[0.65- 1.99] 0.96[0.80 -1.14] <0.64
Certification Level three 0 .94[0.71 - 1.26] 0.98[0.89 - 1.07] <0.67
level at
deployment Level four 0.66[0.41 - 1.04] 0.97[0.84- 1.13] <0.70
Diploma 3.00[2.18 - 4.13 0 .89[0.80 - 0.99] <0.03
Degree 3.55[1.11 - 11.39] 0 .76[0.46 -1.26] < 0.29
Level one 1.00 1.00 1.00
Level two 0.64[0.35 - 1.17] 0.93[0.77 - 1.13] <0.49
Current Level three 0.62[0.45 - 0.86] 0.98[0.89 - 1.09] <0.76
certification
level Level four 0.15[0.10 - 0.22] 1.23[1.08 - 1.39] <0.00
Diploma 1.46[1.02 - 2.08] 0.82[0.73 - 0.93] <0.00
Degree 1.47[0.99 - 2.18] 1.22 [1.06 -1.40] <0.01
Remote 1.00 1.00 1.00
Perceived
Medium 0 .70[0.57 - 0.85] 0.83[0.68 - 1.03] <0.09
distance of HP
Near 0.77[0.64 - 0.93] 0.80 [0.65 - 0.99] <0.04
Agrarian 1.00 1.00 1.00
Woreda type Pastoralist 2.28[1.81 - 2.87] 0.88[0.82 -0.95] <0.00
Urban 2.44[2.04 - 2.91] 0.89[0.84 - 0.95] <0.00
No 1.00 1.00 1.00
Having child
Yes 0.40[0.35 - 0.47] 1.16[1.09 - 1.21] <0.00
No 1.00 1.00 1.00
COC status
Yes 2.69[2.318 - 3.129] 1.84 [1.55 - 2.19] <0.00
page-796
Findings from Cox regression analysis are summarized in table 16, HEWs’ age,
place of birth, type of deployment, current and deployment certification level,
perceived distance of the HP from the WorHO, having a child, and COC
status were able to independently predict attrition. After controlling for the
other variables in the model, those HEWs who had a diploma certification at
deployment had an 11% lower chance of attrition [AHR = 0.89, 95% CI 0.800-
0.991] compared to those with a level I certification. The other categories had
no significant association with time to attrition. On the other hand, those HEWs
who had a level IV current certification had a 23% increased risk of leaving
[AHR=1.23, 95% CI 1.079-1.394] compared to those certified as level I. Similarly,
those with a diploma certification had an 18% lower risk of attrition [AHR =
0.82, 95% CI 0.727-0.934] compared to those who were level I certified after
adjusting for other variables. Likewise, those who had a degree-level certification
had a 22% higher risk of attrition compared to those who had a recent level I
certification [AHR=1.22 95% CI 1.059-1.402] The other categories—i.e., diploma
and degree certifications—had no statistically significant association with the
time factor to attrition (Table 16).
HEWs’ age was another independent predictor of time to attrition. Adjusting

for the other variables, HEWs in the age category 20-24 had an 18% lower
chance of leaving compared to those HEWs aged between 18 and 19 years
[AHR = 0.82 95% CI 0.685-0.986]. HEWs aged 25-29 and those aged 30 and
above had a 29% [AHR = 0.71 95%CI 0.535-0.934] and 44% [AHR = 0.56
95%CI 0.386-0.823] lower chance of leaving, respectively. The other statistically
significant predictor identified was the HEWs’ birthplace. Accordingly, holding
the other variables constant, those who were born in urban areas had an 8%
higher chance of attrition compared to those who were born in rural areas
[AHR = 1.08 95%CI 1.024-1.131]. Another independent predictor of attrition is
the type of deployment of HEWs. Those who were promoted to work as HEWs
had a 74% higher risk of leaving compared to those who were deployed as
HEWs originally after controlling for the other variables [AHR = 1.74 95% CI
1.170-2.587].
The perceived distance of the HP where the HEWs work from its WorHO
is another statically significant predictor of attrition. Holding other variables
constant, HEWs working in HPs perceived as being near the WorHO had a 20%
lower likelihood of leaving their work compared to those working in a remote HP
[AHR=0.8, 95% CI 0.647-0.998]. The category perceived as medium distance
had no statistically significant association with attrition. HEWs who had one
page- 797
or more children had a 16% [AHR= 1.16, 95% CI 1.094-1.219] higher probability
of attrition compared to those who had no children after adjusting for other
variables. Being COC-certified is another statically significant variable. After
controlling for all other variables in the final model, HEWs who were COC-
certified had an 84% [AHR=1.84, 95% CI 1.555-2.196] higher chance of leaving
their work compared to those who were not COC-certified. When the analysis
was made after taking into account the weight, only four variables—recent
certification level, having a child, receiving a reprimand with salary reduction
and dismissal—remained independent predictors. Table 17 shows the weighed
adjusted hazard ratios of the statistically significant variables after weighting.
Table 17: Predictors of time to HEWs’ attrition
Variable Category Weighted HR [95% Conf. Interval]
Urban 1
Birthplace
Rural 0.70 0.51, 0 .97
Level one 1.00
Level two 0.25 0.02, 3.22
Certification level at Level three 1.29 0.91, 1.84

deployment Level four 4.33 2.41, 7.79
Diploma 6.39 2.50, 16.33
Degree 4.59 0.91, 23.24
Level one 1.00
Level two 1.12 0.07, 19.17
Current certification Level three 0.59 0.29, 1.21
level Level four 0.12 0.06, 0.25
Diploma 0.43 0.11, 1.71
Degree 0.47 0.17, 1.27
No 1.00
Children
Yes 0.51 0.41, 0.65
Agrarian 1.00
Woreda type pastoralist 2.19 1.32, 3.62
Urban 0.39 0.21, 0.72
page-798
Based on the result of the weighted Cox proportional hazards regression,

after adjusting for birthplace, certification level during deployment, current
certification level, having a child and woreda type, those HEWs who were
born in rural areas were 30% less likely to leave their jobs than their urban
counterparts [AHR=0.70, 95% CI 0.51-0.97]. Those certified as level IV have a
71% lower likelihood of leaving their jobs compared to those certified as level
I [AHR=0.29, 95% CI 0.13-0.66]. HEWs who had one or more children had
a 59% lower chance of leaving their jobs compared to those with no children
[AHR = 0.41, 95% CI 0.28-0.61] (Table 17).
4.8 Causes of HEW attrition in Ethiopia

In order to understand the causes of HEWs attrition we have conducted focus
group discussion (FGD) and in-depth interview with former HEWs. The FGD
includes both male and females former HEWs who left their job due to different
reasons. Most of the participants were level III HEWs. Some were level IV
HEWs.
The findings from the interviews with HEWs and FGD were consistent. After
analyzing both the interview data and the FGD, four major themes emerged
as causes of attrition: psychosocial factors, administrative or structural factors,
salary and incentive packages and work-related factors. The results below
represents the result from the qualitative findings.
page- 799
4.8.1 Psychosocial factors

The HEWs’ attrition was influenced by several types of factors. One is personal
factors, which may include a lack of skills, personal belief in their work, personal
conflicts, personal health-related issues, a lack of trust and respect from the
community and a lack of community respect for the HEP and/or HEWs. This
category could be seen in terms of psychological or personal and social factors.
Participants demonstrated that their personalities, values and attitudes varied

widely. The report revealed that HEWs entered into organizations with stable
or transient characteristics that affected how they behaved and performed.
Participants reported that their personal beliefs had influenced their expectations
to move to urban areas after employment through transfer or a search for
other job opportunities.
“
Some HEWs have an attitude problem. Because we are from rural
communities, we thought that we would get into the town where we
were recruited. But the reality is not that we would work with the
rural community after the training. So, since they work in the rural
areas, they develop the attitude that they did not improve. But if
they think that they are working to benefit society, they would not
think of leaving the job for a second, if they really got into it to help
the community. But, as I mostly observed, most of them need to live
and work in urban areas.
HEW, Amhara Region
The other issue raised by respondents is the personal interest in working as

HEWs. Respondents stated that they had both positive and negative impressions
regarding being HEWs. A respondent from Hawassa who had interest in the
profession said:
page-800
“
Regarding interest, it is good if there is success at work, especially if
there is a good relationship with the woreda, community and moral
support and a conducive working environment. The job gives you
the opportunity to interact with various people. And they ask you
many questions….Mothers used to ask us questions, and we used to
respond….Also, youths used to ask many questions everywhere they
met us that they might not ask in the HC or hospital….High-school
students also used to ask us questions about illnesses and things
they don’t know. It is very pleasing to satisfy someone who has a
question. Also, when we see mothers giving birth at health facilities,
it also used to satisfy us….The job is interesting when it brings the
desired answer, but if the work is kind of demoting it is hardly
interesting.
HEW, SNNP Region
On the other hand, a respondent from Addis Ababa who had a negative
impression of the profession and joined because it was the only opportunity at
“
that time, said:
Yes, actually I was not interested in being an HEW in the first

place. Because the things you see do not give you satisfaction.
For example, if you are interested in teaching about HIV/AIDS,
thanks to the internet everyone has that information. Thus, given
this reality, going home to home to teach about TB, or cancer or
other things is not that realistic. It is really very difficult. It would be
better to do that only when new things come and in selected areas
will be good. All the reports are false, and then they do not thank
you for what you did. I am telling you the truth.
HEW, Addis Ababa
page- 801
With respect to competence, participants also reported that those who leave
their jobs are those who have been challenged by their skills gap and developed
a negative attitude. Some participants expressed that their job skills were
not up-to-date and felt that this would hinder their career advancement and
chances of promotion accordingly. It is obvious that there is a disconnection
between HEWs and the workplace when it comes to the skills gap.
Program officers who participated in FGDs confirmed that there were HEWs
with problems related to competence and discipline:
“
The problem is that they don’t have good competence; we have said
clearly that we have no problem with attrition, rather not working
regularly and properly....There are those who do not know where
their health post is and have someone to rely on, those, who we even
cannot dare to touch them. Even now in the recruitment process,
we could not make them take exams. We accept who has been
sent. There are problems in competence. Given that incompetence,
they pass the COC, but they could not understand the English
language....
Program Officer
Respondents stated that family conflict happens because of pressure from the
mismatched family and work domains. Participants claimed that work-family
conflict was a form of inter-role conflict in which the pressures from the work
and family domains were mutually incompatible in some respect. They have
also seen it as a rivalry between their professional responsibilities and their
personal and family lives. A respondent from the SNNPR said following from
the perspective of her family’s satisfaction:
page-802
“
No, they are not satisfied. They tell me to leave work and take
care of my children. My family was not satisfied at the beginning
when I was attending education. I attended my education because
I want to go out and work alike any person. When I left the work
and stayed at home, my family was happy....[What about your
husband?] My husband was not satisfied with my being an HEW
from the beginning. He was telling me to be a merchant.
HEW, SNNP Region
In relation to the above sentiment, participants also pointed out that HEWs
leave their jobs for personal reasons, such as having children but no one to
help take care of their children. An HEWs might leave her job when she has a
child and there is no one to take care of their children, and her spouse’s work
transfers him to another place but she is denied a transfer there. A respondent
from Hawassa reported that:
“
She left after giving birth to her baby.…Actually, she initially
requested to be deployed where her husband resides, but the zone
adminstrators refused and deployed her in another place, so she
was not able to travel back and forth to where her husband and
children live on a daily basis.
HEW, SNNP Region
Respondents also stated that interpersonal conflict at their work places could
affect not only morale but also the efficiency and productivity at the workplace.
They reported that it could also lead to strained relationships, grievances,
litigation, absenteeism and employee turnover. One respondent gave personal
conflict as the reason she left her job. This respondent reported that, after she
left her job, she tried to use all means to settle that conflict:
page- 803
“
Then he said he would write a final warning letter to us. Then she
came together with him and insulted me … At that time, I was so
angry and wanted to beat her, but she was pregnant. Finally, when
it was checked, it was found that it was her mistake. Then, she got
a final warning letter for her behavior. Then many people gathered
from the Woreda Health Office, kebele leaders, the community,
and the Health Center to hear what the problem. Both of us talked
about what has happened and the community also gave their
views. Then the woreda identified who was creating the problem.
After that, she was always unhappy with me; she said unpleasant
things about me. Then I decided to leave the job.
Ex-HEW, Amhara Region
4.8.2 Administrative and/or structural factors

In terms of administrative and structural factors, some factors discussed include
the poor timing of HEWs’ appraisal, demands for false reporting, judgmental
appraisal, denial of legitimate leave, officials’ failure to keep their promises,
woreda officers’ unresponsiveness, officers’ negative perceptions about HEWs,
poor support culture, unequal treatment or bias, disrespect from officials, and
lack of educational opportunities.
Work promotion with service years is one of the major work-related factors
reported by the participants. The participants perceived promotion as one of
factors that inspired employees to stay in their jobs. HEWs are in a place,
however, where this opportunity can easily be missed compared to other health
and non-health related work environments, which forces them to confront the
issue that it is only for HEWs that promotions do not work and as a result plan
to leave their jobs:
page-804
“
Others, like teachers, get promoted by their services, but we who
are educated are in remote and arid areas and the new ones are
in town, so when there are differences, why would we want to work?
We will look for other options for work.
HEW, Amhara Region
In addition, participants described promotion within the same place or

organization, reporting that HEWs were forced to continue working in the same
place, with the same responsibilities, even after they upgrade their educational
status. Thus, many employees look for work elsewhere or become homemakers
when they realize they have neither opportunity in their current workplaces nor
opportunities for transfer.
“
I think a majority of HEWs are interested in staying in their work.
What could be raised as a reason is that one HEW will stay for 10
years in one kebele, and after they upgrade their education, they
return to the same kebele. There are many complaints regarding
this.
HEW, Oromia Region
Regarding career structure and promotion, FGD participants reported that if

they have no control over their careers’ structure it might be a danger in the
future. For example, when they begin a degree program, if the program fails
to figure out a career path for these graduates before they arrive, it may lose
“
them after investing so much in them. One FGD participant from Tigray said:
The other thing is that the HEP is work without a career structure
after you serve a certain number of years. So, in the future, there
will not be anyone who will want to take up this work. So, we have
to have a means that could help us value the work experience of
HEWs, and thus they may develop a feeling and hope to gain a
better position that will consider their work experience as HEWs.
Otherwise, when they [prospective employers] abandon HEWs
page- 805
“
because they may not benefit from their work experience they have,
they won’t be considered. Now even their professional licenses will
be given as though they were a beginner or a fresh graduate. This
in turn affects their motivation to continue working as HEWs for
long. If they have a B.Sc. in nursing, their work experience will not
be considered at all, whereas if they were diploma nurses in the
first place and upgraded to a B.Sc. degree, they will be awarded
a senior nurse license, which is not the case for HEWs. This will be
very dangerous in the future too.
Program Officer, Tigray Region
Additionally, respondents reported that there was failure to conduct

performance appraisals or a timely performance appraisal, which increases
claims regarding unfair employment practices. Respondents reported that
supervisors and managers put off the performance appraisals and claim that
they haven’t received an appropriate appraisal. HEWs files included formal
complaints about the manner and timeliness with which supervisors evaluated
their performance. One respondent form Amhara said:
“
I got pregnant after the transfer, and I used to get sick with my
first child, so activities were not implemented, and 2 days after my
delivery, an evaluation was done again. Two years’ results were
combined for transfer, 50% for each year, so during the evaluation
I had delivered and was on leave, so it was low, then after that the
second year an evaluation was done, and I was again sent to a
remote arid area.
HEW, Amhara Region
page-806
“
Another respondent said:
The monitoring and evaluation from the Health Center and Woreda
Health Office was poor, but it exists. There is weakness in providing
monitoring and evaluation from the Health Center and Woreda
Health based on the schedule.
HEW, Amhara
Participants also assumed that the appraisal approach would be supportive and
help them perform their jobs better. They claimed that the current appraisals
are the opposite: unsupportive and judgmental, which does not give them the
chance to improve. A respondent in one woreda reported that the officials used
“
the appraisal as the sole basis on which to judge her performance and fire her:
In any work area, there are weaknesses; a mother or a child might

die in our care. Instead of supporting you so you do not repeat the
same mistake, they evaluate you harshly. They just point all the
fingers at you, put all the blame on you. Some workers request a
transfer, and some get sick due to this stress. Nobody will listen to
you. Then your only option will be to leave.
Ex-HEW
The other issue that could be seen as falling under the theme group of leadership
and governance is that of demanding false reports. Respondents claimed that
there was a tendency to report more politically desirable activities than they
had actually performed. Participants claimed that officials in the kebele and
woreda tended to demand false reports either for political consumption or their
own personal positions. In one way or another, some community members had
a hand in such processes. One respondent reported that she left because of
such tendencies:
page- 807
“
I performed a good-quality job, then it was up to them to accept
my work. Besides this, I do not lie: if I perform 40%, then I report
the exact figure. I am always criticized for my under-performance in
numbers. But when I was criticized, it is not because a single child
is found not vaccinated. I do not care about the percentage, but I
care about quality work. So I was doing this, and the management
people needed a false report, not quality work. And finally when I
left the job, I discussed this issue with my colleague HEW, and we
did not agree. Then I decided to leave the job rather than reporting
falsely and getting paid. Hence, I submitted a letter requesting
leave.
Respondents in all woredas argued strongly that an employee who has followed
the prescribed procedures for requesting and receiving annual leave has the
right to that leave (subject to the supervisor’s right to determine when leave may
be taken). Annual leave may be used for any purpose, including for purposes
related to pregnancy, childbirth and recovery from childbirth, adoption or foster
care, bonding with or caring for a baby, or for other childcare responsibilities,
including taking a child to medical appointments, well-baby doctor visits or any
other purpose. They said that they were denied this right and that they did not
“
get even weekend breaks. One respondent said:
The one who was working nearby requested an annual leave. But
before that, there was one maternal death from her center. So, they
denied her annual leave because of this, and she left her job. There
are many. Would you get them in a meeting? They could have
given you many ideas. It is so challenging.
HEW
page-808
Another respondent affirmed that officials prohibited sick HEWs from getting
more annual leave. One responded from Hawassa said:
“
The other left after an operation did not heal in the given time….
She asked for annual leave, but the woreda refused, so she left the
job. I don’t know if she is at work or not. Whereas the other left for
Kenya….In general, six HEWs have left the job.
HEW, SNNP Region
In addition, regional officers have accepted this claim in their FGDs. They
mentioned that they were not given any leave except for maternity leave.
Respondents consistently reported that the support culture from the HC and
the WorHO was poor. They claimed that support was not assumed unless
someone came from the zone or higher-level offices. They complained that
when supervisors or clinicians came for support, they simply bragged about
their qualifications. HEWs believed that supervisors were necessary in all
settings where they need guidance and supervision to complete tasks, serve
patients, and meet deadlines. They reported, though, that supervisors lacked
effective communication, problem-solving and employee-motivation skills. They
claimed that supervisors were not good leaders, took advantage of HEWs, and
needed to have insight into their strengths and developmental human needs.
“
Supervisors also needed to have a clear set of expectations and goals:
The Health Center doesn’t support us unless the zone comes. If the
zone sends activities, the office passes them to the Health Center,
and the Health Center deploys the professionals, but when they
come, they just talk and don’t support us. I usually disagree with
them because I was alone and did my work efficiently so that I
could argue when they say that we don’t understand the support.
HEW, Amhara Region
What FGD participants raised in relation to supervision was that the supervision
team came only when the HEWs’ performance was failing and things like the
death of a child or an increase in home deliveries. The supervisory team came
page- 809
from officials from thee woreda or RHB and focus on punishment, even though
HEWs are not at fault. Reportedly, all the failings are attributed to HEWs. They
ask supervisors what punishments have been given, whether salary deduction
or a warning letter about a poor performance appraisal. Hence, respondents
believed this was one reason HEWs wanted to leave their jobs. One participant
said:
“
Most of the things have been said. What I want to say is about
the relational hierarchies and the support system. One reality
regarding that is that supervisors are those who have not taken
basic trainings. Mostly such supervisory work is done by delegates
not the right person himself, because it has issues with payment
and benefits. Even in cases where supervisors are hired, they do not
provide enough supervision. They have not adequate knowledge
about HEP activities; HEWs are better than the supervisors in
every possible way.
HEP Program Officer
Respondents reported that factors related to self-development were the

motivating factors that could be felt as life goals. The lack of good opportunities
for education was a self-development factor that reportedly resulted in
the attrition of HEWs. On the other hand, HEWs, being selected from the
community, are expected to serve and continue to reside in that community,
while the HEWs themselves expect to be transferred to better locations, get
“
educational chances and upgrade themselves.
The major factor is education. Everybody needs change, and for

change, education is a must. There was no education opportunity.
Ex-HEW, SNNP Region
This respondent emphasized how important educational opportunity was in their

ability to make meaning of her work life. The lack of educational opportunities
affected HEWs individually and could force them to leave their jobs.
page-810
Parallel to this, the FGD participants reported that it was natural to have an
interest in upgrading and improving oneself in any ways possible. This was
expressed by one FGD participant:
“
For example, the educational level, since the level has been
opened, they all want to go for education and be like other health
professionals. This is just a normal human need, as all individuals
who have a degree want to have a master’s, and those who have a
master’s want a PhD. That is how it is with them too. It is in relation
to their need for education. If they have a diploma, they want to
have a degree, and they leave for that.
HEP Program Officer
Another FGD participant indicated that HEWs leave their jobs after they
upgrade. This participant claimed that HEWs have a higher tendency to leave
after they upgrade to level IV:
“
The second reason to leave is, when they upgrade to level IV, what
was expected was for their position and work environment to be
upgraded together to enable them to use their professional skills.
After upgrading, the HEW is expected to be a supervisor or has
to be assigned to the HCs. If not that, they need to be assigned to
places nearer to town or towns themselves. And since the demands
of these Health Extension Workers were not considered, they leave
for better positions.
HEP Program Officer
Participants also reported that educational opportunities were purposefully

hidden to prevent the HEWs from using them, and as a result, HEWs would
not ask for promotion or transfer after earning the desired qualification. The
participants perceived that officials were sensitive about the issue and changed
the topic during meetings. COC exam dates’ not being announced to HEWs and
doing evaluations were presented as evidence that they were being prevented
intentionally from learning.
page- 811
“
How are you supposed to go to learning? It is intentional not to
send you to education. Even COC, they didn’t tell me the date, they
hide the information from me. Someone else told me about it. But
for work-related evaluations, they even send you an ambulance. But
for things that will benefit you, they are reluctant. I was late for the
COC; I just took it once recently and passed the exam.
HEW
The above quote suggests that even the COC process is being manipulated by
someone who could benefit from it.
The FGD participants accepted that the supervisors did not want the HEWs to
upgrade, and many challenges have faced those HEWs who tried to upgrade
on their own. For that reason, some of them left their jobs, preferring education
over their jobs, while others dropped out of school just to stay in work. This
claim was corroborated by an FGD participant from Mekele, who said:
“
What has been said previously, there was an incident in Mekele city
in which Mekele University used to teach during the weekends as
in-service training, and most of the HEWs got registered for that
upgrading program and started their education. Then, when the
woreda and the supervisor got more controlling, those who had an
option or decided to continue learning left their jobs because they
were not allowed to learn in that crucial time; some left for this
reason. Then afterward, there were some who were complaining
that they were not able to continue their education because of that
strong control and influence, and they left again. Regardless of the
incidents, there was no formal effort to learn why they leave.
HEP Program Officer, Tigray Region
The other topic that was raised and appeared as a sub-theme in the FGD was
that of COC exams. One respondent explained that the COC was essential; if
they could not pass the COC, they could not be employed, or if they had already
failed, they would be fired. In reality, since they have extensive experience, it
would be better to retain them than rather than firing them. If they can keep
them in their jobs, it would be without a change in position, salary increment, or
educational opportunity. One FGD participant from Gambella said:
page-812
“
And this in turn makes them hopeless, and they may leave as long
as they can find any other opportunity. This may be considered one
reason, but not passing the COC was the problem with the change
in the education system and content at the time of their training
and the time they took the COC exam. Some of them even saw a
computer for the first time when they sat for COC exams. So there
were measures to train them before the COC exams; otherwise
most of the HEWs who were level III have had difficulty passing
the COC. So was it a reason, yes it was, and it may be still now,
because they could not get a promotion, and if they could not get
that promotion, they might get hopeless and bored and leave after
a while. They will remain in the same position because they do not
have the COC. We have tried to consider them in the GAG too, but
they still need to have the COC certificate.
Program Officer, Gambella Region
In one woreda, a respondent reported that when the selection system is not
followed and HEWs do not come from the kebele they need to serve, their
relationships with the community are constrained as a result of a lack of trust
from the community side, and this leads to poor performance. Contrary to this,
another HEW reported that being placed in the kebele where HEWs were
born is a challenge in the sense that there will be difficulty in carrying out some
demanding responsibilities when it comes to not following accepting what the
HEWs said, and the matter has to be governed by the law. They have said that
“
there will be conflict of interest because the matter could become a family-level
issue.
From the beginning, the rural HEWs should be someone residing
in the rural area who doesn’t want to leave. The announcement for
HEWs should include the residence of the applicants. In the future,
when there is announcement to enroll people in the rural HEP, they
should consider those who reside in rural areas.
Program Officer, Amhara
FGD participants objected strongly to the current recruitment process, believing

it to be full of problems that hinder the quality of services and performance
page- 813
the nation expected. FGD participants were unhappy that the responsibility
was left only to kebele leaders without checking the mechanism of the process
itself, but they were supportive of the recruitment criteria. The claim here, then,
is that kebeles are not using the criteria to recruit HEWs. One participant said:
“
The Health Extension recruitment process has problems that the
region transfers to the zone and then to the woreda and then to
the kebele. The criteria says those recruits must be from that kebele,
born in that kebele, have their parents there, have completed grade
10 and be able to speak the language and have good manners. But
once it reaches the kebele, the matter of nepotism comes in relation
to friendship, blood relations. Then these guys who have been
recruited not on the basis of the criteria, but relationships, would
come to training. Even the educational level expected and those
who come are very different. That is not the only thing we have
discussed so many times at the regional level: the training manuals
are prepared and the training is given in the Amharic language,
and those recruited do not understand the Amharic language.
Program Officer
The other main threat within the recruitment process is the quality and
competence of recruits. There are quota and politics issues associated with
recruitment; in one case, officials’ wives were selected as HEWs in all kebeles.
Most of these HEWs fail to work wholeheartedly in their assigned places after
returning from training. Participants from RHBs reported that there was a time
where they controlled the kebele and dictated to them regarding who should
“
be recruited as HEW candidates and sent for training:
Then it was decided to translate the manual in to English, and then

there were trainees who do not understand English. So now we are
discussing translating it into the languages spoken in the region,
like Anuak, Nuer and Majang. There are some who agreed on this
and some who do not, and there is still a problem. There is even
a problem after it has been translated into the local language in
which the recruited could not read. So what is best is we have to
be critical in the recruitment process. It is better to have a writing
page-814
“
exam at least. We have seen it by throwing it down to the kebele
for many years, and we are seeing the result. For example, there is
an entrance exam for those who are joining other professions, and
the same should be done for HEWs. For example, if we want two
or three, we will recruit five to ten individuals and then make them
take an exam, so that way we can find better recruits.
The other group of factors relates to officials’ perception and support. These
may include broken promises, poor perception and responsiveness, neglect and
disrespect of HEWs, and unequal and unfair treatment of HEWs.
Participants reported that there was no need to detail the many broken promises
that have accumulated throughout their stories. For example, in case of the
reality of HEWs, they claimed that they were promised a lot when they were
recruited for training back then when the HEP was being planned. For instance,
they claimed that they were promised that they would be given the chance to
upgrade and even become clinicians:
“
So, when we were there, transfer was done since we had a one-year
evaluation, then when we asked, “What is going to happen?” they
told us it would be calculated and there was a place set for us.
So, since they said that, had they told us it was not calculated, we
would have complained at the zone while we were doing education,
but when we came back, they said it was not done and to go back
to our kebele.
Another claim that showed that the promises the officials were making were not
being kept relates to educational achievements and upgrading. Respondents
stated that they were confused about the value up upgrading educationally:
page- 815
“
We have accepted that it was remote, and went there after the
performance evaluation was gathered. Then, we performed well,
but the transfer was not done; this was the obstacle. You learned
and you went backward. We have raised this question as well; when
we attended level IV, our teacher told us to read when we went for
apprenticeship; we used to say that the Health Extension was still
backward. Even if we learned more, what was the use? What is the
use if we pass the COC or not? Even when we meet here among
ourselves, we say when somebody got level IV or passed level III
COC, what did she get? They didn’t get anything new. So this is
discouraging.
HEW, Amhara Region
FGD participants mentioned that Hews’ questions never reached the concerned
body. Somewhere in the chain it was kidnapped. Sometimes there are issues
raised to the RHB that seem surprisingly new but are old for HEWs. There
is also a tendency not to be able to see and accept HEWs as professionals,
perceiving them as inferior. Those in the HCs are said to be professionals while
they are treated as something else. They feel bad about this as human beings.
One FGD participant said:
“
The other point here is that the issue that HEWs raise and questions
they ask on different occasions, like meetings, supervisory visits and
reports, are not compiled at each level and sent to decision-makers.
The supervisor does not compile their complaints and submit them
to HCs, or HCs do not compile them and send them to the woreda,
or the woreda to the zone or region or to the MoH, which may lead
their question and complaints not to be answered. So, HEWs would
say, “We have said this and that previously. What is the verdict on
our questions?” in every encounter we have with them. And officials
go, like, “Forget about them, they are always like that.” So they are
ignored in every way possible.
Program Officer
page-816
Respondents stated that the managers were the type of people who keep piling
it on, oblivious to the telltale warning signs that HEWs show as a result of an
overload and being about to hit the breaking point. They complain that the
woreda officers do not have the courage to see HEWs’ lives and have poor
perception:
“
The woreda itself and the perception of officers in the office are
not good, but it’s because of my children nonetheless that Health
Extension work satisfies me. Woreda officers don’t have good
perception of HEWs; that is one of the reasons I left my work.
Others might have options or have wanted to learn, but the woreda
made me leave, they put aside what the woreda has done for me.
That post is still open.
FGD participants raised the above issue in relation to burnout due to long years
of work without any change that was been expected. One FGD participant said:
“
Because they stay there for a long time, they develop burnout
syndrome. I believe that this is one of the reasons they leave. Plus,
motivation is an important factor, including salary increment,
incentives for mobile and transportation and other issues that cannot
be afforded by woredas due to budget shortage or other reasons.
So, there is nothing that keeps them motivated, and it might cause
attrition. The other observation regarding reasons for attrition is
personal health problems. If they develop chronic diseases, it could
be one root cause of their leaving. This is true because the work is
difficult and need lots of hard work and physical movements, which
need HEWs to be effective.
Program Officer
Respondents in some selected woredas reported that officials did not treat them
equally. Their specific personal experiences could be evidence of their claim. The
reports on such differential treatment involve differences in management style,
page- 817
employee discipline and other actions necessary for HEWs. It also includes
illegal differential treatment, such as singling a worker out because of certain
“
characteristics, which is completely avoidable.
The management was good, but there is a weakness in this area

too. For example, there are conditions of assigning workers to a
good place if they have a good personal relationship with officials.
Some managers do not consider us equally, and they benefit the
HEW whom they know personally more.
HEW, Oromia Region
Participants reported that woreda officers were unresponsive and paid little
attention to the complaints and personal suffering of HEWs. All respondents
expressed the experience of chronic work stress, with low salaries, lack of
opportunities for advancement and heavy workloads topping the list of
contributing factors. On the heels of the recession, many HEWs appear to
feel stuck. Compounding this problem, less than half of the correspondents
reported that they received adequate monetary compensation or non-monetary
recognition for their contributions on the job. One respondent said:
“
I told them “I have a child a year and 5 months old who is also sick.
How am I supposed to go? You may send someone else”—there
were many young people who can go there—and later, when my
child gets stronger, I will go.” They said, “No, you have to go.” And
they banned me from taking my salary for three months. After
that, since I was supporting my family with the salary I got, I was so
stressed. Finally, they decided to fire me, and they [the WoHO] did.
Then I appealed to the Civil Service, and the Civil Service decided
to reinstate me and take my three months’ salary as a penalty.
Then I went to the Regional Health Bureau, and the Bureau called
a representative from the woreda and told them, “How come you
prevent her from taking her salary? She has a sick child, what is
she supposed to eat?!” And they told me to go and try work at the
new place for about four or five months, and if it is hard I can be
transferred by submitting an application.
HEW
page-818
4.8.3 Salary and incentive packages

This category covers factors such as low salary, poor chances of promotion,
benefits and incentive packages.
Most developed countries monitor compensation, making sure that salaries

are basically fair. In the case of Ethiopia, there are indications that things fall
through the cracks. Unfortunately, HEWs have one complaint of this type,
and what makes their claim more interesting is that sometimes the head boss
doesn’t care about fair but focuses solely on paying them as little as possible.
Given the amount of physical or mental exertion needed to perform the job, the
degree of accountability required in performing the job and working conditions
encompasses two factors: (1) physical surroundings, like temperature, fumes
and ventilation and (2) hazards, which strengthen HEWs’ complaints. One
respondent said:
“
The salary was not enough for our work. For example, during our first
employment, our salary was 380 ETB. After a period, 22 ETB were
added for us, and we were getting 402 ETB. It is not comparable to
the services we were providing. Because we leave our homes in the
morning and provide service in the community by travelling long
distances on foot. In my opinion, the Health Extension works, and
the salary paid for them is not equivalent.
HEW
FGD participants mentioned there was no pay rise comparable with their
changes in responsibilities and expansion of packages of the HEP. One
participant said:
“
On my side, what I would like to add finally is that, in the Health
Extension, those packages that have been added at the federal
level have been as an additional workload. Why we say that they
have been a workload is because all the responsibilities of other
sectors, like education, agriculture and women’s affairs, have been
page- 819
“
shifted to HEWs. Even though the packages have been added,
nothing has been thought about HEWs. There is no top-up or
increment in their salaries. So she will get an equal salary doing all
the work she does. Doing all that, she receives the same salary as
teachers. HEWs in some places complain that the government has
added to their workload, but nothing has changed regarding their
salary. When the packages were 16, there were not many, but now
there are complaints.
Program Officer
Respondents recognized that the incentives and rewards for their performance
have benefits for both the HEP and themselves. They believe that when they
are recognized, their performance and productivity increase, as well as their
morale being given a huge boost. But most complaints about the limitedness or
absence of such habits are that they have affected their efficiency negatively
and decreased their productivity. Therefore, we can assume that when rewards
are given regularly for excellence, the result is a win-win situation.
“
Additional incentives for me, during my time? I did not get any
incentives. Maybe sometimes NGOs, it’s only my name. To be frank,
even when there is a good training, I was not selected. I have never
attended training that took place in other areas. Maybe I could get
an incentive like others when there is a campaign. Also, one time,
when I was pregnant, there was a training called PMCT. During
this training, I did not attend....Then they called me for experience-
sharing training, and I went....Apart from this, there is no additional
incentive for HEWs. I do not think it has any.
HEW, Amhara Region
4.8.4 Work environment related factors

The work environment category has already discussed being assigned to
locations other than one’s birthplace, distance of the work site, means of
transportation and difficult topography.
page-820
Respondents claimed that, previously, things were difficult, and even the
community had no awareness of the HEP. The HEWs, then, complain that
their time was difficult. For instance, they had no transportation; thus, there
was no conducive transportation, which the officials did not understand. The
other challenge related to the work environment was community acceptance
and awareness of the role and responsibilities during the initial phase of HEP. It
was also very difficult due to challenges like kebele administrations’ and RHBs’
failure to support HEWs as they should have, the lack of access to transport,
the low responsiveness of officials and resistance from the communities.
A few respondents reported that they had left their jobs due to the difficulty
and hardship linked to the job and the work site and its effect on their personal
“
lives. One respondent said:
I left the job due to the difficulty of the work and its social impact.
I left the HEW due to my personal life. The work is difficult in rural
areas, especially for females. I left the HEW job due to the difficulty
of the work and everything in the rural being uncomfortable for me.
I suffered a lot at that time.
Ex-HEW, SNNP Region
Difficult topography and a sense of insecurity while walking on foot to and from
health post to community and residential places was also a major challenge
reported by both active and resigned.
The FGD participants mentioned that one of the reasons for leaving their jobs
was topography. They have served more than 14 years as HEWs, and after
these long service years, there are no transfers in the HEP, so it is difficult to
get transferred to another woreda, and the topographical challenge is constant
“
and exhausting. One respondent said:
As I have mentioned before, I have told you that the topography is

challenging and we are forcing them [the HEWs] even when they
are pregnant, it is a must to address one household at least every
three months, and to do that, they should not get pregnant or they
could not handle the hardship. We have lost some HEWs because
of this.
HEP Program Officer
page- 821
page-822
5 Discussion
The study has identified that the overall prevalence of attrition of HEWs is 21.1%
over the 15 years period. It was also identified that there is variation in the rate
of attrition among regions, ranging from 0.32% to 30.33%. While there is an
internationally agreed-upon cutoff point to designate a certain level of attrition
as low, medium or high, the figures in this study could be considered high for
a country that relies on these workers almost entirely to meet the crucial needs
of basic preventive and curative health care service at the grassroots level.
The literature usually defines the magnitude of attrition in terms of the value
and performance of those employees who leave their jobs. Since there is no
nationally standardized method of measuring the performance of HEWs, it
might be difficult to classify HEWs as high or low performers with confidence,
but such measures could enable a healthy attrition rate to be defined simply by
identifying the lowest performers, as their attrition could pave a way for new
talent and energy to achieve the HEP’s goals. The current study has analyzed
the data collected on the performance of the HEWs, although it might not be
measured in a standardized way. It shows that a majority (63%) of those who
have scored below the average performance of 80% since such measurement
began also have left their jobs. In this sense, it could be justified that such
attrition, where it is mostly the low performers who leave, is advantageous, as
described earlier. It is also good to note that some of the attrition is due to
official dismissal because of frequent absenteeism, discipline issues and related
factors, which, again, can justify such attritions as beneficial for the program.
Deaths have also contributed to about 3% of the attrition, which is clearly a
cause that cannot be solved by improvements in the program.
Various studies have reported the attrition rate of community health workers
in different countries. A study in Sri Lanka on the feasibility of a large-scale
community health workers program in 1982 identified an attrition rate of 77%,
concluding that interventions involving community health workers usually end
up in high rates of attrition unless a conducive environment is created [1]. There
is a marked difference in the level of attrition from the current study, but the
study conducted in the 1980s may indicate the many challenges to the workers
then compared to the HEWs in the current study. In another study conducted on
page- 823
improving the duration of exclusive breastfeeding in Bolivia, the attrition rate of

voluntary community health workers involved in the program was estimated to
be 43%: again, higher than the current study in relative terms [2]. Yet another
study, this one in South Africa and focused on tuberculosis treatment, has seen
the attrition of 11 out of 12 lay health workers in the community within a one-
year period. This overwhelmingly high level of attrition, although the sample
size is very small, makes it impossible to comfortably compare the results of that
study to the current one [3]. A study conducted in Bangladesh to assess the
factors affecting the recruitment and retention of community health workers in
community-based newborn care identified a 74% (32 out of 43 workers) attrition
rate over a four-year period, but the study had a small sample size, which
might not have adequate power to estimate the attrition rate accurately [4].
An unpublished study found a 27% attrition rate of HEWs in Oromia, slightly
higher than the current national figure. The present study, however, found the
attrition rate to be 9.2% in Oromia, far lower than the Feyisa study. Contrary
to both the above findings and the current study, one study conducted in
Nepal on female community health workers in a program that lasted for over
20 years saw an attrition rate of less than 5% annually. Similar to the current
findings, variations were seen among the districts of the country: it was higher
than the average in some districts [5]. A seven-year attrition of 49.6% was
seen in a Kenyan study on volunteer community health workers, another study
showing a higher attrition rate compared to the current study [6]. Namibia
is implementing an HEP like Ethiopia’s, and, in an evaluation of the overall
program, a low annual level of attrition was found: only 3.6%. Attrition varied
considerably, however, among the regions in the country [7].
The attrition rate of HEWs has varied over the years, just as it did among the
regions during the first 15 years of implementation of the program. The first five
years, 2003- 2008 saw the lowest magnitudes of attrition, and no attrition in the
years 2004 or 2006. After 2009, however, it continually increased until 2018,
which saw the highest magnitude of attrition: 1 999 per 10 000 HEWs. The low
magnitude during the early years of implementation is expected and could be
attributed to the few job opportunities in other sectors, the majority of HEWs
being unmarried and the fact that the program itself was initiated in the rural
parts of the country. As the years have gone by, the HEP’s implementation
was further expanded into urban and pastoralist areas, many HEWs who were
single become married, other job opportunities became available, and in most
regions for most of the implementation period, incentives such as annual leave,
page-824
transfer and educational opportunities were restricted, all of which might have
contributed to the increasing attrition in the later years. The higher attrition also
tends to coincide with the overall political instability in the country, especially
during the years 2016-2018.
Similar studies conducted in other countries also show variation in attrition

during different years. A study in Uganda showed a retention rate of 95%
during the first year and 91% and 86% during the second and fifth years of
implementation, an increasing trend. The study concludes that retention of a
majority of community health workers over a medium term is possible [8]. A
study in Kenya on a community-based program started in 2002 saw an overall
attrition rate of 33%, with dropout of 36% in the first two years. In 2007, five
years after the initiation of the program, 59% of the staff had dropped out,
showing an enormous increase in attrition, a similar trend to that of the current
study. The study identifies the high expectations of and subsequent rejection
by beneficiary families as the main reasons for the high attrition during the
later years [9]. In another related study on the retention of community health
workers in Tanzania, an attrition rate of 12.7% over a four-year period was
observed. The trend of attrition of the community health workers varied over
the years. In the first year, 174 community health workers left, followed by 123
and 174 community health workers in the second and third years, respectively
[10].
The other main finding from the current study is the median time of attrition of
HEWs. It was estimated that HEWs on average serve a median of 5.9 years
before leaving their jobs. In other words, about 50% of HEWs leave their job
after serving not quite six years. The overall median time of service for the
participants in the study is 7 years. Based on this finding, there is only a one-
year difference in the average number of years served among those who are still
working and those who have left their job. One reason for such a small difference
is the lower rates of attrition seen during the early years of implementation.
Higher rates were seen after 2008, and in some regions, attrition was seen only
for short period of time. For instance, in Harari and Gambella, the attrition
began late in implementation: 2017 and 2013, respectively. These all could
have contributed to the small difference in the number of years served. In this
regard, it could be asserted that those HEWs who have left would not have
added much to the program compared to those who have continued to work.
In other words, an average of one-year additional service to the program is
page- 825
not particularly significant. The Ugandan study on the retention of community

health workers has estimated the average length of time until resignation or
death. In this study it was reported that, on average, community health workers
stay for 20 months from initial training until resignation or death, ranging
from 1 month to 60 months [8]. For a five-year program, the average stay of
less than a year is relatively shorter than was found by the current study. In
the Kenyan incidence of attrition study, the median number of years served by
community health workers was 5.7 years, longer than the seven-year estimate in
the current study. In the same manner, most of those HEWs who have resigned
in the current study stayed for almost half of the program’s implementation
period [7]. As described earlier, various factors might have played a role in such
a long tenure, but this trend may not continue the same way, and employees
may start to leave earlier than anticipated.
The incidence rate of attrition—i.e. the number of new cases leaving every
year—was estimated to be about 288 per 10 000 person-years. This result
corresponds to the high prevalence of attrition reported earlier. Similar to the
prevalence, the incidence of attrition has also varied over the years. The highest
incidence was seen in the eight years of implementation, 519 per 10 000 person-
years. When also seen trend-wise, the incidence of attrition steadily increases
over time. The overall incidence of attrition was estimated to be 46.8 per 1000
person-years in the Kenyan study [7]. This figure is almost double the estimate
of the current study.
Several factors were identified as having a significant association with the

incidence of attrition in this study. HEWs’ age, birthplace, type of deployment—
i.e. whether a new deployment or a promotion from another position—the
perceived distance of the HP from the WorHO, the level of certification gained
by HEWs recently and at the time of their deployment, having children and
COC status were all independently associated with the incidence of attrition of
HEWs. Deploying new HEWs, recruiting those born in rural areas and in the
age category 20-24, at all certification levels at deployment except level I and
working in pastoralist and urban settings are all protective of attrition. On the
other hand, having one or more children, being COC-certified, working in a
remote HP and having a recent level IV or degree certification were identified
as increasing the probability that a HEW will leave her/his job. In the selection
and recruitment process of new candidates for training as HEWs, it is good
to consider these attributes of the candidates so that those with higher odds
page-826
of staying longer will be selected. Factors such as recognition status, marital

status, family size, taking annual leave and administrative punishment were
not associated with the incidence of attrition of HEWs, contrary to what was
anticipated.
Various studies have sought to identify the factors associated with either the
retention or attrition of community workers, some of the factors of whom are
similar to those of the current study. Most of these studies directly involved
the community health workers in order to assess such factors, but the current
study could not include these factors, as the data collection was conducted
through a review of records. A study in Ghana on the retention of community
health workers has identified the approval for work by the community and
the workers’ immediate family members as statistically significant predictors of
retention. In this study, the estimated attrition rate was 21.2% and considered
a moderate level [11]. In the Bangladesh study, the poorest community health
workers were more likely to continue working compared to the those from a
higher wealth quintile. Other predictors of retention identified in this study were
social prestige and household responsibilities [4]. A study conducted in India
found that the factors that motivated and led to the retention of community
health workers included their age and the recent provision of training. Other
motivating factors were financial incentives, community support and recognition,
organizational commitment and pride. On the other hand, job burnout, poor
personal health, job insecurity and a lack of career-development opportunities
were found to be de-motivators [12]. A study published by the World Health
Organization (WHO) mentions inadequate and irregular payments, a lack of
family support, age, upgrading of health posts, a lack of profit, a poor selection
process, the availability of better employment positions in other fields and the
loss of economic opportunities as factors significantly associated with attrition
[13]. In the Tanzanian study, factors such as being married, being male and
having prior volunteer experiences predicted retention of community health
workers [10]. Studies in Kenya have found a lack of interest in peer organization
membership, the absence of refresher training, the lack of feedback from
supervisors, high expectations by the community health workers, a perceived
heavy workload, a lack of adequate support or transparent operations of local
NGOs and poor selection criteria of CHWs to be associated with higher rates
of attrition [6, 9].
page- 827
Studies have also described motivation or demotivation factors acting as

factors for the retention or attrition of community health care workers. In a
study on factors motivating community health workers in Uganda, experiencing
issues with supplies and a lack of collaboration with peers were significantly
demotivating, while factors such as social responsibility and aspiration for other
opportunities were motivating factors for community health workers, which were
also identified as contributing to their preference to continue working [14]. A
similar study in a region of Tanzania reported altruism and the intrinsic need to
support and respect the community as incentives to work as community health
workers and hence to continue working [15].
In general, the literature reviewed suggests that the attrition of community health
workers tends to be high and is markedly so when the projects or programs
being undertaken are of a large scale and cover a longer period [1, 4, 5, 13].
As described earlier, a majority of studies saw a much higher level of attrition
[1, 2, 13] than was found in the current study. Only a few studies identified
lower levels of attrition [5,7]. Even when attrition rates are low, however, they
tend to vary by region or district and over time, as was the case in the current
study [2, 3, 5]. Hence, curbing the relatively high level of attrition, especially
of those with better performance and longer experience serving communities
should be a priority both at the regional and national levels. Otherwise, the
sustained attrition and shortage of HEWs will inevitably impede the delivery
of the services at the community level and might even roll back the gains made
so far. Such a problem also leads to a lack of continuity in the relationship
established among the HEWs, the community and the overall health system
[16]. Moreover, a considerable investment is made in recruiting, training and
deploying these HEWs, and refresher trainings are further provided while on
the job. Frequent turnover means than more and more investment is needed
and the cost to the health system increases over time. On the other hand,
higher retention is advantageous in that it has lower costs and higher return in
the long term.
Common issues raised by HEWs, such as the provision of educational opportunities,

various incentives, such as annual leave, sick leave, transfer and related issues,
should be made available to ensure HEWs’ motivation and continued services.
Furthermore, other well-known factors affect the chances that HEWs will stay
longer serving the community. Factors like the regular supervision of HEWs
with supportive and positive feedback, ensuring the continued supply of drugs,
page-828
supplies and consumables needed to deliver services, allocating budget for

running costs and recognizing better performers are important. As described
earlier, during the recruitment and training process, factors indicative of those
who stay longer or those prone to leaving earlier should be considered in order
to increase the chances of retaining these workers longer.
page- 829
page-830
6 Conclusion and recommendations
6.1 Conclusion
The study revealed an overall acceptable level of attrition among HEWs

during the 15-year period implementation. However, the attrition rates in urban
and pastoralist areas need more attention. Furthermore, there has been a
gradual increase in overall attrition over time, mainly during the past five years,
which warrants identification of various ways to retain experienced and well-
performing HEWs.
The average number of years HEWs serve before resigning is long enough given
the short training period each HEW passes through. Through the improved
economy, the wide availability of job opportunities in other sectors and HEWs’
increased awareness of such opportunities, the current trend of attrition could
rise, which will ultimately shorten the number of years served by those who
leave.
Factors such as HEWs’ age, number of children, level of certification, and

birthplace predicted time to leave among HEWs. Reasons for HEWs’s attrition
include personal, administrative, incentives, and workplace hardship related
factors.
6.2 Recommendation
Even though attrition rate among HEWs is within acceptable ranges given the
potential for replacement, retention actions should be in place to ensure that
better performing HEWs stay in the system. Hence, there is a need to make
employment in the HEP more competitive where the benefits of continuing
to work in the community outweigh, as much as possible, the benefits of
employment elsewhere.
page- 831
A comprehensive package of benefits, linked to objectively measured

performance, will help in the retention of better performing HEWs. In addition
to monetary incentives, factors like the provision of refresher trainings, respective
the civil servant rights of HEWs, including rights of annual leaves and transfer,
should be seriously considered for health post staffs.
Recruitment of HEWs should involve provision of adequate information about

the nature of work of HEWs to candidates in order to ensure expectations are
realistic.
page-832
References
1. Walt G, Perera M, Heggenhougen K. Are large-scale volunteer

community health worker programs feasible? The case of Sri Lanka. Soc
Sci Med 1989;29:599-608. doi: 10.1016/0277-9536(89)90179-
2 pmid: 2799410.
2. Tenerio A, Saunero R, Sinani J, Lafuente L, Gutierrez F. Extending

the duration of exclusive breastfeeding in El Alto, Bolvia through a
community-based approach and the provision of health services. Dhaka:
Child Health and Nutrition Research Initiative (CHNRI); 2009.
3. Atkins S, Lewin S. Jordaan E. Enhanced tuberculosis adherence

programme. Cape Town: Cape Town Medical Research Council; 2009.
4. Rahman SM, Ali NA, Jennings L, Seraji MH, Mannan I, Shah R et

al. Factors affecting recruitment and retention of community health
workers in a newborn care intervention in Bangladesh. Hum Resour
Health 2010;8:12. doi: 10.1186/1478-4491-8-12 pmid: 20438642.
5. Glenton C, Scheel IB, Pradhan S, Lewin S, Hodgins S, Shrestha

V. The female community health volunteer programme in
Nepal: Decision makers’ perceptions of volunteerism, payment
and other incentives. Soc Sci Med 2010;70:1920-7 doi: 10.1016/j.
socscimed.2010.02.034 pmid: 20382464.
6. Ngugi AK et al. Prevalence, incidence and predictors of volunteer

community health worker attrition in Kwale County, Kenya. BMJ Glob
Health 2018;3:e000750. doi:10.1136/bmjgh-2018-000750
7. Evaluation of the Namibian Community Health Workers Program.

Ministry of Health and Social Services, Namibia. 2017 November 30.
8. Ludwick T et al. Poor retention does not have to be the rule: Retention
of volunteer community health workers in Uganda. Health Policy and
Planning 2014;29:388–395 doi:10.1093/heapol/czt025
page- 833
9. Olang’oa CO et al. Staff attrition among community health workers

in home-based care programmes for people living with HIV and
AIDS in western Kenya. Health Policy 2010;97:232-237. doi:10.1016/j.
healthpol.2010.05.004.
10. Ngilangwa DP, Mgomella, GS. Factors associated with retention of

community health workers in maternal, newborn and child health
programme in Simiyu Region, Tanzania. Afr J Prm Health Care Fam
Med 2018;10(1):a1506.
11. Abbey M, Bartholomew LK et al. Factors related to retention of

community health workers in a trial on community-based management
of fever in children under 5 years in the Dangme West District of
Ghana. Int Health. 2014 Jun;6(2):99-105. doi: 10.1093/inthealth/ihu007
12. Tripathy JP et al. Measuring and understanding motivation among

community health workers in rural health facilities in India: A mixed
method study. BMC Health Services Research 2016;16:366. doi 10.1186/
s12913-016-1614-0
13. Nkonki L, Cliff J, Sanders D. Lay health worker attrition: Important but
often ignored. Bulletin of the World Health Organization 2011;89:919-
923. doi: 10.2471/BLT.11.087825
14. Brunie A et al. Keeping community health workers in Uganda

motivated: Key challenges, facilitators, and preferred program inputs.
Global Health: Science and Practice 2014;2(1)
15. Mpembeni RNM et al. Motivation and satisfaction among community

health workers in Morogoro Region, Tanzania: Nuanced needs and
varied ambitions. Hum Resour Health 2015;13(44). doi 10.1186/s12960-
015-0035-1
16. Bhattacharyya K, Winch P, LeBan K, Tien M. Community health

worker incentives and disincentives: How they affect motivation,
retention, and sustainability. Basic Support for Institutionalizing Child
Survival Project (BASICS II), United States Agency for International
Development. Arlington, Virginia, October 2001.
page-834
Specific Study 3:
The Role of the Health Extension
Program in the Public Health
Emergency Management
page- 835
CONTENTS
Contents --------------------------------------------------------------------836
List of Tables ------------------------------------------------------------------938
List of Figures ---------------------------------------------------------------838
Executive Summary ---------------------------------------------------------841
1. BACKGROUND ----------------------------------------------------------------845
2. OBJECTIVES ----------------------------------------------------------------------851
2.1. General Objective ------------------------------------------------851
2.2. Specific Objectives ------------------------------------------------------851
2.3. Scope of the Study ----------------------------------------------852
3. SIGNIFICANCE OF THE STUDY -------------------------------------855
4. METHODS -------------------------------------------------------------------857
4.1. Study Area and Period ------------------------------------------857
4.2. Study Design ---------------------------------------------------857
4.3. Source Population ----------------------------------------------------------858
4.4. Study Participants --------------------------------------------------858
4.5. Sample Size ---------------------------------------------------------858
4.6. Data-Collection Tools and Data-Quality Management --------859
4.7. Sampling strategy ------------------------------------------------------861
page-836
4.8. Data management ------------------------------------------------863
4.9. Ethical clearance -----------------------------------------------------864
5. FINDINGS ---------------------------------------------------------867
5.1. Characteristics of Study Participants -------------------------867
5.2. Overall Knowledge Level of HEWs ----------------------------868
5.3. Structure of the Primary Health Care Surveillance System --869

5.4. Core Functions of the Primary Health Care
Surveillance System ----------------------------------------878
5.5. Supporting Functions of the Primary Health Care

Surveillance System -------------------------------------------------892
5.6. Data-Quality Components of the Primary Health

Care Surveillance System ----------------------------------------895
5.7: Variability of HEWs’ Knowledge Scores at the Woreda Level --898
5.8. Variability and Relationships of Surveillance

Components at Different Levels -------------------------------------899
6. DISCUSSION ------------------------------------------------------903
6.1 Health Extension Workers’ Knowledge ------------------------903
6.2 Supporting Function of the System at the

Community Level ----------------------------------904
6.3 Structure of the system at the community level ----------------905
6.4 Core functions of the surveillance system at

the community level ------------------------------908
6.5. Limitations of the study ------------------------------------912
page- 837
7. CONCLUSION AND RECOMMENDATIONS -----------------------------913
7.1. Conclusion --------------------------------------------913
7.2. Recommendations ------------------------------------------------914
REFERENCES --------------------------------------------------------917
List of Tables
Table 1: Main Research Questions Addressed by the Assessment --------852
Table 2: Distribution of Structure of Primary Health Care-Level Surveillance

System Indicators by Region --------------------------------------------------876
Table 3: Distribution of Core Functions of Primary Health Care Level

Surveillance System Indicators, by Region -----------------------------890
Table 4: Distribution of Supporting Functions of Primary Health

Care Level Surveillance System Indicators by Region -----------------896
Table 5. Variability in HEWs’ Knowledge, Structure and

Surveillance Components of the System ------------------------------------898
Table 6: Association of the Supporting Functions and

Core Functions of the Surveillance System -------------------------------------900
List of Figures
Figure 1: Scope of the assessment --------------------------------------852
Figure 2: Sampling strategy and number of study participants

enrolled in the assessment --------------------------------------862
page-838
Figure 3: Geographical distribution of sample zones -------------------------863
Figure 4: Characteristics of Health Extension Workers

included in the assessment ----------------------------------------877
Figure 5: Characteristics of HC- and woreda-level respondents

included in the assessment -----------------------------------------------------868
Figure 6: Distribution of HEWs’ knowledge score by region ---------------869
Figure 7: Proportion of respondents who mentioned specific

disease conditions as reportable diseases/events in their locality --------878
Figure 8: HEWs’ knowledge of immediately and weekly

reportable diseases/events and reporting, by regions ---------------------882
Figure 9: Regional distribution of surveillance data

analysis and visualization practices commonly used by
HEWs at the Health Post level --------------------------------------------884
page- 839
page-840
Executive Summary
Background: The Ethiopian Health Extension Program (HEP) was launched

in 2003 with the aim of delivering a package of basic and essential promotive,
preventive and curative health services targeting households. The public health
emergency management (PHEM) system requires the active involvement of
the community and primary health care system (PHC). Despite the involvement
of the HEP in the PHEM system, the HEP’s contribution and actual role in
PHEM-related activities have not yet been clearly outlined or recognized. This
study aims to assess the role of the HEP in addressing the basic processes of
the PHEM System in Ethiopia.
Method: A cross-sectional study involving multistage sampling was conducted

to assess 4 key areas/components of community-level surveillance system.
These components are the 1) structure of the system, 2) core components of
the system, 3) support functions of surveillance system and 4) data quality
attributes. Its conceptual framework was adopted from the Primary Healthcare
Performance Initiative (PHCPI) and a modified version of the WHO framework
for evaluating communicable-disease surveillance systems. Both quantitative
and qualitative approaches were used to collect the required data, which were
collected using a standardized questionnaire, key informant interviews (KIIs),
focus group discussions (FGDs) and a review of secondary data. Health Extension
Workers’ (HEWs’) knowledge was assessed using a standard questionnaire
addressing 13 questions on the basics of community PHEM, including the public
health conditions under surveillance in context, their case definitions and the
thresholds for public health preparedness and response. Descriptive statistical
methods were used to analyze the quantitative data, and the findings were
disaggregated by region. The variability of the HEWs’ knowledge scores and
surveillance system components’ performance at 3 health system levels were
measured. Findings from the qualitative study were triangulated with findings
from quantitative findings to better understand the bigger picture of the system
in addressing the basic process of PHEM at the community level.
Result: One hundred (70.9%) HEWs scored above the mean score. From the
expected 15 notifiable diseases/events and conditions at the community level,
100 (70.9%) of HEWs reported 1 to 10 diseases/event conditions in their locality,
and 8 (5.7%) reported 11 to 15. On the other hand, 22 (44.0%) of woreda-level
page- 841
Assessment of the Role of the Health Extension Program
in the PHEM System in Ethiopia
PHEM experts and/or HEP coordinators reported only 1 to 5 disease conditions.

In addition, 122 (86.5%) HEWs knew the correct case definition of cholera, and
124 (87.9%) knew the correct case definition of measles. Significant variability
in HEWs’ mean knowledge scores was observed among the study woredas and
HEWs working in different regions (P<0.001). Of the HEWs interviewed, 114
(80.8%) knew that there were mandatorily notifiable disease conditions. Of the
HPs visited, 59 (83.0%) had functional community-based networks, and among
the community networks the Women’s Development Army structure was the
dominant available structure, functioning under 52 (73.2%) HPs. Of the HEWs
working at the Health Post (HP) level, 100 (70.9%) were practicing a mixed
surveillance strategy (both active and passive). Of all HEW respondents, only 71
(50.3%) confirmed the existence of a kebele-level emergency coordination task
force in their locality. Of all the health facilities visited during the survey, only
6 (11.0%) allocated a budget to support community-level surveillance activities,
of which 80.0% was dedicated for per diem and transportation allowance.
Supplies and logistics for emergency-response purposes were available at only
22 (41%) health facilities. At the time of the survey, a total of 30 (60%) woreda
structures were working with different governmental and non-governmental
stakeholders regarding emergency preparedness and response activities. Of
the HEWs interviewed, 100 (70.9%) claimed to have a means of verification of
reported cases from the community, and 70 (49.6%) had engaged in surveillance
data analysis at least once in the past. Regarding surveillance data reporting,
a total of 124 (87.94%) HEWs reported that they had sent a surveillance report
to their catchment health facility. In addition, 111 (78.7%) HEWs confirmed their
participation in different disease-prevention, emergency-response and control
activities, including social mobilization and health-education-related activities.
Of the HEWs, 88 (62.4%) had no prior information or knowledge regarding
emergency preparedness planning in their woreda. Regardless, 21 (14.8%)
HEWs reported an occurrence of a public health outbreak, like measles and
cholera, in their locality in the last 12 months. Reporting formats for reportable
diseases/event conditions were available in 49 (69.01%) of the visited HPs, of
which only 15 (21.1%) had a locally translated version of the reporting formats.
Only 52 (73.2%) HEWs confirmed having a guideline, protocol or package to
guide their surveillance activities at the community level. Sixty-seven HEWs
(47.5%) were not trained on community surveillance. Forty-two health facilities
(79.0%) monitored the quality of the surveillance data by measuring both
the timeliness and completeness of the data. Thirty-seven (52.1%) of the HPs
visited received supportive supervision and got feedback from higher officials.
Significant variability was observed in the availability of structure of the system
page-842
at all 3 health system levels. In addition, how the functions of the system were
supported differed significantly across the 3 levels. The supporting and core
functions of the system have a direct and positive relationship.
Conclusion: The findings reveal a sub-optimal level of engagement from HEWs,

community networks and other relevant bodies in PHEM preparedness and
response activities at the community level, irrespective of the standard. Half of
HEWs were not engaged in surveillance data analysis or information generation
for local use. Furthermore, 13% of HPs were not reporting surveillance data
to their catchment health facility, reducing the geographical representation
of the surveillance data. The HEWs’ and community networks’ lower level of
basic knowledge about community-level surveillance activities, the fragmented
coordination of community-level networks (structures), the lack of and limited
access to locally translated guidelines, case definitions and reporting tools, sub-
optimal monitoring and follow-up, limited capacity-building for the community
health system structure and limited funding for community-level surveillance
activities were identified as common challenges to community-level surveillance
implementation. These findings on surveillance components, performances
and challenges can potentially affect the minimum standard of 80% facility
representativeness and geographic coverage of community-level surveillance
performance to detect and respond promptly to public health emergencies
(PHEs). Thus, we can conclude that the national HEP is not contributing much
to addressing the basic components of PHEM at the community level as per
the expected standard. This study also showed, however, that, if the HEP takes
the required measures to improve and support the system’s structure, it can
play a significant role in the real-time detection, reporting and response to
PHEs at the community level. This may be evidenced by the finding that the
country’s primary-level surveillance system performance (a core function of the
system) depends on the availability of inputs from the system’s structure and
supporting functions.
Recommendations: We recommend improving HEWs’ knowledge of the basics

of PHEM through ongoing training and mentoring, the provision of standardized
PHEM guidelines and reporting formats, upgrading the reporting mechanism
at the community level and improving the HEP’s involvement in the PHEM
system by allocating a sufficient budget and engaging in preparedness and
response activities.
page- 843
page-844
1. Background
Populous developing countries have several challenges in addressing the basic

health needs of their population. Poverty, ignorance, a lack of resources and
health facilities and poor leadership acting in concert produce a wide spectrum
of diseases.1 To address the shortcomings of service coverage within the health
system, accelerating the expansion of and/or strengthening Primary Health Care
services by engaging community members at the grassroots level is critical.2,3
Ethiopia’s health system is structured in 3 tiers: primary, secondary and tertiary

levels of care. The primary level of care includes primary hospitals, Health
Centers (HCs) and Health Posts (HPs). The primary health care unit (PHCU),
where primary health services are provided and the unit is most accessible
to the general population, comprises 5 satellite HPs (the lowest-level health
system facility, at the village level) and a referral to an HC.4 A further network
links the community with the PHC system at the community level. A Women
Development Army (WDA), the basic structural unit, is divided into Health
Development Teams which comprise up to 30 households residing in the same
neighborhood. The HDT is further divided into smaller groups of 6 members
(households), commonly referred to as “one-to-five” networks.3
The Health Extension Program (HEP), which operates at the primary level
of the Ethiopian health system, was launched by the Ministry of Health
(MoH) in 2003 with the goal of improving health outcomes in Ethiopia by
targeting households and communities. The program delivers a package of
basic and essential promotive, preventive and curative health services targeting
households in a community. The focus of the program is on improving household
behaviors and providing basic health services that have high impact and are
cost-effective.5-7 Besides these tasks, the HEWs are also expected to undertake
disease-surveillance activities in collaboration with other community network
leaders at the kebele/community level for selected priority public health
problems by using community case definitions.8
page- 845
The Role of the Health Extension Program
1.1 Disease Surveillance in Ethiopia

Surveillance activities for selected public health problems in Ethiopia were
begun in 1996. In 1998, the World Health Organization’s African office (WHO/
AFRO), following the resolution of the 48th assembly, began promoting
Integrated Disease Surveillance and Response (IDSR) for all member states to
adopt as their main strategy to strengthen their national disease surveillance
systems. As a member state, Ethiopia adopted this strategy, which is district-
centered and outcome-oriented.
After the 2009 nation-wide Business Process Re-engineering initiative, the

country’s surveillance system was upgraded to the Public Health Emergency
Management (PHEM) system in order to integrate events and conditions
having national public health importance along with other infectious conditions.
The surveillance system in Ethiopia operates at both the health facility and
community levels. Community-level surveillance, which begins in the community,
is expected to complement healthcare facility-based surveillance.8
1.2 Public Health Surveillance at

Primary Health Care Level
Nationally, HEWs (who perform their functions at Primary Health Care Units) are
responsible for undertaking surveillance at the community level in collaboration
with community members and available community network leaders. With
respect to nationally notifiable disease conditions in the community, HEWs are
expected to: 8,9
• identify their occurrence,

• report any confirmed or possible cases to the nearest Health Center,
• study suspected cases,
• identify those affected,
• determine where and when the disease is most common,
• actively search for other cases by doing home visits, inform the community
about cases in the area,
• work with community members to find more cases,
• assist the District Health Authorities to treat cases and control the spread
page-846
of the disease,
• mobilize and educate the community to prevent the disease from
spreading and
• keep the community informed about the cases that have been identified
and how they are being managed.
The surveillance system’s viability depends on its components, structure, core

functions, supporting functions and monitoring and evaluation. Ensuring the
availability and functionality status of these components is crucial for the rapid
detection of any public health threats, preparedness related to logistics and
funding administration and a prompt response to and recovery from various
public health emergencies (PHEs).10
1.3 Statement of the Problem

Emerging and re-emerging infectious diseases pose serious public health and
global economic threats in the 21st century. Many of these infectious diseases
originate in Africa, and the continent has experienced, over the past decade,
epidemics of diseases including cholera, dysentery, meningitis, yellow fever, zika
virus, chikungunya and Ebola virus, which have resulted in significant morbidity
and mortality.11,12 Rapid human development, including numerous demographic,
population and environmental changes have accelerated their emergence and
re-emergence. Furthermore, the resurgence of the microbial threat, rooted in
several recent trends, has increased the vulnerability of all nations to the risk
of infectious diseases, whether newly emerging, well-established or deliberately
caused.13
Ethiopia is among the sub-Saharan African countries that experience repeated

occurrences of PHEs. Its commonly encountered public health problems are
meningitis, measles, malaria, diarrhea, intestinal helminthiasis and acute
respiratory infections, including pneumonia and tuberculosis. Outbreaks of
emerging and re-emerging problems like dengue fever, yellow fever, chikungunya,
West Nile virus and cholera are also common.14
page- 847
Even though Ethiopia met United Nations Millennium Development Goal 4

(MDG4) by reducing child deaths by 67%—from 204 per 1,000 live births
in 1990 to 59 per 1,000 live births 3 years ahead of schedule through the
Health Extension Program, communicable diseases and maternal, perinatal
and nutritional conditions still constitute the primary sources of diseases burden
in Ethiopia.15
Studies show that having improved laboratory services and disease surveillance
systems by engaging community members in surveillance activities are vital to
early identification and monitoring of disease trends and the initiation of public
health action.16
Providing an early warning of potential threats to public health and program

monitoring functions are the main goals of surveillance. Information generated
for decision-making from reliable population data and complete coverage
contributes significantly to national health security and helps monitor trends
of endemic diseases and progress toward disease-control objectives. Moreover,
it provides information that may be used to evaluate the effect of disease-
prevention and -control programs at the grassroots level.10,17
Ensuring the availability and functionality status of surveillance components

is critical to availing representative, reliable and real-time data for decision-
making processes for monitoring and ensuring the early detection of public
health problems in order to take action.
page-848
page- 849
page-850
2 Objectives and Scope of the Study
2.1. General Objective

The main objective of this study is to assess the current role of the HEP in the
PHEM System for PHE preparedness and response.
2.2. Specific Objectives

The specific objectives of this study were to:
• assess HEWs’ knowledge of PHE early warning, preparedness and
response;
• assess the availability and functionality of basic components of
surveillance activities at the PHC level;
• assess community-level networks’ engagement in early detection and
surveillance, outbreak investigation and response, recovery and resilience
activities (preparedness); and
• explore the early warning and surveillance, preparedness, response and
recovery-related activities carried out by HEWs.
2.3 Research Question

This study examined the existing PHC-level surveillance system structure in terms
of its availability and stakeholder engagement, the existence of supporting
functions, the core components of functionality, monitoring and evaluation so
as to avail real-time and representative data for decision-making processes at
all levels.
page- 851
Table 1: Main Research Questions Addressed by the Assessment
Area of assessment Research Question

Structure of the systems To what extent are relevant stakeholders engaged in surveillance
activities at the primary health care level?
Supporting functions Does the primary health care surveillance system have sufficient capacity
and resources for detecting, reporting, analyzing and responding to
public health problems?
Core functions To what extent can the primary health care surveillance system detect,
register, report, analyze and provide response and recovery activities
at the community level?
Data quality What is the level of the primary health care surveillance coverage and
representativeness to avail real time information for decision-making?
Outcome What has been the role of the primary health care surveillance system
in providing alerts and availing real-time information to decrease
mortality and morbidity due to PHEs?
2.4. Scope of the Study

This assessment covers all components of the surveillance system at the PHC
level.
Figure 1: Scope of the assessment
page-852
2.5 Indicators of the Study

Selected indicators for measuring the primary-level surveillance system
components include:
1. Structure of the system (6 indicators):

· Proportion of respondents who know of the existence of notifiable
conditions in their locality
· Proportion of structures with a community network for community
surveillance
· Proportion of structures with a mixed (active and passive) surveillance
strategy
· Proportion of structures with an emergency coordination platform
· Proportion of structures with an allocated budget for surveillance
· Proportion of structures with collaborating sectors (partner
engagement)
2. Core functions of the system (9 indicators):
· Proportion of structures with a community case definition
· Proportion of structures with a rumor logbook and data archiving
· Proportion of structures reporting surveillance data weekly and/or
immediately
· Proportion of structures with an available means of verification
· Proportion of structures that conduct data analysis and summary
reports
· Proportion of structure with available information communication
channels
· Proportion of woredas having a preparedness plan
· Proportion of structures with prevention and control activities during
the last 12 months
· Proportion of structures with a regular feedback mechanism
3. Supporting functions (4 indicators)
· Proportion of structures with a reporting format
· Proportion of structure with guidelines or access to guidelines (PHEM
or disease-specific)
· Proportions of staff trained on community-based surveillance
· Proportion of structures with functional supervision
4. Data quality:
· Proportion of structures that regularly monitor timeliness and
completeness measures
page- 853
page-854
3. Significance of the study
The current PHEM system has been in place since 2009, replacing the
Integrated Disease Surveillance and Response (IDSR) approach, which focuses
on epidemics and fails to consider public health conditions like nutritional
problems or maternal and perinatal deaths. The data reporting for the existing
PHEM approach, which starts from the community level and is integrated
with PHC service by using community-level networks, can contribute better
geographical coverage and data representativeness.8
During its years of implementation, the PHEM system has contributed greatly
by providing real-time public health alerts, availing real-time data for program
monitoring, planning and strategic decision-making at all levels, coordinating
PHEs in collaboration with different stakeholders and participating in emergency
recovery activities. This, in turn, contributes strongly to a significant reduction in
mortality and morbidity resulting from PHEs.
Despite these achievements, PHEs continue to affect the community and cause
significant financial damage to individuals, communities and the nation. This
can reflect functionality and/or availability problems in the surveillance and
early warning components of the system, which is responsible for generating
reliable, representative and real-time data for the monitoring and forecasting
of PHEs. In addition, the engagement of relevant stakeholders, including the
community, may also contribute to effective PHEM.
Assessing the existing system’s capacity and reviewing its strengths, weaknesses
and opportunities for further strengthening of the system are crucial steps in
the development of a strategic plan of action for implementation. This study
aims to assess the functionality and existing capacity of PHC surveillance in
addressing the basic processes of the PHEM system for PHE preparedness and
response at the community level.
page- 855
page-856
4. Methods
4.1. Study Area and Period

Ethiopia has 9 regional states and 2 city administrations. All regions are
administratively divided into zones, woredas and kebeles. Woredas constitute
the lowest budgetary structure, and the kebele is the lowest administrative unit.
City administrations are also divided into sub-cities, which are further divided
into woredas and kebeles. This study was conducted in all 9 regional states:
Tigray, Afar, Amhara, Oromia, Somali, the Southern Nation Nationalities and
Peoples Region (SNNPR), Benishangul-Gumuz, Gambela and Harari, as well
as one city administration: Dire Dawa. According to estimates from the Central
Statistical Agency, in 2019, the population of Ethiopia was 98.7 million with
78.8% living in rural areas.18
The HEP has been implemented since the second phase of the Health Sector
Development Program (HSDP-II) in early 2004. The HEP was started with
the first version of the program targeting agrarian communities. Subsequently,
the program was adapted to pastoralist communities and urban settings.
Currently, the program encompasses customized sets of packages for agrarian,
pastoralist and urban populations. HEWs working in community HPs linked to
HCs constitute grassroots-level implementers of the HEP. Health administrative
institutions at the woreda, zone, region and federal levels are expected to
provide programmatic guidance and inputs to the HEP.19,20
Currently, the HEP is being implemented through more than 40 thousand

HEWs and 17 thousand HPs. This study was conducted from June to July 2019
.4.2. Study Design

A cross-sectional study design with multi-stage sampling was used to investigate
the role of the National HEP in the PHEM System. Using a conceptual
framework from the Primary Healthcare Performance Initiative (PHCPI) and a
modified version of the WHO framework for evaluating communicable disease
surveillance systems, four key areas/components of the surveillance system were
page- 857
assessed: the system’s structure, core components, support functions and data-
quality attributes.
4.3. Source Population

All HPs, HCs, Woreda/District-level Health Offices, HEWs and WDA leaders
constitute the source population for the assessment.
4.4. Study Participants

The study participants for the quantitative part of this assessment include:
1. Woredas/districts selected from all9 regions and the Dire- Dawa city
administration;
2. HCs selected from the study woredas;
3. HPs selected from the catchment HCs; and
4. HEWs working in the study HPs.
Study participants for the qualitative part of the study include purposively
selected informants from:
1. Health Offices of the study woredas;

2. HCs of the study woredas;
3. HEWs; and
4. Women’s Development Army leaders at the community level.
4.5. Sample Size

4.5.1. Sample size for quantitative study
Quantitative data were collected from the woredas/districts, HCs and HPs
selected for the study. Catchment woredas/districts and HCs were automatically
selected based on the HPs included in this study.
The number of HPs required for the assessment was calculated using a sample
size determination formula to estimate a single population proportion. The
2016 service availability and readiness assessment (SARA, 2016) was used
as the source of initial estimates for different proportions related to the HEP.
Sample size was calculated, assuming a 95% confidence in the estimate of the
true proportion and a 5% margin of error (d=5%) for each of the following
page-858
HP-level variables: the percentage of HPs with latrine facilities (60%), the
availability of basic equipment among HPs (57%), the percentage of HPs with
non-zero stock of ORS (40%), percentage of HPs providing family planning
services (95%) and the percentage of HPs with at least one staff member
trained to diagnose and treat malaria (47%). The HCs included in the study
were determined based on the pre-estimated sample size for the HPs. With due
consideration to the assumptions behind different statistical tests and logistics-
related issues, 2 HPs per each woreda were selected through a lottery method
in areas with more than 2 HPs. A total of 142 HEWs, 96 HPs, 52 health facilities
and 54 woredas were included in this assessment (Figure 2).
4.5.2 Sample size for qualitative data

The sample size for the collection of qualitative data (focus group discussions
[FGDs] and key informant interviews [KIIs]) was determined by considering
the diversity of potential information sources about the HEP. Data were
collected at all levels of the health system in each region and the Dire Dawa
city administration. At each level, data were collected both from units directly
responsible for the management of the HEP and from those who use the HEP
as a platform to implement their programs.
4.6. Data-Collection Tools and Data-Quality

Management
Four PHEM experts prepared standardized qualitative and quantitative
assessment tools (guides and questionnaires) incorporating the required
information at the Woreda Health Office, HC, HP and the community levels.
The locally translated tools were pre-tested in communities outside of the sample
woredas prior to data collection. The data-collection tools and procedures were
refined based on observations from field-level pre-testing. All quantitative data
collection tools were entered into the ODK electronic data entry template with
integrated data quality assurance features to be administered using Android-
based tablet computers.
A quantitative data-collection method was applied to address the basic

components of the PHEM system carried out by the HEP. Separate standard
page- 859
questionnaires were administered for: 1) HP heads, who are responsible for

leading the HPs, 2) HEWs in the visited HPs, 3) HEP supervisors and PHEM
focal persons at the catchment HCs and 4) HEP and PHEM coordinators
at District (woreda-level) Health Offices. In addition to the standard
questionnaire, secondary documents and onsite observation methods were
reviewed to identify the PHEM-specific components like keeping the reported
surveillance data and supplementary materials. The quantitative method also
incorporated assessment questions to review the basic surveillance components
in order to identify the capacity and functionality of the HEP in addressing the
basic PHEM process, including: 1) the system’s structure, 2) the system’s core
components, 3) the system’s supporting functions and 4) the quality of the data.
In addition, the tool includes questions to determine HEWs’ knowledge of basic
components of PHEM and the engagement level of HEWs and community
networks on the implementation of early warning, preparedness, response and
recovery activities.
The qualitative assessment includes FGDs with the community-level structures

(e.g., community leaders, Women’s Development Armies) and KIIs with officials
of the Woreda Health Offices, HCs and HPs. The qualitative assessment tool
incorporated qualitative inquiries that addressed: 1) a general overview of
surveillance activities at HPs; 2) infrastructure and communication needed for
surveillance; 3) capacity-building training related to surveillance; 4) available
health information systems; 5) access to essential resources for surveillance-
related activities and 6) leadership/governance regarding surveillance activities.
All data collectors were trained on the general guidelines for data collection,
data-collection methods, sampling and data-collection procedures, the contents
of each data-collection tool and the basic components of PHEM practices. The
collected data were verified in the field for accuracy and completeness. The
validated data were uploaded from each individual tablet to a secure central
server residing in the MERQ data center. Each team’s survey implementation
was tracked by the central data manager and project coordinator based on
the output from the collected data. Data cleaning (i.e., screening data for
duplication, internal consistency, out of range and invalid values and outliers)
was carried out during and after the completion of the survey.
page-860
4.7. Sampling strategy

The study participants for the qualitative-data collection (FGDs and KIIs) were
selected by considering the diversity of potential sources of information about
the HEP. A total of 96 HPs were visited, representing 9 regions and 1 city
administration, 43 zonal structures, 54 woreda structures and 52 catchment HCs.
From the selected administrative structures, a total of 141 HEWs, 1 manager
or official and/or officer from Woreda Health Offices and 1 HC manager
from HCs were contacted for the KII. Figure 2 presents the summary of study
participants by selected administrative structures.
page- 861
Figure 2: Sampling strategy and number of study participants enrolled in the

assessment
page-862
Figure 3: Geographical distribution of sample zones
4.8. Data management

4.8.1. Quantitative data analysis
Quantitative data from the ODK database were exported to Stata 14 for
cleaning and analysis. Descriptive statistics with regional disaggregation were
used to determine the implementation status of the HEP in each of the 9
regional states and 1 city administration. The findings were presented in tables,
graphs and narrative summaries.
4.8.2. Qualitative data processing and analysis

The qualitative data collected from the KIIs and FGDs were transcribed and
translated into English within 2 days of the interview or discussion to maintain
fresh memories. Thematic content analysis assisted by NVIVO version 12
page- 863
software was used for data analysis. A common coding framework aligned
with the PHCPI theoretical framework was applied. During data analysis, the
focus was on identifying major determinants of the performance of the HEP
in addressing the basic PHEM process. The qualitative data were triangulated
with quantitative findings to determine the bigger picture of the HEP’s role in
addressing the basic PHEM process at the community level.
4.9. Ethical clearance

Ethical clearance was obtained from the Institutional Review Board of the
Ethiopian Public Health Institute, and permission to conduct the study was
obtained from different levels of the health system before initiating data-
collection activities. Before each interview, all study participants were informed
about the purpose, significance and content of the study. Informed consent
to take part in the study was obtained from each respondent after they were
provided with adequate information about the study.
page-864
page- 865
page-866
5. Findings
5.1. Characteristics of Study Participants

A majority (77.54%) of HEWs were unmarried, lived inside the kebele where
they worked (72.59%), had more than 5 years’ experience (78.4%) and had a
level III or IV (86.4%) educational status. A majority of HC-level respondents
(83%) were B.Sc. nurses and health officers, and 63% had 1-5 years of work
experience in the health system (Figures 4 and 5).
Figure 4: Characteristics of Health Extension Workers included in the assess

ment
page- 867
Figure 5: Characteristics of HC- and woreda-level respondents included in the

assessment
5.2. Overall Knowledge Level of HEWs

The knowledge of HEWs related to PHE condition surveillance (i.e., the
reporting, alert and action threshold for response) was assessed. The mean
knowledge score of the respondents was 8.16 out of 13 questions. Most (100;
70.9%) HEWs scored above the expected mean knowledge score. Among the
respondents, 41 (29.1%) scored below 55%, and 11 (7.8%) scored above 85%.
The highest-scoring HEWs were from the Amhara 6 (28.6%), Afar 1 (16.7%),
Oromia 3 (8.8%) and SNNPR 1 (3.1%) regions (Figure 6).
page-868
Figure 6: Distribution of HEWs’ knowledge score by region
5.3. Structure of the Primary Health Care

Surveillance System
5.3.1. Knowledge on the existence of notifiable diseases/event

conditions
A total of 114 (81%) HEWs knew the existence of mandatory notifiable disease
conditions in their locality. Of the interviewed HEWs, all from the Gambela
region (19; 90.48%) and 14 (90.33%) from both the Amhara and Somali
regions, respectively, knew the nationally notifiable disease conditions. Forty-
nine (94.0%) of their catchment HC-level respondent and 45 (90.0%) woreda-
level experts knew the existence of diseases/event conditions that are under
surveillance in their catchment community structures.
Among the HEWs, 100 (70.9%) identified 1-10 notifiable diseases/event

conditions, and 8 (5.7%) identified 10-15 in their locality against the total
number of 15 diseases/events and conditions to be notified. The number of
disease conditions identified ranged from 1 to 22. On the other hand, 22
(44.0%) woreda-level PHEM experts and/or HEP coordinators identified only
1-5 disease conditions expected or needing to be reported by HEWs working
in their catchment; 16 (32.0%) respondents listed 15-22 reportable disease
conditions.
page- 869
“
Regarding community members’ awareness of and engagement in the detection
and reporting of cases, during the FGD 1 WDA leader reported the following
We always report cases like tetanus, vomiting, bleeding, neonatal
disease and fever. These are disease conditions that need
immediate reporting. If there are no cases, we report zero.
FGD discussant, WDA leaders
“
Another participant described case reporting as follows:
Maternal death, influenza-like illness and rabies (dog bite) are
also diseases that need immediate reporting.
Awareness of the reportable diseases or conditions can significantly improve

the early identification and reporting of cases in addition to its benefits for
improving health-service-seeking behavior of the community. Regarding
improvement in the community’s health-seeking behavior, one FGD participant
“
assertively described how the HEW could facilitate community-level health-
seeking behavior:
Meningitis, even though I have not encountered it, I have seen
people with a stiff neck, so when this happens, I told them to go
to the health care immediately and get treated.
5.3.2. Available community networks for surveillance activities

Among the visited HPs, 59 (83.10%) have functional community-based networks
that support surveillance activities. Of these networks, the Women’s Development
Army is the dominant structure; it was available at 52 (73.23%) HPs. Different
types of community networks were mentioned by HC-level respondents as
surveillance agents. A total of 43 (81.0%) mentioned Health Development
Army structures, 25 (47.0%) mentioned clan leaders, 16 (30.0%) mentioned
Agricultural Development Armies, 20 (38%) mentioned local respected cabinet
members in the community and 25 (47%) mentioned community key informants,
like religious leaders, teachers, traditional healers, traditional birth attendants,
militia and police forces. Other community structures mentioned by participants
as taking part in community surveillance activities were community volunteers,
epidemic task forces and ‘got’/village leaders.
page-870
Information gathered from Woreda Health Office respondents showed that

15 (30%) of woredas have a disease/unusual event notification and reporting
mechanism through house-to-house visits by the HEW and community
involvement. Community members in their catchment play a key role in
reporting diseases/events directly to the nearest facility either by informing
them in person, informing the HEWs or reporting to health facilities in writing.
The linkage of community structures and HEWs in most of the visited HPs was
found to be weak, however. One key informant said:
“
The linkage between the community and the Health Post is not
functional. Initially, there was a strong WDA in the kebele, and
reporting had been good. Now, Women’s Development Armies
in the kebele are not performing their duties and responsibilities.
Key informant, HEW
The lower level of attention paid by political leaders to engaging community

networks in the job was indicated as the main reason for the current status by
another key informant at the HP.
“
We already stopped working with community groups and networks.
Initially, it had been good. The problem is, there is no political
commitment to engage them in this job.
Key informant, HEW
Another key informant also responded similarly:
“
Our community linkage is influenced by the attention of higher
government leaders. When they pay it good attention, it becomes
strong. When they pay it no attention, everything from the woreda
to kebele becomes weak. So, we don’t have a strong force, and we
can do nothing regarding this issue.
Key informant, HEW
page- 871
5.3.3. Surveillance strategies

Based on the findings, 100 (70.92%) HEWs are practicing a mixed surveillance
strategy (both house-to-house visits and an HP register) to collect and compile
disease conditions for reporting. These HEWs were from Harari (n= 4), Dire
Dawa (n= 9) and the SNNPR (n=30).
Forty health facilities (77.0%) use both community reports and facility visit
registers. In addition, 48 (96.0%) woreda-level respondents confirmed that they
undertake surveillance activities with the involvement of HEWs working in their
catchment areas.
Key informants were also asked about the engagement of community-level

networks and community participation in the community-level surveillance-
related activities. The findings clearly show that there are some areas where
the community network is functioning very well.
One key informant explained the status of community participation:
“
You know that our community networks, especially the 1-to-5
and 1-to-30, are active in the community. We help them with
various activities, and they help us a lot. It’s an amazing thing.
I go together with the network members to help those affected
and at-risk community groups in collaboration with other Health
Extension Workers.
Key informant, HEW
Twenty-eight HEWs (54.0%) stated that health facility surveillance focal persons
were responsible for compiling and reporting the community surveillance data
collected and reported by HEWs. On the other hand, at 20 (38%) health
facilities, Health Extension supervisors were responsible for compiling and
reporting the surveillance data. Among woreda-level respondents, 20 (40%)
said that the primary responsibility for community-based surveillance and
community health services were given to HEW coordinators, and 26 (52.0%)
stated that the responsibility fell to woreda-level surveillance coordinators.
page-872
5.3.4. Emergency coordination platform
One preparedness activity that needs to be done before any PHE is to have a
coordination and collaboration structure at all levels. Among study participants,
71 (50.4%) confirmed the existence of an emergency coordination task force
in their kebele. More than 50% of each region’s HEWs, except for Tigray
(100%), Somali (73.33%), Gambela (66.67%) and the SNNPR (59.38%), said
that they had no kebele-level emergency coordination platform for emergency
preparedness and response.
At the level of the catchment HC, 38 (72%) facilities have a rapid response
team, and only 28 (74%) had a Health Extension supervisors as a member of
the Rapid Response Team (RRT). Only half of the HPs conducted a regular
RRT meeting.
Of the visited woredas, 40 (80%) have an RRT comprising HEWs and

community members. A coordination meeting between the HEP and PHEM
section exists in 34 (68%) woredas. Among these, a quarterly coordination
meeting is held in 18 (36%) of the woreda structures.
One KII interviewee from an HP where kebele-level emergency coordination
was available said:
“
We had emergency coordination platform at the kebele level. We
have a strong connection and linkage with the WDAs, community
and religious leaders. We work together on all issues regarding
our community health.
Key informant, HEW
One HP-level key informant suggested that to have a strong emergency

coordination at the kebele level, local political leaders should be involved in
PHE-related preparedness and response activities:
page- 873
“
In the case of the emergency plan and response, even though we

have an emergency plan, we are not performing it properly. There
are no drugs that are allocated only for emergency conditions,
even though the drugs are provided for in the government budget.
To improve such challenges, training must be provided for us and
logistics should be availed for all emergency-related problems
and conditions. Lastly, to have better public health emergency
coordination at the community level, the government body must
emphasize the issue, and it should be handled by local political
leaders.
Key informant, HC Head
5.3.5. Budget allocation for surveillance-related activities
Only 6 (11%) health facilities allocated part of their budget to support community-
level surveillance activities. Of the allocated budgets, 4 (80%) allocated for a
per diem and transportation allowance. Among all the visited health facilities,
only 22 (41%) have supplies and logistics reserved for emergency response
purposes. In addition, the allocation of the budget and the provision of a
transportation mechanism to HEWs for surveillance-related activities were
implemented in only 9 (18%) of the woreda structures. (Table 2)
Based on the findings of FGDs and KIIs, there were insufficient resources to
provide the required medical supplies, drugs and logistics during emergencies.
In addition, they were not provided with adequate health care services during
emergency situations. The respondents also pointed out that the available
medication did not reach HPs in a timely manner.
“
Sometimes when we go to the HC to seek health services, we don’t
receive adequate service due to electricity problems. Mothers and
sick people visit the Health Center, but there is no electricity.
When the grid power supply is interrupted, a generator should
be installed to back up the grid power supply. But there is no
generator in our Health Center. This needs an urgent solution.
Also, medications are not available in a timely way, and clients
consider the health care providers as refusing to provide them. In
our Health Center, it is a big problem, and we need a solution.
page-874
Furthermore, almost all respondents reported that there was no training or

orientation given for HEWs and WDAs on the issue of handling emergency
cases.
“
To effectively deliver the expected service, a budget should be
allocated for prior preparedness and capacity-building activities. If
we get the required budget, we can arrange logistics and training
regarding public health emergency management for HEWs
and WDAs. In addition, there are no guidelines or protocols or
emergency medication. So, if we think of having a better public
health emergency system, these things need to be fulfilled. Many
staff and WDAs have to be trained.
Key informant, Woreda PHEM officer
5.3.6. Partnership for community-level surveillance activities
Forty-three (86%) woredas had engaged members of the community or social

groups and associations, the Women’s Development Army or community members
in public health surveillance and response-related activities. On the other hand,
30 (60%) Woreda Health Offices have been performing surveillance-related
activities with different governmental and non-governmental stakeholders in
the past several years.
Regarding the participation and engagement of all stakeholders in kebele-level

surveillance and response-related activities, one key informant from an HP
“
expressed the challenge they face:
The main challenge is the low involvement of stakeholders during

normal times and other times when we need them. For me, that
is the big challenge. Politically appointed leaders show some
reluctance when we ask them for some help. It is obvious that
some activities cannot be handled with our capacity. In that case,
the politically appointed leaders should actively help solve the
problems.
Key informant, HEW
page- 875
Table 2: Distribution of Structure of Primary Health Care-Level Surveillance System Indicators

Afar Amhara Dire Dawa Gambela
Level Categories
No % No % No % No %
2.1 Knowledge of existence of
notifiable conditions in their 5 83.3 19 90.5 2 20 9 100
Health Post level
locality
2.2. Availability of a community
5 83.3 8 88.9 1 100 2 66.67
network
2.3. Surveillance strategy 5 83.3 13 61.9 9 90 7 77.78
2.4. Available emergency
1 16.7 9 42.9 1 10 6 66.67
coordination platform
notifiable conditions in their 3 100.0.0% 9 100 2 100% 2 100%
locality
Health Center Level

2 67.0% 9 100 1 50% 1 50%
network
2.3. Surveillance strategy 3 100% 8 89.0 2 100% 1 50%
0 0% 9 100 1 50% 1 50%
2.5. Budget allocation for
0 0% 0 0 1 50% 0 0%
surveillance-related activities
notifiable conditions in their 3 100% 8 89% 1 100% 1 50%
locality
Woreda/district level

0 0% 2 22% 0 0% 2 100%
network
2.3. Surveillance strategy 0 0% 3 33% 0 0% 1 50%
1 33.0% 8 89% 0 0% 1 50%
2.5. Budget allocation for
0 0% 2 22% 0 0% 0 0%
surveillance-related activities
2.6. Partnership 2 67% 3 33% 1 100% 1 50%
page-876
s by Region
Harari Oromia SNNP Somali Tigray Grand Total
No % No % No % No % No % No %
3 75 26 76.47 28 87.5 14 93.33 8 80 114 80.85
2 50 19 86.36 10 90.91 8 72.73 4 100 59 83.1
4 100 17 50 30 93.75 9 60 6 60 100 70.92
2 50 12 35.29 19 59.38 11 73.33 10 100 71 50.35
3 100% 10 91% 10 100% 4 67% 6 100% 49 94%
2 67% 7 64% 9 90% 3 50% 6 100% 40 77%
2 67% 9 82% 7 70% 4 67% 4 67% 40 77%
3 100% 6 55% 7 70% 5 83% 6 100% 38 73%
1 33% 1 9% 0 0% 0 0% 2 33% 5 10%
1 100% 10 83% 9 90% 7 100% 5 100% 45 90%
0 0% 4 33% 2 20% 0 0% 2 40% 12 24%
1 100% 4 33% 2 20% 0 0% 3 60% 14 28%
1 100% 9 75% 8 80% 7 100% 5 100% 40 80%
0 0% 2 17% 2 20% 1 14% 2 40% 9 18%
0 0% 8 67% 6 60% 5 71% 4 80% 30 60%
page- 877
5.4. Core Functions of the Primary Health Care

Surveillance System
5.4.1. Case detection

Uniform understanding of the number of diseases/event conditions to be reported
at all health system levels is crucial for ensuring a functional early-warning
system, monitoring diseases and events under surveillance and providing a
timely response for all relevant PHEs.
Among the interviewed respondents from the HPs, 93 (66.0%) claimed to know
the number of diseases/event conditions to be reported immediately. In addition,
111 (78.7%) HEWs reported knowing the number of diseases/event conditions
to be reported weekly. Among the reportable events, malaria (n=84; 59.6%),
cholera (n= 72; 51.1%), anthrax (n=91; 64.5%) and measles (n=78; 55.3%) were
the most common diseases/event conditions reported by a majority of HEWs as
notifiable diseases/event conditions in their localities. Conditions and diseases
like Guinea worm, maternal death, acute febrile illness, perinatal death and
VHF (Viral hemorrhagic fever) are mentioned by a smaller proportion of
participants (Figure 7).
Figure 7: Proportion of respondents who mentioned specific disease conditions

as reportable diseases/events in their locality
page-878
The qualitative study also found that malaria, measles, scabies, the common
cold/influenza-like illnesses and diarrheal disease are some of the major public
health problems that require immediate reporting. In addition, typhoid fever
and typhus fever were mentioned as critical public health challenges. According
to the respondents, the higher burden of these public health problems in the
community were due to shortages of water. The accessibility and availability
of water supply in most parts of the community residing in rural villages was
“
significantly low. Clothes-sharing among children also makes them highly
susceptible to disease transmission.
Our main health problem is that we have stopped keeping our

sanitation in good condition because of a water shortage, and
as a result, children were mostly affected with scabies, measles,
diarrhea and malarial disease.
In addition to knowing the number and types of diseases/event conditions for

surveillance, knowing the community case definition for diseases/events under
surveillance is key to detecting and reporting them to the next higher level so
appropriate actions can be taken. Among the 126 (89.36%) HEWs who claimed
to know the community case definition of cholera, 4 (6.2%) missed its correct
case definition. On the other hand, among the 124 (87.9%) HEWs who claimed
to know the community case definition of measles, none missed its correct
case definition. All respondents who were able to correctly define measles and
cholera claimed that they treat and send these patients to catchment health
facilities when they get these types of cases in their catchment area or kebele.
5.4.2. Case registration and archiving
Of the total visited HPs, 50 (70.4%) kept a copy of reports at the HP. In
addition, 117 (83.0%) HEWs confirmed that they kept a copy of the surveillance
reports they sent to catchment HCs. Only 16 (31%) catchment health facilities
had a rumor logbook for capturing any disease surveillance-related rumors
from different information sources. Of these facilities, 14 (27.0%) registered
rumors in the registration book within the past 12 months.
page- 879
5.4.3. Case reporting
HEWs are expected to report surveillance data for their catchment health
facilities. Based on the results of the assessment, 124 (87.94%) HEWs claimed
that they sent surveillance reports to their catchment health facilities.
Of the woreda-level respondents, 47 (90.0%) agreed that HEWs reported

surveillance data to the facility even if the reporting timeline varied among
respondents. About 13% of woreda-level respondents (n=6) expected a
surveillance report from HEWs immediately, 2 (4%) expected the report daily
and 39 (83%) expected the report weekly regardless of the type of the diseases/
events reported.
Based on the findings from key informants, routine reporting from the HP is
done on a weekly basis. The HPs may report daily, however, when they observe
emergency conditions that need attention. On the issue of reporting timelines
“
and information sources, one HP worker stated:
We report on a daily and weekly basis. The weekly report is a

routine activity. If there is some new issue or problem, we can
inform and report to the Health Center on a daily basis. Here,
we get reports mostly from Women’s Developmental agents of
the community. However, we can also get information and report
from the community leaders, religious leaders and other sectors,
as they are close to the community.
Key informant, HEW
“
On the type of diseases and conditions they report, another participant replied:
What we report to PHEM are malaria, measles and malnourished

children. We send it also to the Health Center monthly after
compiling the weekly reports.
Key informant, HEW
page-880
Regarding the role of community networks in the reposting of cases, one key
informant stated:
“
I think a key aspect of case surveillance is the involvement of
community agents in identifying and reporting diseases and
conditions to us as soon as possible. This helps us minimize further
deterioration and further damage to the community due to the
problems.
Key informant, HEW
Another respondent also spoke about the reporting of cases by community
“
networks:
Our duty is coordinating the community. We get enough

information from the heads of the kebeles and zones. They will
bring the report to the Health Post or inform us by phone if cases
and new issues happen in their community.
Key informants also explicitly mentioned the delays in reporting disease

conditions by the community and the lack of attention paid by the WDAs to
the identified cases in the community as the main issues and challenges they
faced from community networks:
“
Delay of the report and giving less attention have been a
challenge, as most Health Developmental agents are illiterate.
To solve this problem, we included male leaders—those who read
and write properly.
Key informant, HEW
page- 881
Regarding capacity-building activities for better detection and reporting of
“
diseases/events and conditions, one HP key informant said:
We have people who were trained from each ketena of the

community. In case of an emergency, they will send us a report.
They continue saying “the trained people are from the kebele’s
women leaders and WDAs. During the training/orientation
session, we will inform them about the signs and symptoms of
disease like measles, polio and other emerging diseases and how
to send and organize reports. Here we can meet with them every
2 weeks.
Key informant, HEW
Figure 8: HEWs’ knowledge of immediately and weekly reportable diseases/

events and reporting, by regions
Regarding the means of surveillance reporting, 47 (28%) health facility-level

respondents claimed that the HEWs used SMS messaging and 47 (87%)
used paper-based communication. No respondent mentioned the HEWs’ use
of internet service as a means of reporting. Only 1 (2%) and 15 (29%) used
other communication channels: in person/physical contact and phone calls,
respectively.
page-882
5.4.4. Available means of verification
Regarding the availability of the means of verification for reported diseases/

event conditions by community networks or any community member, 100
(70.92%) HEWs working at HPs reported that they had a means of verification
for the reported cases. The largest segments of HEW respondents claiming this
were from the Tigray (n=10, 100%) and Amhara (n=20, 90.9%) regions.
Woreda-level respondents also said that household-level supervision was carried

out by the HEWs and Woreda Health Office focal points at 68% of woredas as
a means of verification of the reliability of reported diseases or conditions from
any community member. This strategy contributes significantly to ensuring that
surveillance reports are communicated to the next-highest level.
Regardless of this evidence, HP-level respondents raised the issue of monitoring
“
and evaluation. One study participant said in this regard:
There’s a system for collecting and compiling data for

reporting purpose. However, the system is not receiving as
much emphasis on monitoring and evaluation activities to
improve surveillance data quality as is expected, considering
that the real data is expected to be generated by the Health
Extension Worker and the health care provider.
Key informant, HEW
5.4.5. Surveillance data analysis

Among the HEWs participating in this assessment, 70 (49.64%) had engaged
in surveillance data analysis one time in the past. The most commonly practiced
data analysis and visualization approaches are producing figures through a
weekly summary report (n=67, 47.5%), monitoring diseases using dashboards
63 (n=63, 44.7%) and using charts and tables 48 (n=48, 34.0%).
On the other hand, 47 (90%) HC respondents said that they analyzed the
surveillance data that came from HEWs. In most health facilities (n=46, 88%),
PHEM focal persons were in charge of doing surveillance data compilation
and data analysis, followed by HEW supervisors (n=11, 21%). In addition, 5
(10%) facility disease-prevention process owners were responsible for this.
Woreda-level findings also showed that most HEWs (88.0%) practiced paper-
based reporting and documentation. The production of summary reporting and
page- 883
surveillance data visualization, however, was practiced in less than half of the
Woreda Health Offices.
Figure 9: Regional distribution of surveillance data analysis and visualization

practices commonly used by HEWs at the Health Post level
5.4.6. Surveillance data communication
Among the HPs visited, 51 (71.8%) communicated their surveillance data

and information generated to relevant stakeholders like catchment HCs
and community members. In this regard, 100 (70.9%) HP-level respondents
also stated that they had an available channel for surveillance information
communication with relevant stakeholders in their locality.
5.4.7. Emergency preparedness activities

Eighty-eight (62.4%) HEWs had no prior information or knowledge about
emergency preparedness planning in their woreda (Table 3).
Almost all interviewed participants at the woreda level underscored the

importance of community engagement and its role in surveillance, with special
emphasis on reportable diseases and conditions like maternal health. They
recognized that engaging the community in surveillance could reduce damage
to the community through early detection and participation during the response
to public health problems in the community.
page-884
One woreda-level key informant asserted this:
“
Engaging them [community members] in emergency surveillance
and response gave the professional the peace of mind that, if an
emergency were to occur, measures were in place to mitigate the
damage as soon as possible….I feel like I’m as well prepared as I
can be, and it gives me peace of mind.
—KII, Woreda PHEM Officer
“
Another participant reported:
Being active in such activities [engaging community members] will
decrease our fear and the damage that can follow any occurrence
of an emergency situation. Its contribution to problem detection
as early as possible is great.
Key informant, WPO
Community education materials for nationally notifiable disease conditions

were available in 70% of woredas, and risk assessment for PHE preparedness
was carried out with the involvement of HEWs and community volunteers in
72% of the studied woredas.
One woreda-level key informant spoke regarding the sample preparedness

activities performed in the woreda and the involvement of the WDA:
“
Not all people have a simple pit latrine. Even those who have a
latrine might have a poor practice of latrine use due to incomplete
excavations and incomplete construction of a concrete slab and
a superstructure for privacy. As a result, health education and
promotion are conducted by WDAs. They teach the community
by planning home-to-home visits and in community gatherings.
Key informant, Program Officer at Woreda Health Office
page- 885
Based on what the HP-level respondents reported, much has been done
on emergency preparedness and control-related activities. Most HEWs, in
collaboration with the available community networks, were engaged in teaching
the community to use insecticide-treated bed nets (ITNs) and providing health
education to the community on environmental hygiene. Based on information
received from HEWs, they work with the aim of persuading every house to
construct and use a latrine and dig a means of solid and liquid waste disposal.
One HP staff participant responded to the issue of emergency preparedness

and controlling activities as follows:
“
To prevent outbreak occurrences, we are engaged in providing
health education for the community on environmental hygiene,
the use of latrines and solid and liquid waste disposal. Once the
outbreak happens, the Health Post, while giving the available
treatment, prepares the patient for a referral to a place like the
Health Center.
Key informant, HEW
The level of engagement and support of community networks for prevention

and control activities, however, is limited. The absence of a well-organized
system for capacity-building activities was mentioned as the main reason for
their lower engagement, in addition to the poor channels of communication
with catchment HC and woreda leaders.
One HP-level key informant said:
“
There are 1-to-5 groups, and they usually interact with Health
Extension Workers. But, like I told you before, in areas where there
is malaria, they implement water drainage only based on the
education they were given, so it is somehow present. This is the
only thing that they are trained in.
Key informant, HEW
Participants noted that the failure to properly ensure community engagement

leads to lower effectiveness of any community problem that can’t be solved in
a timely way.
page-886
Regarding the significance of environmental and personal hygiene to prevent
“
occurrences of disease, one FGD participant pointed out:
If the plate we use for meals and the bottle for water are clean,
there won’t be any contamination. I believe we are not vulnerable
to hygiene-related diseases.
FGD participant, community member
WDAs were also asked about the level of preparedness and prevention activities
carried out by the community members. One FGD participant said:
“
Ideally, everybody has been taught to construct a latrine, but I
don’t think that all in the community have a latrine because we
can see excessive defecation in the open fields.
FGD discussant, WDA leader
Regardless of the WDAs’ tremendous effort and contribution, they also face
some resistance from community members. Sometimes the community does not
“
accept what they are told by the WDAs. One FGD participant commented on
the challenges they faced from the community:
We are teaching them to clean their clothes and their bodies, but
they are not doing it. And when we told them to renovate the
toilet, they replied “Okay,” but they didn’t do it.
FGD participant
5.4.8. Emergency response and control activities
Of the HEWs included in the study, 111 (78.7%) claimed that they participated
in different social mobilization and disease-prevention-related activities and
emergency preparedness, response and control activities in their locality.
Twenty-one (14.9%) reported the occurrence of public health outbreaks, like
measles and cholera, during the last 12 months.
Regarding challenges during an emergency response, woreda-level participants
mentioned several emergency situations in which their capacity to handle an
emergency was relevant. Although respondents’ responses varied, at least 3
page- 887
participants from the KIIs mentioned blamed a poor governance structure

for their poor response to emergency situations in their catchment. Experts’
lower levels of participation in different capacity-building activities were also
mentioned as factors in their inadequate capacity during emergency response
“
activities:
All the trainings are handled by the leader. There are emergency
threats, like conflict, infectious disease and drought, that could have
a significant impact on individual households and communities at
large. For this, the warranted precautionary measures are not
available. The thing I’ve thought about for a year during and after
the internal displacement was: How can I provide an adequate
water supply, solve the shortage of essential medication and food
supply for the displaced? But I failed to provide a response to
the issue immediately before severe damage happened to the
community. To improve our emergency-response capacity, the
governance structure and system should be strengthened.
Key informant, Program officer at Woreda Health Office
In addition to this challenge, the absence of a coordination platform at the

community level, as well as shortage of drugs and logistics at HPs contribute
greatly to providing a prompt response to emergencies. One key informant also
said the following:
“
In our case, there are problems of community coordination and
shortages of medications.
Key informant, HEW
Among FGD respondents, a significant improvement was observed in the

community’s health status in recent years. Almost all discussants had witnessed
a variety of changes in the health status of children due to vaccination and
other health services. The most frequently mentioned health services that helps
improve the well-being of the community are vaccination, the active engagement
and involvement of HEWs and community awareness-related activities.
One FGD participant identified the change seen in the community health status
and their contribution related to reporting:
page-888
“
Maternal deaths and neonatal deaths have declined. Because

of the vaccination service, our children’s deaths and other deaths
due to unknown causes have declined. We report such conditions
immediately, and the WDA transfers the message to the Health
Center. As a result of this, we can get medication early. Thanks to
God and the government, there is no problem regarding maternal
and child health in our area.
FGD discussant, WDA leader
5.4.9. Feedback
Routine monitoring of the surveillance system using supportive supervision and

providing feedback significantly improve the quality of data for information
generation. Of the visited HPs, 37 (52.1%) claimed that they received supportive
supervision and feedback from officials of both the woreda and the HC, but the
feedback was seen and verified in only 15 (21.1%) HPs.
The qualitative finding also showed a delay in feedback, as mentioned by an
HP key informant. One HP head spoke about this issue:
“
We have registration and reporting forms given to us from the
Health Center. We use the reporting form to send our reports.
However, we do not get timely feedback when we send the report.
The feedback comes 1 to 3 months after the reporting.
Key informant, HEW
One FGD participant also pointed out that there is limited monitoring and
follow-up from higher-level components of the health system:
“
We are learning from the Health Extension Workers and physicians
here. We have not gone anywhere to do training or any course,
and we are discussing with the community in a way that they can
accept during coffee ceremonies. But we need monitoring and
follow-up. They [HEWs] are following us here, but nobody follows
us from the woreda or zonal level.
FGD participant, WDA leaders
page- 889
Table 3: Distribution of Core Functions of Primary Health Care Level

Surveillance System Indicators, by Region

Level Categories
No % No % No % No %
Community case
0 0 3 33.33 0 0 0 0
definition
Case registration and
2 33.33 4 44.4 1 100 1 33.3
rumor logbook
Surveillance data
3 50 21 100 5 50 8 88.89
reporting
Available means of
3 50 20 95.24 1 10 3 33.33
verification
Surveillance data analysis 3 50 10 47.62 2 20 7 77.78
Health Post Level
Data archiving 6 100 20 95.24 5 50 9 100

Means of information
3 50 19 90.48 4 40 8 88.89
communication
Emergency preparedness
3 50 10 47.6 0 0 6 66.7
plan
Any community
mobilization or awareness 3 50 15 71.4 2 20 6 66.7
creation activity
Existence of PHEs during
0 0 1 4.76 0 0 4 44.44
the last 12 months
Public health prevention
2 33.33 16 76.2 5 50 8 88.9
and control activities
Feedback (checked) 1 16.67 2 22.2 0 0% 0 0
Community case
1 33.33 4 0.4444 0% 0%
definition
Health Center Level

0 7 0.7778 1 50% 1 50%
rumor logbook
Surveillance data
1 33.33 9 100 2 100% 1 50%
reporting
Surveillance data analysis 3 100 9 100 1 50% 1 50%
HEWs engagement on
0 7 77.78 1 50% 1 50%
preparedness
Community case
definition (locally 1 33.33 6 67 1 100% 0%

translated)
0 0 0% 0%
rumor logbook
Availability of means
of information 3 100 9 100 1 100% 2 100%
communication platform
page-890
Harari Oromia SNNPR Somali Tigray Gr. Total

0 0 6 27.3 1 9.09 2 18.2 3 75 15 21.13
3 75 15 68.2 9 81.8 2 18.2 4 100 41 57.75
4 100 29 85.29 30 93.73 15 100 9 90 124 87.94
2 50 24 70.59 27 84.38 10 66.67 10 100 100 70.92
2 50 14 41.18 20 62.5 6 40 7 70 71 50.35

4 100 27 79.41 29 90.63 9 60 8 80 117 82.98
2 50 22 64.71 30 93.75 2 13.33 10 100 100 70.92
4 100 9 26.5 1 3.13 3 20 8 80 44 31.21
4 100 16 47.1 19 59.4 8 53.3 9 90 82 58.16
0 0 4 11.8 5 15.6 2 13.3 5 50 21 14.89
3 75 29 85.3 28 87.5 11 73.3 9 90 111 78.72
1 25 3 13.6 4 36.4 0 0 4 100 15 21.13
2 67% 7 64% 4 40% 1 17% 3 50% 22 42%
1 33% 1 9% 1 10% 1 17% 3 50% 16 31%
3 100% 10 91% 9 90% 6 100% 6 100% 47 90%
2 67% 11 100% 10 100% 5 83% 5 83% 47 90%
2 67% 4 36% 5 50% 3 50% 5 83% 28 54%
0% 10 83% 5 50% 6 86% 5 100% 34 68%
0% 0% 3 30% 0% 0% 3 6%
1 100% 11 92% 9 90% 7 100% 5 100% 48 96%
page- 891
5.5. Supporting Functions of the Primary Health

Care Surveillance System
5.5.1. Availability of reporting format and logbook
Reporting formats for reportable diseases/event conditions were available in
49 (69.0%) HPs. Of these HPs, only 15 (21.1%) have locally translated version
of the reporting formats. At the catchment HC level, 47 (90%) facilities had
a weekly reporting formats for weekly summary reports, 23 (44%) had a daily
epidemic reporting format and 28 (54.0) had a case-based reporting format
for selected diseases like cholera AFP and measles. On the other hand, only
22 (42%) HCs had a case definition for the diseases targeted by surveillance.
Locally translated case definitions for nationally identified reportable disease
conditions were available in 34 (68%) woredas. Forty-one woreda-level
respondents (82.0%) confirmed that the weekly reporting form was the primary
reporting tool used by HEWs.
5.5.2. Availability of guidelines, protocols or packages for surveillance
Of the visited HPs, only 52 (73.2%) had guidelines, protocols or packages

to guide surveillance activities at the community level. Among these HPs, 36
(50.7%) had a translated version of the guidelines in the local language. At
the woreda level, the national PHEM guidelines were available in 39 (78.00%)
woredas. Only 5 (10.00%) woredas, however, offered access to the guidelines
that served as a guiding document for HEWs and PHEM officers.
5.5.3. Training status of HEWs on community-level surveillance
Based on woreda-level respondents, regular training for HEWs was delivered

on addressing the topics of surveillance/PHEM and reporting. Of the visited
woreda structures, 39 (78.0%) claimed that they were providing training for
HEWs on the surveillance report notification process, but 67 (47.5%) HEWs
claimed that they were not trained on community surveillance.
page-892
In addition, evidence from key informants at HPs also showed that emphasis
should be given to enable, motivate and encourage HEWs so that they would
be able to handle problems of the community more easily. In addition, high
staff turnover should also receive due attention to preserve the current service
“
delivered by HEWs. Regarding this, one HP-level key informant clearly stated:
On the contrary to the lower level of attention given by the

current government to HEWs, they play a great role in the early
detection of community problems. To strength this we have to be
equipped with basic knowledge that is updated.
Key informant, HEW
One woreda-level respondent also mentioned the importance of training and

the availability of guidance and protocols for emergency response at the
“
community level:
To effectively deliver the service, there must be training regarding

public health emergency management. In addition, there is no
available guideline, protocol or emergency medication. So, if we
want to have a better public health emergency system, things
have to be fulfilled. Many staff have to be trained.
Key informant, Woreda PHEM officer
Another key informant also discussed the importance of having basic training
“
and other required guidance and materials:
Hopefully, we can bring a better change if we are trained and get

the required material or documents. However, lots of things were
handled at the woreda level. There is a big gap in reaching the
required documents, training and logistics.
Key informant, HEW
page- 893
5.5.4. Supportive supervision
Around three fourths of the facilities (n=39; 74.0%) monitored the activities
of HEWs under their catchment either through on-site supervision or by other
means. Among them, 38 (72.0%) health facilities provided regular feedback
based on their supervision findings. Forty-four woredas (88.0%) also conducted
supportive supervision to monitor and support primary-level surveillance
activities in their catchment areas.
For better community-level implementation of PHEM activities, collaborative

work should be implemented. Besides, WDAs, WDAs, HEWs, and health
“
care providers must work to address community needs. One FGD participant
commented on how to better achieve emergency activities:
We can bring a better change if we work together to address the

community need and work in improving our community health. The
Health Extension Workers are helping us like a family member. To
do this, we have to get support from the government side.
FGD participant, WDA leader
page-894
5.6. Data-Quality Components of the Primary

HealthCare Surveillance System
5.6.1. Timeliness and completeness of surveillance data

To have a prompt and timely response for PHE events and diseases, the quality
of data received at each level should meet the minimum standard. To this end,
there is a need for regular monitoring of the surveillance data at each level.
Forty-two health facilities (79.0%) monitored the quality of the surveillance data
by measuring both its timeliness and completeness status. Almost all facilities
(n= 41; 95.00%) monitored only the completeness of the data they received
from the HEWs, and 3 (7.0%) monitored only the timeliness of the report.
Furthermore, 47 (94.0%) woredas have established a timeline for surveillance
data reporting by HEWs.
page- 895
Table 4: Distribution of Supporting Functions of Primary Health Care Level Surveillance Syste

Level Categories
No % No % No % No %
3.1. Availability of reporting format 4 66.67 7 77.78 1 100 3 100
3.2. Available guidelines or access to

0 0 2 9.52 0 0 6 66.67
Health Post Level
guideline (English version)
3.3. Available guideline or access to

2 33.33 7 33.33 0 0 1 11.11
guideline (locally translated version)
3.4. Training status of community-based

2 33.33 12 57.14 1 10 7 77.78
surveillance
3.5. Supportive supervision 4 66.67 4 44.44 0 0 0 0
3.1. Availability of reporting format

3 100% 7 78% 1 50% 2 100%
(WRF)
Health Center Level

1 33% 9 100% 2 100% 0%
guideline

0% 5 56% 0 1 50%
surveillance
3.4. Supportive supervision 1 33% 7 78% 1 50% 0%
3.1. Availability of reporting format

3 100% 7 78% 1 100% 2 100%
(WRF)

3 100% 9 100% 0% 1 50%
guideline (PHEM or disease-specific

1 33% 5 56% 1 100% 1 50%
surveillance
3.4. Supportive supervision 3 100% 9 100% 1 100% 1 50%
page-896
em Indicators by Region
Harari Oromia SNNPR Somali Tigray Grand Total

4 100 16 72.73 7 63.64 3 27.27 4 100 49 69.01
1 25 2 5.88 4 12.5 1 6.67 0 0 16 11.35
0 0 11 32.35 3 9.38 7 46.67 5 50 36 25.53
1 25 9 26.47 14 43.75 7 46.67 4 40 57 40.43
2 50 12 54.55 6 54.55 5 45.45 4 100 37 52.11
3 100% 11 100% 8 80% 6 100% 6 100% 47 90%
3 100% 10 91% 10 100% 3 50% 6 100% 44 85%
1 33% 3 27% 4 40% 0% 2 33% 16 31%
3 100% 9 82% 9 90% 5 83% 4 67% 39 75%
1 100% 10 83% 7 70% 6 86% 4 80% 41 82%
1 100% 8 67% 10 100% 4 57% 3 60% 39 78%
1 100% 6 50% 4 40% 0% 3 60% 22 44%
1 100% 10 83% 9 90% 5 71% 5 100% 44 88%
page- 897
5.7: Variability of HEWs’ Knowledge Scores at the

Woreda Level
The knowledge score of HEWs at the woreda level was assessed to check
whether their mean knowledge level was similar. Based on the findings, HEWs’
mean knowledge level differs significantly between woredas and among regions
(P < 0.001; Table 5).
Table 5. Variability in HEWs’ Knowledge, Structure and Surveillance

Components of the System
Sum of Mean
Variables df F P-value
Squares Square
Variability in HEWs’ knowledge scores at woreda level
Between groups 4.261 53 .080 4.642 < 0.001
Woreda
Within groups 1.507 87 .017
level
Total 5.767 140
Between groups 2.857 8 .357 16.201 <0.001
Regional
level
Total 5.767 140
Variability of structure of the system
Between groups .238 2 .119 10.144 <0.001
Woreda
level
Total 1.293 92
Between groups .143 8 .018 2.093 0.068
Health Post
Within groups .257 30 .009
level
Total .400 38
Between groups .105 5 .021 2.223 0.081
Health
facility level
Total .359 32
Variability in core functions of the system
Between groups .054 2 .027 1.739 0.179
Woreda
level
Total 2.399 153
page-898
Between groups .196 8 .024 2.500 0.025

Health Post
level
Total .617 51
Between groups .096 8 .012 .685 0.703
Health
facility level
Total 1.047 62
Variability in supporting functions of the system
Between groups 1.300 2 .650 48.686 <0.001
Woreda
level
Total 3.035 132
Between groups .178 8 .022 2.105 0.046
Health Post
level
Total .634 51
Between groups .439 7 .063 5.257 <0.001
Health
facility level
Total .845 41
5.8. Variability and Relationships of Surveillance

Components at Different Levels
The hypothesis is that the mean scores of different surveillance components
(i.e. the structure of the system, a core function of the system and supporting
functions) differ at the levels of the 3 health systems (i.e. the regional, HC and
HP levels).
The findings clearly show that there is variability in the structure of the system’s
mean score at all health system levels. In addition, the supporting functions
of the system differ significantly in their mean score among all health system
levels (Table 5).
Based on the findings, the performance of the core functions and supporting
functions at the HC level show a significant difference. The surveillance
supporting functions also vary between HPs and health facility levels (Table 5).
Regarding the relationship among the different PHC-level surveillance
components, the supporting function and core function of surveillance system
have a direct relationship. The structure of the system and core function of the
page- 899
system also have a positive relationship but not one that is significant compared
to supporting functions. In general, the supporting functions of the PHC-level
surveillance system are high predictors of the core function of the surveillance
activities at the community level in Ethiopia (Table 6).
Table 6: Association of the Supporting Functions and Core Functions of the

Surveillance System
Model B Std. Error Beta t p-value.
(Constant) .451 .136 3.320 0.002
1 SS_ score .103 .110 .122 .935 0.354
SF_ score .254 .105 .316 2.421 0.019
a. Dependent Variable: CF_ score
page-900
page- 901
page-902
6. Discussion
Because a large proportion of the cases of major infectious diseases did not
seek medical treatment at public health facilities, the surveillance system at
the community level is more efficient than the facility-based one in its rapid
detection of outbreaks, monitoring of communicable diseases and notification
of vital events. In addition, implementing public health surveillance at the
PHC level can supplement indicator-based surveillance and help improve
the coverage, sensitivity and responsiveness of the available routine health
facility-based surveillance that focuses primarily on the diseases and conditions
considered to be of particular public health importance for their context. This
can also improve relationships and links between communities and their local
health system and health sectors to leverage community structures for better
surveillance, disease prevention and disease control.21-24
6.1 Health Extension Workers’ Knowledge

The HEWs’ and community structures’ knowledge of case definitions, surveillance
approaches, identification requirements for reportable events and reporting
mechanisms remains crucial for the early detection and reporting of reportable
disease conditions happening at the community level. To determine HEWs’
knowledge level, 13 questions on the basics of PHEM were asked. Based on the
t result, about 100 (70.92%) HEWs scored above the mean knowledge score
level, 8.16 of 13 questions. About 41 (29.1%) scored below 55, and only 11 (7.8%)
scored above 85%. This finding indicates that a significant proportion of HEWs
working at the community level lack basic knowledge of the mandatory level
and list of public health problems, as well as their case definitions, reporting
timelines, alert and action thresholds for response and other basic PHEM-
related concepts. This argument can be supplemented by the finding that
93 (65.9%) HEWs at the HP level knew the existence of diseases and event
conditions to be reported immediately.
The assessment result regarding the availability and functionality of structure of

the system found that 114 (78.7%) HEWs working at HPs, 49 (94%) catchment
HC-level respondents and 45 (90%) woreda-level experts, respectively, knew
the mandatory notifiable disease conditions in their locality. This finding clearly
page- 903
calls into question HEWs’ understanding of the number of reportable diseases

or events at different health system levels and the activities that were performed
to organize the system before its implementation with respect to the division
of responsibility for the uniformity problem. This may indicate that there are
public health problems that are not detected in their totality or are detected
but not reported since the responsible persons are not aware of their existence
or reportability status in their context. Based on WHO recommendations, the
availability and proper functionality of the surveillance system components’
national list of conditions and diseases for surveillance are required for uniform
understanding of the number and case definitions of diseases/event conditions
to be reported at all health system levels. A uniform understanding by all
involved personnel can be achieved by standardizing a list of diseases and
conditions mandated for reporting at all levels. This is crucial for ensuring
a functional early warning system, monitoring and a timely response for all
relevant PHEs at all health system levels and in all geographical areas.10
6.2 Supporting Function of the System at the

Community Level
This study considered the availability of standards and guidelines, trainings,

supervision, communication facilities and resources at the community level in
order to evaluate the level of the HEP’s supporting functions within the PHEM
system. Only 52 (73.2%) HPs had guidelines, protocols or packages to guide
surveillance activities at the community level. In addition, case definitions of
reportable diseases and event conditions were available at only 26 (36.6%) of
the visited HPs. On the other hand, only 22 (42.0%) HCs had a case definition
for the surveillance of targeted diseases. In addition, educational materials for
community members on nationally notifiable disease conditions were available
only in 70% of woredas. The knowledge problems seen with HEWs may also
be related to the availability of supporting documents and guidelines they
can use as a reference. HEWs’ levels of knowledge of community surveillance
and response also shows variability at different levels of the health system.
HEWs working at different woreda health system levels have different levels
of knowledge (at P < 0.001). In addition to the availability of PHEM-related
documents and guidance, gaps seen in HEWs’ knowledge level differences may
page-904
be attributable to differences in capacity-building activities in different woredas

and regions. This is reflected in the finding that 67 (47.5%) HEWs at the HP
level were not trained on community surveillance and only 39 (78%) woredas
were providing training for HEWs on the surveillance report notification process.
A higher level of awareness and engagement of the local community as well
as HEWs is a pre-requisite for effective community-level surveillance activities
and its sustainability.
The finding shows that level of awareness and engagement of HEWs on ongoing
PHEM-related activities is not sufficient as much as it is needed to ensure
community ownership and sustainability. Based on the finding, 88 (62.4%) of
community-level service providers (HEWs) do not have any prior information
or knowledge regarding emergency preparedness planning in their woreda.
Furthermore, only 111 (78.7%) of HEWs claimed that they participated in the
social mobilization, prevention, emergency preparedness, response and control
activities that took place in their locality.
6.3 Structure of the system at the community level
The presence of community networking and partnership, a coordination

platform, a mandatory notification mechanism and a surveillance strategy
at the community level were assessed to understand the adequacy of the
current system’s structure in addressing PHEM activities, as uniform community
engagement at all levels and geographical areas is vital. Based on the assessment
result, however, risk assessment was being carried out with the involvement
of HEWs and community volunteers in only 72% of woredas. In addition, 43
(86%) and 30 (60%) of woreda structures engaged community members,
social groups or associations, community networks and other governmental
and non-governmental stakeholders in public health surveillance and response-
related activities. The programs that have been shown to be the most effective
and sustainable have included partnership participation, empowerment
and ownership of the local communities to underpin their sustainability.6,25
Furthermore, national strategies and policies are insufficient by themselves to
prepare at-risk communities for disasters and emergencies, such as a pandemic.
They also require community ownership, participation and consultation, linkages
page- 905
with partners at all levels and operational planning—not just at the national
level but at the regional and woreda levels, again with the involvement of the
community. This level of performance can have negative effects on the early
identification and reporting of diseases or conditions at the community level.
Community-level engagement and partnerships with governmental and non-
governmental organizations are of paramount importance in managing gaps
in a timely way.
Community engagement for public health surveillance and response, starting

from the planning stage, is vital to increasing early-detection capacity, accurate
reporting and acceptance by key community leaders, villagers and team
members working in case detection and response.26,27 One study performed
in Sierra Leone showed that community event-based surveillance (CEBS) staff
identified 4 out of 6 cases with no epidemiologic links. These cases were identified
more rapidly by CEBS than those detected by the national indicator-based
surveillance system.28 Another study on community-based measles mortality
surveillance in 2 districts of Katanga Province in the Democratic Republic of
Congo revealed that a community-based network of volunteers recorded a
much larger number of deaths (376) than did HCs (27) during a measles
outbreak.17 Surveillance at the primary health service level is also significant for
tracking conditions like nutritional problems or maternal and perinatal deaths
at the community level. The evaluation results of implementation research on
community-based surveillance for nutritional screening showed that engaging
the local community, beginning with planning and optimum capacity-building
and monitoring activity, is far less costly than performing surveys.27
This study showed that, among the visited HPs, 59 (84.29%) had functional
community-based networks that were engaged and supportive of local
surveillance activities. This suggests that the absence of community engagement
in the remaining 15% of kebele structures leaves them without a functional
community structure and with no chance to include community-level reports
of public health problems. This may raise a representativeness issue regarding
the reports received from the remaining HPs that lack a functional community
structure. In addition, these structures lose the proven advantages given by
community engagement in community-level surveillance activities identified
by the studies mentioned above.26-28 This argument can be supplemented
by other findings. In this study, 100 (70.9%) HEWs claimed to be using a
mixed surveillance strategy (house-to-house visits and an HP register) to survey
disease conditions at the grassroots level.
page-906
Community-level public health problems are easily identified and reported by

well-structured and capable functional community networks. Different studies
have shown that community volunteers can identify diseases and conditions.
Information gathered from the community can be used to epidemiologically
monitor malnutrition. The community surveillance data revealing a 28.8%
prevalence rate of notifiable disease conditions is similar to other surveys
reporting an overall underweight prevalence rate of 27.2% for the same
population. To ensure the data’s continued reliability, HCs are recommended
to provide sufficient and ongoing training to community health workers. WHO
also clearly states that the functionality of community-level surveillance activities
depends on the availability of legislation for implementation, strategies for
implementation, clear roles for implementers and stakeholders and clear
relationships with each other and the various networks and partnerships. All
of these are very useful for planning, the uniform implementation of a routine
surveillance system and management.10 Our study result showed that uniformity
in the type of community-level network structures used for surveillance is poor.
WDA networks were found to be a prominent functional community structure;
they were mentioned by 52 (73.3%) HP respondents. Other community
structures mentioned by HC respondents as the functional structures used as
public health surveillance agents at their catchment HPs include clan leaders,
Agricultural Development Agents, kebele cabinet members in the community
and community key informants like religious leaders, teachers, traditional
healers, traditional birth attendants, militia and police.
It is recommended that the already available community structure be used for

their adaptability. The absence of uniformity in community-level surveillance
implementation at different places and health system levels may be due to the
absence of clear guidance and proper monitoring and evaluation activities.
This argument is supplemented by other findings that clearly show an absence
of clear demarcation of the roles and responsibilities of personnel working at
all levels of the health system. About 20 (38%) of health facility respondents
and 20 (40%) of woreda-level respondents perceived that community-based
surveillance activities are the primary responsibility of HEP supervisors and
coordinators at the health facility and woreda levels, respectively. In addition,
PHEM focal persons at 46 (88%) health facilities were responsible for
surveillance data compilation and data analysis, followed by HEW supervisors
(11; 21%) and facility disease-prevention process owners (5; 10%).
page- 907
In addition to the availability of community network structures, the motivation

status of the available networks is critical to their functionality. Based on
studies conducted in different countries, coordination status, team spirit and
knowledge of the type of work to be performed are crucial to the motivation
status of community health volunteers. The availability of guidance on the
roles and responsibilities of stakeholders, a list of diseases and conditions for
surveillance and a clear surveillance strategy can also improve volunteers’
motivation and significantly improve report flow from the affected community
to the volunteer and facilitate patient access to medical services. These factors
can also facilitate timely reporting to the nearest health institution, which can
point to the occurrence of outbreak of one or more of the diseases under
surveillance.29,30 In this regard, our study also shows a significant emergency
coordination gap at the lower levels of the health system, which can be attributed
to the absence of guidance material, clear guidance, and regular feedback and
support from higher-levels of the health system. Only 71 (50.3%) HPs, 38 (72%)
HCs and 40 (80%) woredas had an emergency coordination taskforce or RRT.
More than half of HPs in all regions—except for Tigray, Somali, the SNNPR
and Gambela—had no emergency coordination platform for an emergency
response at the kebele level. Among facility-level RRTs, only half of conducted
regular meetings, and only 28 (74%) engaged health extension supervisors as
members of the facility-level RRT.
6.4 Core functions of the surveillance system at the

com munity level
To evaluate the core functions of the PHEM system, we assessed the level of
HEP involvement in case detection, surveillance-data reporting, analysis and
interpretation for public health action, feedback mechanisms, engagement in
epidemic preparedness and response activities.
Around three fourths (39; 74%) of health facilities and 44 (88%) of woreda
structures monitored the activities of HEWs in their catchment either by on-site
supervision or by other means. Among them, 38 (72%) health facilities provided
regular feedback based on their supervision findings. During the assessment,
37 (52.11%) HPs claimed to receive supportive supervision and get feedback by
page-908
officials from the woreda and HC levels, but feedback was seen and verified
at only 15 (21.1%) of the HPs that claimed to receive supportive supervision
and get feedback during the last fiscal year. Quality data is a prerequisite for
planning appropriate interventions and providing a real-time response for a
significant reduction of public health outbreak-related morbidity and mortality.
Routine monitoring of the surveillance system by using supportive supervision
and providing feedback is one strategy to improve staff performance and the
quality of data for information generation. The present finding clearly indicates
that the routine monitoring of community-level health system structure is not
good enough to improve community health workers’ motivation. This argument
can be supported by the finding that a transparently low salary and position,
poor access to training, a heavy workload and exhaustive job description, a lack
of recognition, a lack of supervision, and poor communication and transport
were unsatisfactory and causes of the lower motivation levels of community
health workers. This reflects that 124 (87.94%) of the HEWs working at HPs
reported surveillance data to their catchment health facilities, so the national-
level surveillance data lack the reports of 13% of the HPs and raise a question
of representativeness.
Studies also showed that the provision of training in community-based

networks, using simple case definitions understandable by the community,
was identified as effective for community-based surveillance.29,31,32 Locally-
translated surveillance materials are more accessible and comprehensible,
which can potentially improve the quality of data reporting and the motivation
of community health workers. In a study on learning from local knowledge to
improve disease surveillance, perceptions regarding the Guinea worm illness
experience showed that, among 164 volunteers, only 2 submitted false reports
due to an incorrect disease definition. By contrast, local government health
workers conducting village searches during the same period were significantly
more likely to register false positive reports.31 As indicated in this study, locally-
translated versions of reporting forms were available only at 15 (21.1%) HPs.
In addition, 52 (73.2%) HPs had the translated version of the guidelines in
the local language, and locally-translated case definitions of the reportable
diseases and event conditions were available at only 15 (21.1%) of the HPs and
34 (68%) visited woredas. This may contribute to the lower engagement and
functioning status of community-level surveillance performance.
page- 909
The qualitative findings of this study also indicated many issues to be fixed
or addressed to improve the community-level surveillance system with respect
to early detection and reporting of the reportable disease conditions. Several
KII and FGD participants stated that the lack of basic refresher training
for community-level networks and HEWs and shortages of PHEM-related
documents and guidance were the reasons for the community-level PHEM
system’s failure to work as intended. The HEWs’ low knowledge level in Dire
Dawa, Somali and other regions in which the mean knowledge score was less
than 50% can negatively affect their motivation level and suggests the need
for uninterrupted capacity-building activities that can form a basis for effective
communication and sustainability. This finding was consistent with qualitative
research conducted to design sustainable community-based surveillance for
rabies in northern Australia and Papua New Guinea, which stated that effective
communication is vital for community-based surveillance; participants must
understand the reporting requirements and reports needing to be collected in
a timely manner for analysis.33
Moreover, different studies done in different countries also showed that a simple,
more direct and useful, less-extensive community-based surveillance system
with sufficient financing and trained personnel has improved performance and
data quality.29,31,32 Health workers performing surveillance activities at the
community level were affected negatively by low salary and position, poor
access to training, a heavy workload and exhaustive job description, a lack of
recognition, a lack of supervision and poor communication and transport.29
To sustain community-based programs, long-term mechanisms must be
established to ensure commitment at all levels and the continued availability
of resources after the initial funding has ceased.30 Consistent with these study
findings, our study suggests the existence of a significant gap regarding budget
allocation for community-level surveillance activities that are very determinant
for functionality of the system and motivation levels of surveillance staff and
respondents at different levels. Our findings show that only 9 (18%) woreda
structures and 6 (11%) health facilities had allocated a budget to support
surveillance activities at the community level. Majority (80%) of the institutions
that allocated budget did so for per diem and transportation allowance.
In addition to making available the required materials for registration and

reporting, appropriate and timely monitoring and evaluation play a role in
improving the motivation level of health workers and data quality. This can be
page-910
reflected in routine surveillance activities that can have the potential to determine
data-quality issues with respect to data reliability and representativeness. The
assessment results showed that 100 (70.9%) HPs had a means of verifying the
reported cases from the community. A majority of HEW respondents working
at the HP level in Tigray and Amhara said that they verified the reported cases
with house visits with health development leaders. In addition, 117 (82.9%) HPs
and 50 (70.42%) health facilities kept a copy of reports in their HP and health
facility, respectively. Only 16 (31%) catchment health facilities had a rumor
logbook for capturing any disease-surveillance-related rumors from various
information sources. Among them, only 14 (88%) facilities had registered
rumors in the registration book within the past 12 months, and 42 (79%) health
facilities monitored the quality of the surveillance data by measuring both its
timeliness and completeness. Of the facilities, 41 of 43 (93%) monitored only
the completeness of the data that they received from the HEWs, and 3 of 41
(7%) monitored only the timeliness of the report.
In general, the findings of the assessment at different levels of the health

system clearly indicate variability in the mean score of both the structure of
the system and its supporting functions at all levels. This variability suggests a
significant difference in support for primary-level surveillance activities, but the
core function of the system shows no significant difference among health system
levels. Among the different components of primary-level health care surveillance,
the supporting function and core function have a direct positive relationship.
The supporting functions of the primary-level health care surveillance system
are predictors of the core function of surveillance activities at the community
level in Ethiopia.
page- 911
6.5 Limitations of the study
This study primarily addressed the system structure, core functions, supporting
functions and data-quality components related to the PHEM system implemented
by the HEP. The limited number of included woredas, HPs, health centers, and
HEWs from each region can be considered a limitation of this assessment that
might mean it is not fully representative of the national status. The lack of
well-documented secondary surveillance data and reports of the PHC system
available to use in evaluating the level of contribution and effectiveness in the
detection of epidemics, reported disease conditions and response activities by
the HEP is also a limitation.
page-912
7. Conclusion and Recommendation
7.1 Conclusion
Although the national HEP is not contributing greatly to addressing basic

components of PHEM as per the standard, this study has shown that it plays a
significant role and has great potential in detecting, reporting and responding
to PHEs at the community level. The detection and reporting of the disease
conditions under surveillance, community-level health education and health-
promotion activities, disease prevention and control activities by the HEWs and
the available community networks have contributed significantly to kebele-level
PHE preparedness, prevention and management of public health conditions.
A significant proportion of HEWs working at the community level lacked

basic knowledge of the mandatory level and list of public health problems,
their case definitions, reporting timelines and alert and action thresholds for
response and other PHEM-related basic concepts. Gaps were observed in
community engagement and partnership at lower levels of the health system.
Almost half of HEWs working at HPs had no knowledge of the emergency
preparedness activities happening in their woredas. In addition, around 1 in
4 HEWs had not participated in social mobilization, prevention activities,
emergency preparedness or response and control activities that have taken
place in their localities. Regarding partnership, only 3 of 5 woreda structures
engaged different governmental and non-governmental stakeholders in public
health surveillance, preparedness and response-related activities.
Regarding the budget allocation for community-level surveillance activities,

this assessment revealed the existence of a significant gap in the budget
allocation for community-level surveillance activities. Around 1 in 4 woredas
assign a budget for emergency-related activities, which is very determinant for
the functionality of the system and motivation level of surveillance staff and
respondents at different level.
page- 913
In addition to the gaps in emergency coordination identified at the lower

level of the health system, the absence of guidance material, clear roles and
responsibilities in community-level surveillance, monitoring and evaluation
activities and regular feedback and support from higher-level health systems
is sub-optimal. Moreover, locally-translated surveillance materials, which are
more accessible and comprehensible, were available in only 1 in 4 HPs.
In addition, there is variability in the structure of the system and its supporting
functions at all levels. This variability at all levels shows significant variation
in the support for primary-level surveillance activities in different HPs. The
supporting function and core function of the surveillance system have a direct
positive relationship. The supporting functions of the primary-level health care
surveillance system are predictors of the core function of surveillance activities
at the community level in Ethiopia.
In general, the common challenges for lower-performing community-level

surveillance activities throughout the nation are identified as the limited
knowledge of HEWs and community networks about community-level
surveillance activities, the fragmented coordination of community-level networks
(structures), the lack of and limited access to locally-translated guidelines, case
definitions and reporting tools, sub-optimal monitoring and follow-up, limited
capacity-building and engagement of the community health system structure,
and limited funding for community-level surveillance activity.
7.2 Recommendations
Based on the findings of this study, the following points are recommended at
different levels of the health system to improve the HEP’s role and contribution
in addressing the major components of surveillance activities at the community
level:
National and regional levels

• Standardize national-level and community-level surveillance
implementation by designing uniform implementation guidance,
standards and strategies needs to be given due attention.
page-914
• Integrate basic PHEM components in the HEP packages and training

curriculum needs to emphasize their sustainability.
• Consider engaging the national- and regional-level HEP in the
vulnerability risk assessment and mapping and emergency-
preparedness planning.
• Emphasize the allocation of a proportional and sufficient budget
for community-level surveillance and emergency preparedness and
response activities.
• Refocus on the preparation and distribution of locally-translated
materials, guidance, roles and responsibilities, case definitions
and documents to support community-level health education and
promotion.
• In view of advancing the work, undertake routine monitoring and
evaluation of community-level surveillance activities at all health
system levels.
• Standardize guidance, orientation, training and educational materials
for community-level networking and operationalization. This work is
essential.
Zone, woreda and health facility levels

• Enhance provision of continuous capacity-building (training and
resources) for community-level surveillance workers and community
networks, like distribution and follow-up for reporting tools, guidance
materials and standards at the lower level. This work is essential.
• Reconsider allocation of a proportional and sufficient budget for
community-level surveillance, emergency preparedness and response
activities.
• Improve the routine monitoring and evaluation and feedback activities
for community-level surveillance activities.
• Pay due attention to routine evaluation and filling gaps on community-
level network availability and functionality.
• Ensure the integration of PHEM components in the HEP packages
and integrate the basics of PHEM into the HEP training curriculum
to help foster their impact.
page- 915
• Consider engaging HEP coordinators and the community structure

in emergency-preparedness planning, assessment of the vulnerable
conditions at the community (district) level and building of a resilient
health care system.
• Ensure the availability and functionality of a kebele-level emergency
coordination platform to enhance the engagement of stakeholders.
• Reserve finances, emergency drugs and logistics at all levels for a
timely emergency response.
Non-governmental organizations working in surveillance-related activities

at all levels
• Provide technical and financial support to strengthen community-
level surveillance-related activities at all levels and ensure meaningful
contribution.
• Fill the gaps seen in emergency coordination and partnership activities
at lower health-system levels to help improve the situation.
page-916
References
1. D. M. Bowley, E. J. Dickson, N. Tai, J. Goosen, K. D. Boffard. A theme

issue by, for, and about Africa. Bmj. 2005;331:780-781.
2. WHO. Declaration of Alma-Ata. Paper presented at: International
Conference on Primary Health Care1978; Alma-Ata, USSR.
3. FMoH. Health Sector Transformation Plan (HSTP) 2014/15 - 2019/20.
2015.
Sector Development Program IV (2010/11 - 2014/15). 2010.
5. Wang H, Tesfaye R, Ramana GNV, Chekagn CT. Ethiopia Health
Extension Program: An Institutionalized Community Approach for
Universal Health Coverage. 2016.
6. Banteyerga H. Ethiopia’s health extension program: improving health
through community involvement. MEDICC review. 2011;13(3):46-49.
7. FMOH. Guideline for Implementation of Health Extension Program.
Addis Ababa, Ethiopia 2008.
8. FMOH, EPHI. Public Health Emergency Management Guidline for
Ethiopia. 2012.
9. The Open University. Communicable Diseases Module: 41. Integrated
Disease Surveillance and Response. 2020.
10. WHO. Communicable disease surveillance and response systems -
Guide to monitoring and evaluating. 2006.
11. O. Fenollar, F. Mediannikov. Emerging infectious diseases in Africa in
the 21th century. New Microbes and New Infections. Elsevier. 2018;28.
12. Emerging and re-emerging infectious diseases as public health challenges of
the 21st century. Ethiop J Heal Dev. 2015;29(2).
page- 917
13. D. E. Mackey, T. K. L, B. A. C, R. H, R.,, Brouwer K, C. Lee.

Emerging and Reemerging Neglected Tropical Diseases: a Review
of Key Characteristics, Risk Factors, and the Policy and Innovation
Environment. Clin Microbiol Rev. 2014;27(4):949–979.
14. Lilay A, Asamene N, Bekele A, et al. Reemergence of yellow fever
in Ethiopia after 50 years, 2013: epidemiological and entomological
investigations. BMC infectious diseases. 2017;17(1):1-6.
15. Institute for Health Metrics and Evaluation. Global Burden of Disease
(GBD) Country Profile for Ethiopia. http://www.healthdata.org/
ethiopia. Published 2017. Accessed 20 March 2019.
16. P. Shears. Emerging and reemerging infections in Africa: The need for
improved laboratory services and disease surveillance. Microbes and
Infection. 2000.
17. N’Goran A A, Ilunga N, Coldiron ME, Grais RF, Porten K. Community-
based measles mortality surveillance in two districts of Katanga
Province, Democratic Republic of Congo. BMC research notes.
2013;6:537.
18. CSA. Population Projections for Ethiopia 2007 - 2037. 2013.
Sector Development Program II: 2002/03 – 2004/05. 2002.
Sector Development Program III - 2005/6 - 2009/10. 2005.
21. S. Oum DC, S. Cairncross. Community-based surveillance: A pilot study
from rural Cambodia. Trop Med Int Heal. 2005;10(7):689–697.
22. JICA Amrids Project. Lessons learned from the community surveillance
activities. December, 2014.
23. D. A. Larsen ea. Malaria surveillance in low-transmission areas of
Zambia using reactive case detection. Malar J. 2015;14(1):1-9.
24. WHO Regional Office for Africa. A GUIDE FOR ESTABLISHING
COMMUNITY BASED SURVEILLANCE August, 2014.
page-918
25. P. Wendy, M. Vincent, B. Heather. Community - Based Pandemic
Preparedness Multi‐sectorial action for safer, healthier and more
resilient communities. 2001.
26. Crowe S, Hertz D, Maenner M, et al. A plan for community event-
based surveillance to reduce Ebola transmission - Sierra Leone, 2014-
2015. MMWR Morbidity and mortality weekly report. 2015;64(3):70-73.
27. Ahmed AEAM, Ahmed IAM. Nutrition surveillance in the Sudan: a
community-based approach. 1996.
28. R. Ratnayake ea. Assessment of community event-based
surveillance for Ebola virus disease, Sierra Leone. Emerg Infect Dis.
2016;22(8):1431–1437.
29. Kok M, Muula S. Motivation and job satisfaction of health surveillance
assistants in Mwanza, Malawi: an explorative study. Malawi Medical
Journal. 2013;25(1):5-11.
30. E.A.S.F.D.N. Coates. CORE Group Polio Project Final Evaluation. 2012.
31. Brieger WR, Kendall C. Learning from local knowledge to improve
disease surveillance: perceptions of the guinea worm illness experience.
Health education research. 1992;7(4):471-485.
32. Cerón A, Ortiz MR, Álvarez D, Palmer GH, Cordón-Rosales C.
Local disease concepts relevant to the design of a community-based
surveillance program for influenza in rural Guatemala. International
journal for equity in health. 2016;15:69.
33. V. J. Brookes, E. Kennedy, P. Dhagapan, M. P. Ward. Qualitative
Research to Design Sustainable Community-Based Surveillance for
Rabies in Northern Australia and Papua New Guinea. Front Vet Sci.
2017;4:1-9.
page- 919
page-920
Specific Study 4:
Cost-Effectiveness Analysis of the
page- 921
CONTENTS
Contents ------------------------------------------------------------------------922
Executive Summary -------------------------------------------------------------926
1 INTRODUCTION ----------------------------------------------------929
1.1 Background -------------------------------------------929
1.2 Statement of the Problem ---------------------------------------932
2 LITERATURE REVIEW --------------------------------------------935
3 OBJECTIVE --------------------------------------------------------937
3.1 General objective --------------------------------------------------937
3.2 Specific objectives ---------------------------------------------------937
4 METHODS ----------------------------------------------------------939
4.1 Overview of the costing methodology -----------------------------939
4.1.1 Costing approaches --------------------------------------939
4.1.2 Cost analysis from providers’ perspective ----------940
4.1.3 Cost summary measure from providers’ perspective --943
4.1.4 Client perspective costing analysis ---------------943
4.1.5 Community Contribution to the HEP ----------------945
4.1.6 Cost analysis from the societal perspective ---------945
4.1.7 Cost savings analysis ------------------------------945
page-922
4.2 Effectiveness measure -----------------------------------------947
4.3 Cost-effectiveness analysis ---------------------------------------952
4.4 Sensitivity analysis ----------------------------------------------952
5 RESULTS -------------------------------------------------------------953
5.1 Cost Analysis of the HEP from Providers’ Perspective --953
5.1.1 Unit cost by components --------------------------------953
5.1.2 Unit costs of selected HEP services by

delivery modality -------------------------------------955
5.1.3 Unit cost disaggregated by cost center

and HEP component ---------------------------------------957
5.2 Cost of providing healthcare service at HCs and HPs --958
5.3 Client perspective cost and cost saving analysis ------------959
5.4 Cost analysis from the societal perspective -----------------959
5.5 Cost-effectiveness analysis --------------------------------962
5.6 Sensitivity analysis -------------------------------------------963
6 DISCUSSION ------------------------------------967
7 STRENGTHS AND LIMITATIONS --------------------972
8 CONCLUSION ------------------------------------------------------973
9 RECOMMENDATIONS -------------------------------975
REFERENCES -------------------------------------976
ANNEX 1: COST OF SERVICES AT HEALTH CENTERS --------981
page- 923
ANNEX 2: COST OF SERVICES AT HEALTH POSTS ---------------984
ANNEX 3: ONE-WAY SENSITIVITY ANALYSIS OF ALL

INTERVENTIONS ---------------------------------------986
LIST OF TABLES
Table 1. HEP service packages: family health services, disease

prevention and control, and hygiene and sanitation -------------------931
Table 2. Description of HEP input costs and sources of data ----------946
Table 3. Coverage inputs for effectiveness measures -----------------------950

Table 4: Unit costs (US$) of selected HEP interventions, by
component, in Ethiopia, 2018. ------------------------------------------954
Table 5: Unit cost of selected HEP interventions by service

delivery modality in Ethiopia. -------------------------------------956
Table 6: Unit cost of providing selected interventions at

Health Centers and Health Posts. ---------------------------------959
Table 7: Total (direct medical, non-medical, and productivity loss) cost per
client and cost savings of selected HEP interventions from client
perspective, Ethiopia, 2018. ----------------------------------961
Table 8: Unit cost of selected HEP intervention from societal

perspective in Ethiopia, 2018. --------------------------------------------962
Table 9: Cost-effectiveness of selected Health Extension

Program interventions in Ethiopia, 2005/10-2018. --------------------------964
page-924
LIST OF FIGURES
Figure 1: Percentage distribution of family health services

unit costs, by component ----------------------------------------------957
Figure 2. Percentage distribution of DPC unit costs, by component --958
Figure 3. One-way sensitivity analysis showing an incremental

cost-effectiveness ratio of pneumonia treatment over a
range of key parameters -----------------------------------------------------------966
page- 925
Executive Summary
Background: Ethiopia launched the Health Extension Program (HEP) in 2004

with a view to achieving universal health coverage of primary healthcare
(PHC) among the rural and underserved population. It was intended to
provide equitable access to healthcare at the community level through Health
Extension workers (HEWs) deployed in each kebele. The HEP has attained
remarkable achievements in family health service outcomes. Despite being a
flagship program, there is limited evidence available on the cost-effectiveness
of the HEP to inform policy and program-management decisions. To help fill
this gap, this study assesses the cost from a societal perspective, the cost-
effectiveness of selected HEP interventions from a provider perspective, and
the savings from the client perspective.
Methods: The costs of providing the HEP service packages were analyzed
using mixed methods (bottom-up and top-down). From the HEP program 21
interventions were chosen, covering the hygiene and sanitation, maternal and
child health, and disease prevention and control packages. The cost components
include personnel, medicine, supplies, infrastructure, capacity-building, and
equipment. The study reports all costs in 2018 United States dollars (US$, using
an exchange rate of US$1=27.6677 Ethiopian birr [ETB]) and uses a 1-year
timeframe. The final cost summary measure is cost per person per service used.
The measure of effectiveness is life years gained (LYG), calculated from the
standpoint of the HEP’s impact on the coverage of the health system. A full
economic evaluation was applied to address the cost-effectiveness of introducing
the HEP into the existing health system from the provider perspective. The cost-
effectiveness of the interventions are reported in cost per LYG, compared with
Ethiopia’s per capita gross domestic product (GDP), in order to determine the
cost-effectiveness of the interventions.
Results: The average unit cost of providing hygiene and environmental sanitation,
family health services, and disease prevention and control packages through
the HEP were $0.70, $4.90, and $7.40, respectively. Across these different
service packages, the cost of drugs and supplies account for 53% to 68% of the
total cost and are the major cost driver. The cost of HEP interventions ranges
page-926
between $0.60 (for an improved water source) and $43.10 (for tuberculosis
[TB] treatment). Each year, the provision of the HEP would be able to save
$2.30 per capita for a client seeking healthcare at a Health Post (HP). The
average annual cost of delivering the selected 21 healthcare interventions
through the HEP at the HP level was $9 897. The incremental cost-effectiveness
ratio (ICER) of all the selected HEP interventions lies within 100% of the GDP
per capita per LYG, which are considered very cost-effective. Overall, the HEP
is found to be very cost-effective: for an investment of an additional $77.40, an
additional life year is gained.
Conclusion: The HEP is assessed as a very cost-effective way of delivering

services to the community. Besides improving health outcomes at a lower cost
to the provider, its implementation saves an enormous amount of money for
clients each year. This evidence should be helpful in guiding policy-makers
in financial planning, decision-making, and allocating resources to the HEP.
Expanding access to essential health services should maximize the use of health
posts by at the same time ensuring that quality of care is not compromised.
page- 927
page-928
1 Introduction
1.1 Background
Ethiopia has a 3-tier healthcare delivery system comprising primary, secondary,

and tertiary healthcare units. The primary healthcare unit (PHCU) is the first
level of the tier system: it comprises primary hospitals, Health Centers (HCs),
and their satellite Health Posts (HPs), which are connected by a referral system.
It serves about 100 000 people. HPs are the most peripheral unit of healthcare
delivery, providing mainly preventive care, with some curative services for
selected illnesses.
The HEP is one of the most innovative government-led, community-based

health programs in Ethiopia. It was adopted in Ethiopia in 2004 with a view
to achieving universal health coverage for PHC among the rural population by
providing equitable access to healthcare services in each kebele (i.e., community;
a kebele is the lowest administrative unit in Ethiopia) through Health Extension
Workers (HEWs), high-school graduates who receive 1 year of training and are
deployed with a salary to each sub-district or kebele to implement the Health
Extension packages at the HP.1,2 There was a steady increase in the number of
HPs constructed over the past few years, from 6 191 in 2013 to 17 187 in 2017.3
In order to implement the packages, the HEP serves as a primary vehicle
for health promotion, disease prevention, behavioral change communication,
and basic curative services through effective implementation of the defined 16
Health Extension packages.
Ethiopia has made impressive progress in health outcomes, including substantial

reductions in maternal, neonatal, infant, and child mortality. Among these gains
is a reduction in under-5 mortality of 67% from the 1990 estimate, which led to
an increase in average life expectancy at birth from 45 to 64 by 2014. During
the same period, a 69% decline in maternal mortality was reported from a
high estimate base of 1 400 per 100 000 live birth2 The HEP’s contribution to
decreasing morbidity and mortality related to pneumonia, diarrhea, and malaria,
as well as improving the use of family planning (FP), has been substantial: it
page- 929
Cost-Effectiveness Analysis of the Health Extension Program
has increased the coverage of fully immunized children (>80%), increased the
use of insecticide-treated bed nets (ITNs, 60%) and minimized unmet FP needs
(25%).2,4 This success is partly attributable to the extensive investment in the
HEP and would otherwise not have been attainable. As a result, Ethiopia has
had remarkable success in meeting most of the Millennium Development Goals
(MDGs). HEWs provide the Health Extension packages through 3 modalities:
static service at the HP, outreach, and home-to-home visits.5 The current HEP
interventions encompass 16 packages. Based on the availability of data related
to cost-effectiveness measures and the identification of the interventions that
are significantly shared among HEWs’ roles and responsibilities, the following
interventions were selected for this study (Table 1).
page-930
Table 1. HEP service packages: family health services, disease prevention

ancontrol, and hygiene and sanitation
1 Family health services: preventive

1.1 Maternal and child health
Antenatal care
Tetanus-toxoid vaccination
Iron folate supplementation
Family planning – oral contraceptive pills, condoms, injectables,
and implants
1.2 Expanded program of immunization
Pentavalent vaccine
Measles vaccine
Pneumococcal vaccine
2 Disease prevention and control (curative services)
2.1 HIV testing and counselling
2.2 Tuberculosis prevention and control
Directly observed treatment short course therapy
2.3 Malaria prevention and control
Insecticide-treated bed nets
Indoor residual spray
Malaria treatment
2.4 Diarrheal disease management (oral rehydration salt and zinc)
2.5 Pneumonia treatment (cotrimoxazole, amoxicillin, gentamicin)
3 Hygiene and environmental sanitation
3.1 Improved water source
3.2 Handwashing with soap
3.3 Hygienic disposal of children’s feces
3.4 Latrine use
page- 931
1.2 Statement of the Problem
The initiation of the HEP in Ethiopia has improved the health outcome of
communities through delivering preventive, promotive, and curative services.4,6
The per-capita health expenditure on essential health services has grown
significantly over the past 2 decades, from $4.50 in 1995/96 to $28.65 in 2013/14.7
Since resources are scarce, however, countries need to focus on investing in
targeted, high-quality interventions with the highest impact and finding ways
to achieve the same results at a lower cost. Program decisions can be informed
through unit-cost and cost-effectiveness analysis estimates of interventions in
the HEP to help decision-makers determine priorities in resource allocation. The
unit-cost and cost-effectiveness analysis estimates provide a message about
how to improve health outcomes by using scarce healthcare resources in a
way that produces a high value for each dollar spent.8,9 Thus far, few cost and
cost-effectiveness studies related to the HEP have been conducted, despite
the program’s having been in place since 2004. Therefore, this study provides
estimates of the cost and cost-effectiveness of the HEP. The outcome of the
research is expected to help program managers, policy-makers, and healthcare
managers make informed decisions about the HEP in order to help improve
rural healthcare.
page-932
page- 933
page-934
2 Literature review
A few studies carried out in Ethiopia, as well as other countries, provide an

overview of the cost-effectiveness of certain interventions linked to community
or volunteer health workers. One study, conducted from the perspective of
the health systems in Ethiopia, Indonesia, and Kenya to evaluate the cost-
effectiveness of community-based practitioner programs (in pregnancy, child
health, breastfeeding, and postnatal care), found that the services were cost-
effective, using the country’s willingness to pay the threshold as a reference. The
incremental cost per LYG (ICER) was $999 in Shebedino, Ethiopia; $3 396 in
Southwest Sumba, Indonesia; $2 470 in Takala, Indonesia; and $82 in Kasarami,
Kenya.10
A costing study conducted in Ethiopia showed that the HEW fees for the full
cost recovery of the provision of services ranged from 55.1 ETB to 209.1 ETB
per encounter in urban areas and from 19.6 ETB to 219.4 ETB in rural areas.11
In a community-randomized trial (CRT) in Southern Ethiopia, the societal cost
per successfully treated smear-positive tuberculosis (TB) cases through HEW
was $60. Program costs were by far the largest share of the cost (82%), and the
rest represented out-of-pocket payments by patients (18%).12 A costing study
done in Ghana on the cost of delivering healthcare services in selected primary
health facilities (community-based Health Planning [CHP]) and services at
HCs estimated the average cost of delivering services at $10 923 and $44 638,
respectively. Personnel costs were the largest cost share in both facilities.
A cost-effectiveness analysis (CEA) done in the Mareko woreda showed that

implementing 22 healthcare interventions at a cost of <100 ETB per DALY
gained at the health-post level would avert 52% of the burden of disease,
whereas implementing 17 interventions at the hospital level and 18 interventions
at the HC level would avert about 22-34% of the burden of disease.
A cost-effectiveness analysis to assess the costs of a community health worker

(CHW) program in Mozambique over the years 2010-2012 found that the total
costs of the CHW program increased from $1.34 million in 2010 to $1.67 million
in 2012. In addition, the highest incremental cost-effectiveness ratio was for the
cost per beneficiary covered (including CHW salaries), which was estimated at
page- 935
$47.12, while the smallest incremental cost-effectiveness ratio was for the cost
per household visit (not including CHW salaries) was estimated at $0.09. On
the other hand, the study showed that increasing the CHWs’ salaries would not
only have increased total program costs but would also have led to the largest
efficiency gains in program implementation by 56% in cost per output over the
long run, after including CHW salaries.
Another cost-effectiveness study comparing the provision of antenatal care

(ANC) services by 2 auxiliary nurse-midwives and 1 auxiliary nurse-midwife in
sub-centers (i.e., with the same structure and training as the HEP) indicated
that the introduction of 1 auxiliary nurse-midwife at the sub-center level would
be very cost-effective. The same report found the annual cost of delivering
preventive, curative, and promotional services by CHWs to be $19 381.
There are very few economic evaluations available on community-based

practitioners in low- and middle-income countries, including Ethiopia.13 Among
the few studies carried out in Ethiopia were those on the new initiatives of
the Urban HEP (such as the Family Health Team [FHT] at the piloted health
facilities in Addis Ababa) and a cost-effectiveness analysis of 3 community-
based programs: HEWs in Ethiopia, village midwives in Indonesia, and
community HEWs in Kenya.10,13
Neither of these studies comprehensively evaluated the cost or cost-effectiveness

of healthcare services provided through the HEP in Ethiopia or considered the
different perspectives (i.e., societal, healthcare provider and client perspectives),
service delivery modalities, or broader HEP interventions. To fill this gap, a
costing and cost-effectiveness analysis of HEP interventions was conducted
under the broader perspective of economic evaluation.
page-936
3 Objective
3.1 General objective
The overall objective of this study is to assess the cost and cost-effectiveness of
delivering HEP interventions in order to assist decision-makers and program
managers in identifying interventions representing the best value for money
and allocating the scarce resources of the HEP.
3.2 Specific objectives
The specific objectives include the following:

• To estimate the provider and societal costs of HEP interventions;
• To analyze the potential cost difference due to HEP interventions for the
client; and
• To assess the cost-effectiveness of the HEP from the provider’s perspective.
page- 937
page-938
4 Methods
4.1 Overview of the costing methodology
4.1.1 Costing approaches
This study uses both bottom-up and top-down costing approaches to estimate
the cost of providing the HEP service packages. The bottom-up costing approach
is a costing method in which actions to be taken as part of an intervention are
listed, the specific resources needed to implement the intervention are described,
and prices are assigned to all the resources based on the opportunity costs
used for the intervention. The top-down costing approach considers the overall
expenditure at a central level in order to allocate costs to each intervention.14
The bottom-up approach was done by identifying the relevant cost center
(human resources, supplies and equipment, building, and capacity-building or
trainings) for the selected interventions. The top-down costing techniques were
applied for the different activities that were performed at the federal (central)
level, but some activities are essential to strengthening the provision of the HEP
at the woreda and grassroots levels. These include pre-service training, in-service
trainings (integrated refreshment trainings [IRTs]), supportive supervision, and
review meetings.
Primary and secondary data sources were used for costing, which was conducted
from 3 perspectives: healthcare provider, client, and societal. The healthcare
provider costing focuses on the resources depleted for the provision of the HEP
through HEWs from the government perspective, while the client perspective
takes into account the out-of-pocket spending attributed to care-seeking at
an HP (family health service or disease prevention and control interventions).
This spending includes direct medical, direct non-medical, and productivity
losses. Finally, to gain a much broader societal perspective, the community
contribution related to the healthcare system was extracted from the recent
national health account survey.7 Financial reports, cash receipts, procurement
units, market prices of commodities, and various data sources were reviewed
page- 939
and triangulated to accurately measure and value the resources used. All
resources were identified, measured, and valued. The detail costing for the
respective perspectives is explained in sub-sections 4.1.2, 4.1.3, and 4.1.6.
4.1.2 Cost analysis from providers’ perspective
COST ANALYSIS OF THE HEP
To conduct the costing from the provider’s perspective, both primary and
secondary data sources were used. The cost data were collected from HP
records, expert opinions, and interviews with HEWs from June 7, 2019 to July 1,
2019. The data were gathered through a cross-sectional health-facility-based
design that contacted more than 300 HPs, 54 HCs, and 57 Woreda Health
Offices (WorHOs) through standard questionnaires. In addition, secondary
data sources were used to collect the unit costs of items. The cost of each
service was annualized and allocated based on the overall and specific service
coverage each year. All shared costs were identified and distributed to the
services under study based on the annual number of clients served by HEWs.
The total annual number of clients served through the HEP was calculated by
multiplying the number of the population expected to be served by the program
(e.g., 5 000 in a village) by the proportion of the population seeking care at
an HP (0.8).15 This product was then multiplied by Ethiopian Demographic and
Health Survey (EDHS) coverage information and the percentage of the target
population to arrive at the number receiving each intervention. The cost of
the unshared supplies was allocated to their respective interventions. All costs
related to healthcare service delivery at HPs through HEWs were expressed in
Ethiopian birr (ETB) and converted to US dollars in 2018, with an exchange
rate of $1=27.6677 ETB.
The cost data include personnel, drugs/medicine, supplies/consumables,

building, capacity-building (i.e., start-up, pre-service and in-service trainings,
and monitoring and evaluation), and equipment. The costs were then classified
as: supplies, equipment and furniture, personnel, capacity-building, or building.
page-940
A. Supply costs
Supplies include medicines, consumables, stationery, and other utilities. The

costs included in this category are cost of the item, the cost of pharmaceutical
supply management, and wastage for vaccines. The total supply cost was then
estimated by multiplying the average unit cost per client by the number of
follow-up visits (if any) related to receiving the selected health service.
B. Personnel costs
We obtained the proportion of HEWs at each career level from the Ministry of
Health (MoH) human resources department in 2019.16 We collected the gross
salary of HEWs by level from primary data and calculated the average salary
of an HEW per minute using 22 working days in a month. Then, through an
HEW time-motion study, we estimated the average working time HEWs spent
for each type of service and multiplied this by the average HEW salary per
minute and number of follow-up visits (if any) to arrive at the personnel cost
per client .
C. Capital costs (Building, Equipment and furniture costs)
To estimate the annual building cost, the total cost by type of HP construction
(hollow block, wood, or stone) was collected retrospectively between 2004 and
2010 Ethiopian fiscal year (EFY) through interviews with the WorHO. The
unit cost of equipment and furniture for each intervention was obtained from
the pharmaceutical fund and supplies agency. All costs were adjusted to their
2018 value using Ethiopia’s consumer price index. The capital cost items with
a useful lifespan of more than 1 year (30 years for hollow block, 20 years for
stone (assumed) 15 years for wood or mud, and 5 years for all equipment and
furniture) were annuitized based on the discount rate of 3% 17. The weighted
average building costs of the 3 types of building (hollow block, wood or mud,
and stone) were calculated by multiplying the proportion of each type of HP
page- 941
construction types by their respective annuitized cost. These costs were then
added to estimate the annual unit cost of building. This annuitized cost was
then allocated to each service based on the total clients served per year to get
the unit cost of the building, equipment, and furniture per single visit, which was
then multiplied by the number of follow-up visits (if any) to compute the unit
cost per client. The capital cost was annuitized using
r
E= K*
1-(1+r)-n
where E is the annual equivalent costs,
K is the current price of a capital good,
n is the expected lifespan of the capital good, and
r is the discount rate.
D. Pre-service (start-up) and in-service training (IRT)
The pre-service and in-service training costs per HEW were collected from the
MoH and RHBs through interviews and document review. The costs of pre-
service and in-service training were annuitized with a useful life years of service
of 10 and 5 years, respectively. The annuitized costs were allocated to each
service based on the coverage of each intervention per year.
COST ANALYSIS OF HEALTHCARE SERVICES AT HCS
To estimate the cost implication of the addition of the HEP, we compared the
unit cost of providing a health service through HEWs with the unit cost of
providing the client with the same service at the HC level. Both primary and
secondary data were used to estimate the unit costs of providing tetanus-toxoid
(TT) vaccination, pentavalent vaccination, and iron folic acid supplementation
for pregnant women, diarrheal disease management, and antibiotics for
pneumonia treatment. The inputs included to estimate the cost were building,
equipment, supplies, personnel time, and review meeting and supervision costs.
In addition, 10% of the unit costs were added for administrative costs.10
In addition, an extensive search of the scientific literature was performed to

obtain the cost of other interventions. The databases reviewed were PubMed,
page-942
Elsevier, the website of the peer-reviewed journal Health Policy & Planning, and
the Tufts cost-effectiveness analysis registry. The keywords used for searching
the literature were: unit cost, cost, Ethiopia, Health Center, primary healthcare,
exempted services, family planning, directly observed therapy, tuberculosis
treatment, pneumonia, diarrhea, malaria, immunization, vaccination, measles,
pentavalent, antenatal care, tetanus-toxoid, iron folate, east Africa, and
developing countries. The main criteria for inclusion were that all the studies
must have been performed in Ethiopia and on the HC level. All reported costs
were first converted from US dollars into ETB, then adjusted to the 2018 price
year using a consumer price index. Inputs missing from the estimated costs were
addressed by gathering additional primary data at the HC level, analyzing
them, and adding them to the unit cost of the literature reviewed to fill the gap.
4.1.3 Cost summary measure from providers’ perspective
All unit costs of the cost centers of the respective interventions were added to
generate the total unit cost. The final cost summary measure is cost per person
per service used. Similarly, for FP services, the cost per couple year of protection
(CYP) was also estimated. Then, to estimate the total cost of selected HEP
interventions, the unit cost (i.e., cost per person) was multiplied by the number
of clients served at the HP for that particular intervention.
4.1.4 Client perspective costing analysis
COSTING APPROACH
The cost to the client when seeking a healthcare service was estimated by
considering 3 basic sub-components: direct medical, direct non-medical,
and productivity losses (i.e., opportunity costs) when seeking the service. An
exploratory review of the scientific literature was conducted to obtain the
indirect costs related to the clients when seeking healthcare services. These
costs were compared between those healthcare services provided at HCs and
those at HPs.
page- 943
The client-side costs at the HC and HP were collected through a review of

different literature from secondary sources of scientific literature. Direct
medical cost mainly includes the consultation or registration fee, laboratory or
investigation fee, and drug cost. The keywords used in searching the literature
were: out of pocket, expenditure, direct medical cost, drug cost, consultation
fee, laboratory/investigation fee, cost in Ethiopia, client cost in lower income
countries, and direct medical cost at Health Centers in sub-Saharan Africa.
In addition, we searched for antenatal care, family planning, immunization
pentavalent and measles, tetanus-toxoid vaccination for women, HIV testing
and counseling, TB treatment, under-5 pneumonia treatment, diarrheal case
management, and malaria treatment over the years 2007-2018.
All client costs collected were adjusted to the 2018 costing base year. For the
exempted family health services (e.g., immunizations or FP) and those for which
we were unable to find the real client cost from the literature, we obtained an
expert opinion based on the healthcare financing strategy and assumed their
cost to be zero.
Direct medical costs take into account the use of resources directly related
to the treatment at an HP or HC.17 Examples of direct medical costs include
outpatient fees, inpatient costs, and drug costs.19 Direct non-medical costs
consider the cost of non-medical supplies or services, which are not typically
linked to diagnosis or treatment. These include items like the transportation,
food, and accommodation costs incurred by clients or other parties.20
Furthermore, productivity losses, which are commonly referred to as economic

productivity losses for the clients and caregivers as a result of their accessing
health services or the best opportunity forgone because of the disease were
considered. The productivity loss considered in this study was the time spent
traveling by foot and transport and the time spent in the HC waiting for and
receiving the service. To calculate the productivity losses, daily income from
the United Nations’ Food and Agriculture Organization’s (FAO) report on the
economic productivity of smallholder household farmers’ data was considered.21
The detail description of the reviewed literature across the identified interventions
is presented in Annex 2.
page-944
4.1.5 Community Contribution to the HEP
To obtain a comprehensive costing approach from different perspectives, the

community contribution attributed to the HEP packages or interventions was
considered. The most applicable literature search was applied to take into
account the viable costs spent by the community. The Ethiopian National
Health Account VI (NHA) documented the community contributions from 2
perspectives: contributions through the Women’s Development Army (WDA)
and those through malaria control programs. Considering this literature, the
community contribution cost per capita was considered in the costing analysis.
Through a review of the NHA, the community contribution to health system

was estimated in the NHA VI report. Accordingly, the per-capita community
contribution to strengthening the health system was estimated at about 36
ETB. Of this, 55% and 45% were contributed through the WDA and the
malaria control program, respectively.7 Using this data, we allocated the WDA
contribution equally to all interventions except malaria and the malaria-specific
contribution for malaria intervention during our calculation of the community
contribution for each intervention.
4.1.6 Cost analysis from the societal perspective
The costing analysis of the HEP was also conducted from the societal perspective.
All the identified resources of the interventions in the family health service
and communicable disease prevention and control service packages were
assessed, taking into account 3 viewpoints: healthcare provider perspective,
client perspective, and community contribution.
4.1.7 Cost savings analysis
The cost savings to the client were calculated as the change in the difference
in cost of seeking similar services at the HP and at the HC considering direct
medical costs, non-medical costs, and productivity loss. We extracted the
proportion of HPs providing the respective services.22 Consequently, the cost
saving/difference to the client was calculated using the following formula:
page- 945
CS=Δ C * Ps * nHP
where Cs: cost savings due to visiting an HP instead of an HC;
ΔC: difference of cost at an HC and at an HP;
Ps: total population served per year at the HP; and
nHP: Number of HPs providing the respective service.
Table 2. Description of HEP input costs and sources of data
S.No Cost inputs Inputs Source of data

Average cost of constructing Primary sources
different types of Health Post
1 Building Proportion of Health Posts
constructed of different materials Primary sources
(hollow block, wood or mud, stone or
brick)
Personnel time (time spent per
client) Both primary and
2 Personnel Authors’ calculation based on time- secondary sources
motion study and total clients served
per month
Unit cost of drugs, supplies,
equipment, and furniture for selected Secondary sources
interventions were collected from
EPSA
Drug, Average functional furniture
equipment, available and supplies used over the Primary
furniture, period
3 and supply Expert opinion and literature were
costs used to estimate the average number
of follow-up visits per client. Expert and
Antenatal care (3), tetanus-toxoid secondary data
(2), family planning (injectable) sources
(1), pentavalent 3 (3), tuberculosis
treatment (48)
page-946
Start-up costs (pre-service training): Primary and

Number of clients served secondary data
Annual cost estimation of pre-service sources
training
In-service training Primary and
Number of clients served secondary data
Annual cost of providing training sources
4 Other costs services
Supervision and review meeting
Number of clients served, total
annual distance of vehicles; cost Expert primary and
of vehicles, and transport cost, secondary data
and total duration for supervision source
from MoH, RHB, and Zonal health
department.
Community contribution to health Authors’ calculation
Community system strengthening, per capita
5 based on secondary
contribution malaria prevention and control data sources
(45%), WDA (55%)
Abbreviations: EPSA, Ethiopian Pharmaceuticals Supply Agency; MoH, Ministry of Health; RHB,
Regional Health Bureau; WDA, Women’s Development Army.
4.2 Effectiveness measure
The primary outcome measure used for this analysis was the number of lives
saved through the HEP between 2005 and 2018, which was calculated using
the lives saved tool (LiST), FamPlan, and TB impact and modeling estimate
(TIME) spectrum models, which translate the measured coverage changes into
estimates of reductions in mortality. The LiST model estimates the number of
lives saved due to maternal and child healthcare interventions, TIME due to
the TB-related mortality effect, and FamPlan due to FP interventions. The LiST
model estimates the effect of coverage changes in maternal and child health
services on mortality. The model:
- is linear,
- has a fixed relationship between input and output,
- is mathematical,
- assumes that the causal pathways of interventions reducing cause-
specific mortality via reducing risk factors are correctly defined,
page- 947
- is deterministic in that it produces the same result each time the

model is run with identical inputs,
- is a population-based intervention, and
- produces results by age cohort.
The interventions are services at the preconceptual, pregnancy, childbirth,

vaccine, and curative levels. The FamPlan module combines the percentage of
users of a particular method, the proximate determinant of fertility (women in
sexual union, postpartum insusceptibility, unintended pregnancy, and sterility),
method attribute, effectiveness of the FP method, and impact and miscarriage
rates to estimate the maternal and child deaths attributable to the coverage
of contraceptive uses. TIME is an epidemiological compartmental transmission
model that projects the drug-susceptible (DS-TB) and multidrug-resistant TB
(MDR) burden.
The mathematical formula to estimate the lives saved in LiST due to the
intervention is:
LS=cause specific mortality*intervention coverage change*AF*E

where LS: lives saved;
AF: affected fraction; and
E: effectiveness of the intervention.
The effectiveness/health outcome measure for the study was then converted to
the final outcome measure of life year gained (LYG).23 The LYG was calculated
as the number of deaths prevented (i.e., lives saved) by an intervention multiplied
by the remaining life expectancy at the point of each averted death and was
retrieved from Ethiopia’s life table.24 Under this method, saving the life of a
younger person accrues more LYGs than does saving the life of an older person.
The LYG was calculated as follows:
LYG=LS*average remaining life expectancy
A discount rate of 3% was used to estimate the current value of future years
gained.
page-948
To estimate the number of lives saved attributable to the HEP, 2 scenarios were
created:
• Scenario 1: Data related to the coverage of the health system
implementation were extracted from the 2016 EDHS.25 This reflects
the coverage of various health services through the HEP in 2016 (i.e.,
the current health system, which includes the HEP).
• Scenario 2: By using the available impact studies, we created a
counterfactual scenario of what would have happened to the health
system coverage in 2016 if there had been no HEP (i.e., a health
system without the HEP). We used various previous studies in Ethiopia
concerning the effect of the HEP on health service coverage. As
there are only a few such studies, we used all the best available and
most recent studies. The non-HEP scenario (i.e., the counterfactual
health system coverage in which there had not been the HEP) is the
difference between the current coverage and the coverage due to
the HEP, as estimated from impact studies on the HEP. Most HEP
services were initiated in 2005, and other health services begin in
2011.
The baseline coverage for both scenarios was considered similar, extracted
from EDHS 2005 for those interventions implemented since the initiation of
the program and from EDHS 2011 for recent intervention scale-up.26,27 In both
scenarios, linear interpolation was used between baseline and current coverage
(Table 3).
For example, to estimate the number of deaths averted by measles vaccine

attributable to investment in the HEP, we used LiST to model changes in the
coverage of measles vaccine between 2005 and 2018, during which period the
interventions are projected to have reached 54.3% of the target population.
Next, we modeled the change in the coverage of measles vaccination over
the same period years as 33.2% in the non-HEP scenario. We then took the
difference between the estimated lives saved in years 2005-2008 in the HEP
and non- HEP scenarios to calculate the effect of the HEP.
page- 949
Table 3. Coverage inputs for effectiveness measures
With
Baseline HEP Change
Target
Intervention coverage (2016 in Source
population (2005) Coverage coverage
(%)
Antenatal care – Total number
provision of iron, of estimated
pregnant mother 28% 62% 20% 25,26,28,29
pregnancies/
malaria treatment, births
and TT vaccination
Percentage
receiving 2 or more Total number
TT injections during of estimated 32% 41% 7% 10,25,26
the pregnancy for pregnancies/
the most recent live births
birth
Total number
Iron of estimated 10% 42% 18.00% 25,26,30
supplementation pregnancies/
births
Non-
pregnant
Family planning 24% 35.30%
women of
fertile age
Short term – OCP (method mix) 0.70% 0.80%
Short term – 11% 25,26,28
0.70% 0.80%
condom
Short term – 17.70% 63.70%
injectable
Long term – 3.40% 22.10%
implant
EPI – Pentavalent 3 Estimated 29% 53% 22.90% 6,25,26
live births
Estimated
EPI – Measles 28.50% 54.30% 21.10% 6,25,26
live births
Number of
populations
TB –DOTs* 84% 89% 6.20% 12,31,32
in need of TB
treatment
page-950
Population at
Malaria – ITN risk (60%) of 33% 40% 7.90% 4,33,34
malaria
Population
residing in
highland
Malaria – IRS 14.20% 29% 7.90% 4,33,34
fringe or
epidemic-
prone areas
Malaria – diagnosis Prevalence of
malaria in all 24.20% 38.20% 20% (27,28,33
and treatment* age groups
(25,27,
Diarrheal treatment Diarrhea 0.30% 33% 5.50% primary
(zinc)* prevalence data)
among <5
Diarrheal treatment years children 40% 46.40% 6.30% (4,25,27
(ORS)*
Acute respiratory
tract and Pneumonia
Pneumonia prevalence 27% 31.30% 9.60% (25,27,35
treatment (cotrim., among <5
amoxa., and years children
gentamicin)*
Improved water Households 51 62 3.1
source
Handwashing with Households 50 60 2.8 25,26,28
soap
*baseline year is 2010.

Abbreviation : TT, tetanus-toxoid; OCP, oral contraceptive pills; EPI, expanded program of immunization;
ITN, insecticide-treated bed net; IRS, indoor residual spray; TB, tuberculosis
page- 951
4.3 Cost-effectiveness analysis
Acost-effectiveness analysis evaluates whether an intervention

provides value relative to an existing intervention (with value defined as the
cost relative to the health outcome).
Since 2004, the health system has incurred additional costs for the provision
of health services due to the initiation of the HEP. All the costs of services
provided through HEWs are incremental or additional to the costs of the health
system. Due to the additional costs of the HEP to the health system, the health
system will bring additional benefit to the community through the HEP. The
incremental cost-effectiveness ratio is calculated as the ratio of additional costs
to the health system because of the HEP divided by the additional benefit due
to the HEP.
Inc.CEA=(Inc.cost)/(Inc.benefit)
Where: Inc. CEA: incremental cost-effectiveness;
Inc. cost: average incremental cost; and
Inc. benefit: average incremental benefit.
The incremental cost-effectiveness ratio (ICER) is calculated for particular

interventions, as well as for selected HEP interventions as a whole.
To determine whether investing in the above services is worth the expense, the
World Health Organization (WHO) recommends a set of rules. Ethiopian GDP
per capita for the year 2018 ($852.80) is used for this threshold. Interventions
that lie within 200% of the GDP per capita of a country are deemed very
cost-effective, and interventions that lie within 300% of GDP per capita are
cost-effective.
4.4 Sensitivity analysis
One-way sensitivity analysis is conducted by varying the unit costs, the discount
rate for effectiveness and cost, the life years of the equipment, the building, pre-
service and in-service training, the total annual number of services provided,
the number of services provided for each intervention, and the salary of the
HEW.
page-952
5 Results
5.1 Cost Analysis of the HEP from Providers’ Perspective
The results of the provider’s cost are divided into 3 groups: unit cost by
components, cost by service delivery modality, and cost by component.
5.1.1 Unit cost by components
The unit cost of providing hygiene and environmental sanitation interventions,

family health service, and disease prevention and control (DPC) interventions,
along with their respective cost centers, are presented in Table 4.
The unit cost of the hygiene and environmental sanitation packages, improved
water sources, handwashing with soap, hygienic disposal of children’s feces, and
latrine use ranges between $0.60 and $0.80. Among the selected family health
services, the provision of full dose pentavalent vaccine for a child immunization
has the highest cost ($15.20). The lowest unit cost is observed in the provision
of iron folate supplementation for pregnant mothers ($0.70).
Similarly, the unit cost of providing the 4 types of FP method offered through
HEWs—oral contraceptive pills (OCP), condoms, injectables, and implants—
is $18.00, $7.20, $12.70, and $44.20 per CYP, respectively. The provision of
implants is the most expensive among the FP services. The average unit cost of
FP is $4.20. The unit cost of ANC for pregnant mothers, measles vaccination for
a child, and tetanus-toxoid vaccinations for pregnant mothers is $1.90, $2.50,
and $3.00, respectively.
Among the DPC interventions, the highest unit cost for a client service is
tuberculosis directly observed treatment (TB DOT, $43.10), followed by malaria
prevention through indoor residual spray (IRS, $3.40), long-lasting insecticidal
nets (LLIN, $2.10), diarrheal disease management ($2.00), and pneumonia
treatment ($2.60).
page- 953
Table 4: Unit costs (US$) of selected HEP interventions, by component, in

Ethiopia, 2018.
Capacity- Unit
Interventions Personnel Supply Equipment Building
building cost
Improved water source 0.11 0 0 0 0.53 0.6
Handwashing with soap 0.2 0 0 0 0.53 0.7
Hygienic disposal of
0.2 0 0 0 0.53 0.7
children’s feces
Latrine use 0.25 0 0 0 0.53 0.8
Antenatal care 0.30 0.75 0.38 0.13 0.32 1.88
Family planning – OCP 0.15 0.34 0.13 0.18 0.41 1.20
Family planning – condom 0.15 0.94 0.13 0.18 0.41 1.80
Family planning –
0.15 2.30 0.13 0.18 0.41 3.16
injectable
Family planning – implant 0.11 9.43 0.19 0.25 0.47 10.45
Pentavalent vaccination 0.16 11.92 2.13 0.54 0.46 15.20
Measles vaccination 0.16 1.01 0.71 0.18 0.46 2.51
Tetanus toxoid 0.24 0.80 1.22 0.36 0.41 3.02
Iron folate 0.11 0.09 0.01 0.14 0.32 0.67
HIV/AIDS testing and
0.27 0.42 0.02 0.18 2.81 3.70
counseling
TB treatment (DOT) 6.54 23.46 2.97 7.50 2.66 43.13
Malaria Case
0.17 0.88 0.14 0.16 0.43 1.78
Management
Malaria prevention –
0.03 2.00 0.00 0.00 0.11 2.14
LLIN
Malaria prevention – IRS 0.49 2.84 0.03 0.00 0.00 3.36
Diarrheal disease 0.64 0.69 0.06 0.18 0.46 2.02
Pneumonia treatment 0.53 1.19 0.10 0.25 0.54 2.60
Pneumococcal vaccination 0.16 6.10 2.13 0.54 0.46 9.38
*If all HEWs’ salary was changed to level IV, the unit cost would increase by
0.1% to 3.06%.
Abbreviation:OCP, oral contraceptive pills; HIV, human immunodeficiency virus; AIDS, acquired
immunodeficiency syndrome; DOT, directly observed treatment; LLIN, long-lasting insecticidal net;
IRS, indoor residual spray.
page-954
Personnel and capacity-building costs are the only cost centers in the hygiene and
environmental sanitation package. Personnel costs contribute to 18%, 29%, and
31% of the costs of an improved water source, handwashing with soap, hygienic
disposal of children’s feces, and latrine use. The rest of the costs are the costs of
capacity-building. Among the cost centers, drug and supply costs contribute the
most to the unit cost of malaria LLINs (93%), implanon (90%), malaria IRS
(84%), pentavalent vaccination (78%) and pneumococcal vaccination (65%).
The lowest percentage contribution by supplies or drugs is for iron folate (13%),
condom provision (28%), and ANC (40%). The share of contribution of the
building cost ranges from 5% for injectable to 20% for the provision of iron
folate (26%). Personnel cost represents the highest percentage of the unit cost
of diarrheal disease management (31%) and pneumonia treatment (20%) and
the lowest of pentavalent vaccination (1%), injectable contraception (1%), and
pneumococcal vaccination (2%). Equipment represents 40% of the cost of TT
and 22% of pneumococcal vaccination. The cost of capacity-building has the
highest share of the unit cost of iron folate (47%), OCP (34%), and malaria
case management (24%).
5.1.2 Unit costs of selected HEP services by delivery modality
The unit cost of providing HEP interventions, whether static provision at the
HP, outreach in the community, or home-to-home visits, is described. The mean
cost of providing hygiene and environmental sanitation interventions through
the HEP is $0.70, and the cost for family health service and DPC interventions
is $4.90 and $7.60, respectively.
The costs of hygiene and environmental sanitation interventions are found

only in home visits. Services provided through outreach and home-to-home
visits cost comparatively less than services provided at a static site. Healthcare
interventions at a static site have higher costs because they include building
and furniture costs, which are not included in the home visit or outreach services.
This is not a recommendation to stop providing service at static sites but an
indication of the coupled benefits of the HEP, which helps provide services
close to the community at a lower cost. The provision of TB DOTS to the clients
through a home visit costs $32.30, less than providing the same service at a
static site. Similarly, providing FP services (OCP, condoms, injectables, and
page- 955
implants) costs less through outreach service than providing the same service
at a static site. ANC and the provision of iron supplementation for pregnant
mothers also costs less for services provided through home-to-home visits than
at the HP. The weighted mean cost of providing the services through each
modality is also described in Table 5.
Table 5: Unit cost of selected HEP interventions by service delivery modality in

Ethiopia.
Outreach Home unit

Package Service Type Static visit cost
Improved water source 0.6 0.6
Handwashing with soap 0.7 0.7
Hygiene and Hygienic disposal of
environmental children’s feces 0.7 0.7
sanitation
Latrine use 0.8 0.8
Weighted mean cost 0.7
Antenatal care 2.59 1.3 1.1 1.88
Family planning – OCP 1.37 0.8 1.2
Family planning – condoms 2.0 1.4 1.8
Family planning – injectables 3.7 1.9 3.2
Family planning – implant 10.5 10.5
Family health
service Pentavalent vaccine 16.4 12.3 15.2
Measles vaccine 3.0 1.2 2.5
Tetanus toxoid vaccine 3.7 1.4 3.0
Iron folate supplementation 0.9 0.4 0.4 0.7
Pneumococcal vaccine 9.9 7.2 9.4
page-956
HIV testing and counseling 4.4 1.6 3.7

Tuberculosis treatment 49.6 32.3 43.1
(DOTS)
Malaria case management 2.1 1.3 1.8
DPC Malaria prevention – LLIN 2.1 2.1
Malaria prevention – IRS 3.4 3.4
Diarrheal disease 2.0 2.0 2.0
Pneumonia treatment 2.6 2.6
Abbreviations: OCP, oral contraceptive pills; DOTS, directly observed treatment; LLIN, long-lasting
insecticidal net; IRS, indoor residual spray.
5.1.3 Unit cost disaggregated by cost center and HEP component
The major cost drivers of the cost of family health services were drugs and
supplies, which account for 68% of the total cost, followed by equipment cost
(15%), capacity-building cost (8%), construction cost (8%), and personnel cost
(4%; see Figure 1).
Figure 1: Percentage distribution of family health services unit costs, by

component
page- 957
The study also showed that drugs and supplies are the leading cost driver of
DPC interventions (53%; see Figure 2). The personnel cost (15%), followed by
the costs of building (14%), equipment (6%), and capacity-building (12%), also
contributes to the unit cost of DPC interventions.
Figure 2. Percentage distribution of DPC unit costs, by component
5.2 Cost of providing healthcare service at HCs and

HPs
The unit cost of providing healthcare services at an HC and through HEWs

is described in Table 6. The provision of health services through the HEP costs
less than providing the same services at an HC. Providing ANC at an HP
costs much less than at an HC, adjusting for possible input differences (e.g.,
laboratory costs). Similarly, providing other family health services and DPC
healthcare interventions cost less in HPs than in HCs. As presented in Table 6,
the provision of healthcare services at an HP costs 20% to 85% less than the
same services provided at an HC. This comparison makes adjustments for the
type of services (e.g., outpatients) and costs, but the quality of the services
provided through the HEP and the HC were assumed to be similar for curative
intervention.
page-958
Table 6: Unit cost of providing selected interventions at Health Centers and

Health Posts.
Health Health
No. Interventions Difference
Post Center
1 Iron supplementation 0.67 1.24 85%
Tetanus-typhoid vaccination for
2 3.02 4.30 42%
pregnant mothers
3 Pentavalent vaccination 15.20 18.19 20%
4 Measles vaccination 2.50 3.94 57%
5 Diarrhea disease management 2.00 2.86 43%
6 Oral antibiotic for pneumonia 2.05 3.51 71%
7 Antenatal care 1.80 14.2
8 Malaria case management 1.80 7.53
At an HC, the costs of drugs and supplies contribute 11-87% of the total cost.
The cost of building contributes to 5-36% of the total cost. Similarly, equipment
contributes to 1-27% of the total cost. Personnel, supervision and review meetings,
and administrative costs contribute to the rest of the unit cost. Although the
difference in the unit cost of providing the services at an HC or HP lies within
all the cost centers, most of the difference is due to the cost of supplies and
drugs, building, and equipment.
5.3 Client perspective cost and cost saving analysis

As illustrated in Table 7, the client cost is analyzed in 2 scenarios: the cost
incurred by clients while receiving the services at an HC, and the cost incurred
for the same services at an HP. This is done by considering the client’s direct
medical cost, direct non-medical cost, and productivity losses due to seeking
the services at an HC or HPs. The highest total client cost per year is reported
for TB treatment: $58.82 at an HC and $9.56 at an HP; the lowest total client
cost per year is for FP (implant), which is $0.19 at an HC, and FP (condoms),
with a cost of $0.01 at an HP.
page- 959
Similarly, the potential cost saving from the client perspective was analyzed
across the family health service and DPC service packages. Each year, the
society (considering direct medical cost, direct non-medical cost, and productivity
losses) would save $26 379 586 as a result of the implementation of the HEP.
This is equivalent to saving $2.30 per capita every year (see Table 7).
page-960
Table 7: Total (direct medical, non-medical, and productivity loss) cost per
client and cost savings of selected HEP interventions from client perspective,
Ethiopia, 2018.
Direct Direct non- Productivity

Total client
medical medical loss cost Cost saving
S.N. Interventions cost (2018) cost (2018)
cost (2018) (2018)
HC HP HC HP HC HP HC HP Difference Cost saved
1 ANC 5.62 0.00 2.81 2.69 2.69 0.76 9.99 3.45 6.54 11094622
2 FP (OCP) 0.00 0.00 0.96 0.91 0.91 0.21 1.67 1.12 0.55 135193
3 FP (condoms) 0.00 0.00 0.01 0.01 0.01 0.00 0.02 0.01 0.01 386
FP
4 0.00 0.00 0.32 0.31 0.31 0.07 0.56 0.38 0.18 564284
(injectables)
5 FP (implant) 0.00 0.00 0.11 0.11 0.11 0.03 0.19 0.14 0.06 60167
6 Iron folate 0.00 0.00 0.7 0.67 0.67 0.19 1.09 0.86 0.23 83369
7 HIV testing 0.00 0.00 0.7 0.67 0.67 0.19 1.09 0.19 0.90 62956
Pentavalent
8 0.00 0.00 2.81 2.69 2.69 0.68 4.33 3.37 0.96 7104852
immunization
Measles
9 0.00 0.00 2.81 2.69 2.69 0.68 4.37 3.37 1.00 1097494
immunization
TT
10 0.00 0.00 2.56 2.69 2.69 0.34 4.08 3.03 1.05 1865993
immunization
Diagnosis and
11 treatment of 3.51 0.00 0.88 2.69 2.69 0.72 5.99 3.41 2.58 1803408
pneumonia
Diagnosis and
12 treatment of 3.34 0.00 0.83 2.69 2.69 0.72 5.77 3.41 2.36 1216840
diarrhea
Diagnosis and
13 treatment of 0.53 0.00 1.98 1.39 1.39 0.36 4.11 1.75 2.36 680447
malaria
Treatment of
14 0.00 0.00 41.8 5.08 5.08 4.48 58.82 9.56 49.26 609575
tuberculosis
Total saving 26 379 586
Abbreviations: ANC, antenatal care; FP, family planning; OCP, oral contraceptive pills; TT, tetanus-
toxoid.
page- 961
5.4 Cost analysis from the societal perspective
The unit cost incurred by society as a result of the HEP ranges from $2.20 (for
iron folate) to $53.40 (for TB prevention and control services). Among family
health service interventions, the highest societal cost is that of pentavalent
vaccination ($19.90) and FP service through implants ($11.30). Among the
DPC interventions, the highest societal cost was the cost of TB treatment and
follow-up (DOTS, $53.40) and the lowest societal cost was the cost of malaria
case management ($3.60). The costs of community contribution, the costs of
the providers, and the costs to the client for the selected Health Extension
service is described in Table 8.
Table 8: Unit cost of selected HEP intervention from societal perspective in

Ethiopia, 2018.
Total
Package Service type Provider Clients Community
cost
Antenatal care 1.9 3.45 0.69 6.0
Family planning – OCP 1.2 1.12 0.69 3.0
Family planning – condoms 1.8 0.01 0.69 2.5
Family planning – injectables 3.2 0.38 0.69 4.2
Family
health Family planning – implant 10.5 0.14 0.69 11.3
service Pentavalent vaccination 15.2 3.37 1.38 19.9
Measles vaccination 2.5 3.37 0.69 6.6
Tetanus toxoid 3.0 3.03 0.69 6.7
Iron folate 0.7 0.86 0.69 2.2
Pneumococcal vaccine 9.4 3.37 0.69 13.5
HIV/AIDS testing and counseling 3.7 0.19 0.69 4.6
DPC TB treatment (DOT) 43.1 9.56 0.69 53.4
Malaria case management 1.78 1.75 0.69 4.2
Malaria prevention – LLIN 2.1 0 3.25 5.4
Malaria prevention – IRS 3.4 0 1.38 4.7
Diarrheal disease 2.0 3.41 0.69 6.1
Pneumonia treatment 2.6 3.41 0.69 6.7
Abbreviations: ANC, antenatal care; FP, family planning; OCP, oral contraceptive pills; TB, tuberculosis;
DOT, directly observed treatment; LLIN, long-lasting insecticidal nets; IRS, indoor residual spray.
page-962
5.5 Cost-effectiveness analysis
The cost-effectiveness results are presented, from the most to the least cost-
effective intervention, in Table 9. The incremental cost-effectiveness ratio of
improved water source, measles vaccination, handwashing with soap, TT
vaccination for pregnant women, ANC, and iron supplementation for pregnant
women lie between $21.60 and $57.50 per LYG. This indicates that providing
that the above interventions through HEWs would incur an additional $21.60
to $57.50 to the health system for every additional LYG. Similarly, pentavalent
vaccination, oral antibiotic for pneumonia treatment, ORS for diarrheal disease
management, zinc and ORS for diarrheal case management, and malaria case
management have an ICER that lies between $64.80 and $81 per LYG. This
would mean, that to gain 1 more life year through the HEP, the health system
would invest $64.80 to $81 for each respective intervention. Pneumococcal
vaccination, TB treatment follow-up (DOTs), LLINs, and the provision of FP
services have an ICER that lies between $103.70 and $295.40 per LYG.
All the interventions are very cost-effective according to WHO cost-effectiveness

criteria, in that the cost per LYG lies within 200% of GDP per capita when
compared with Ethiopia’s GDP per capita income of $852.80. Among the
interventions, improved water source, measles vaccination, handwashing with
soap, TT vaccination for pregnant women, ANC, and iron supplementation for
pregnant women lie between 2.5% and 6.7% of the country’s GDP per capita
for each additional LYG. Interventions such as pentavalent vaccination, oral
antibiotics for pneumonia treatment, ORS for diarrheal disease management,
zinc and ORS for diarrheal case management, and malaria case management
have costs that lie between 7.5% and 9.4% of GDP per capita for each additional
LYG (Table 9). This means all of the above interventions far less than 100%
of GDP per capita to gain 1 more additional life year: i.e., they are very cost-
effective interventions. Similarly, pneumococcal vaccination, TB treatment
follow-up (DOTs), LLINs, and the provision of FP services have costs that lie
between 12% and 34% of the country’s GDP per capita to gain 1 more life year.
This is also a very cost-effective intervention.
page- 963
Table 9: Cost-effectiveness of selected Health Extension Program interventions

in Ethiopia, 2005/10-2018.
Life year
Intervention Lives saved Total cost ICER
gained
Improved water source 369.2 10 976 237 037 21.6
Measles vaccination 1 068 31 761.8 979 293 30.8
Handwashing with soap 248.1 7 376.0 248 288 33.7
Tetanus toxoid vaccination 330 9 812.2 419 786 42.8
Antenatal care 505 14 357.0 677 272 47.2
Iron supplementation 160 4 174.0 239 877 57.5
Pentavalent vaccination 3 311 98 443.8 6 376 381 64.8
Oral antibiotics for pneumonia 804 23 916.5 1 607 311 67.2
Diarrheal disease management
1 468 43 633.2 3 409 507 78.1
(zinc and ORS)
Oral rehydration solution (ORS) 1 301 38 690.1 3 125 879 80.8
Malaria case management 85 2 467.0 199 953 81.0
Pneumococcal vaccination 1 084 32 219.5 3 341 373 103.7
Tuberculosis treatment (DOTs) 95 1 957.4 222 713 113.8
Long lasting insecticide net 67 1 936.5 315 210 162.8
Family planning services 536 14 098.8 4 164 292 295.4
Overall, selected HEP intervention 10 927 321 463.0 24 886 899 77.4
Abbreviations: ORS, oral rehydration salt; DOT, directly observed treatment; LLIN, long-lasting
insecticidal nets.
Over 1 year, the selected interventions save 10 927 lives, or 0.254 lives saved per
1 000 population. The incremental cost-effectiveness ratio of the HEP is $77.40
for each additional LYG. This value lies within the country’s GDP per capita of
$852.80, making it a very cost-effective program. The program costs 9% GDP
per capita to gain an additional life year, a very cost-effective program.
For example, if we consider the cost of $100 per LYG gained as a reference
for an attractive use of healthcare resources in low-income countries, then
implementing 11 low-income healthcare interventions costing less than $100 per
LYG gained at the HP level will be necessary (Table 9).
page-964
5.6 Sensitivity analysis
The results of the sensitivity analysis, which was conducting by varying the
parameters, such as cost, effectiveness, life year of capital items, pre-service
and in-service trainings, and salary, are presented in Figure 3 and Annex 3.
Among all interventions, age discounting has the highest effect on the baseline
ICER, which changes slightly when the discount rate per LYG is altered to non-
age-discounting and 6% age-discounting. In almost all interventions, the age
discounting, when changed from 3% to 6%, increases the base case ICER by
200%. Conversely, when it is changed from a 3% discount to no discount, the
base case ICER falls by half. Next to age, discounting life years of equipment
has the highest effect on ICER of the intervention. For example, the ICER of
TT vaccination increases by 30% when the life years of the equipment are
changed from 5 to 3 years, as presented in Figure 3. Conversely, it decreases
by 10% when the life years of equipment are changed from 5 to 7 years. The
change in the ICER of TT vaccination is a good example to indicate the effect
of poor-quality equipment and furniture, which incur a 30% additional cost to
gain 1 more life year due to their decreased service years. Other variables, such
as unit costs, number of services provided for each service, discounting costs,
total number of service users, building year of services, building costs, HEW
salaries, and pre-service and in-service training years of service have a minimal
effect on the baseline ICER. Although change in age discounting has some
effect on the ICER, the range still lies within 100% of the country’s GDP per
capita, and the overall finding is robust with the baseline ICER.
page- 965
Figure 3. One-way sensitivity analysis showing an incremental cost-effectiveness

ratio of pneumonia treatment over a range of key parameters
page-966
6 Discussion
Ethiopia has achieved a massive reduction in maternal and child mortality
since launching the HEP in 2004. This study has analyzed the unit costs, client
costs, cost savings, and cost-effectiveness of providing selected Health Extension
service packages.
This study used the tailored costing of the HEP as a cost benchmark to guide
the planning for the HEP. The costs of HEP interventions are presented in broad
ranges that provide a reasonable guide for planning purposes. The unit costs
of providing HEP services range from $0.70 (iron folate supplementation) to
$15 (pentavalent vaccination) for family health service interventions. Similarly,
the unit cost of providing DPC services ranges from $2 (diarrheal disease
management) to $43.10 (TB DOT). The unit costs of the HEP, based on health
needs and supply constraints, will provide contextually appropriate input
for decision-makers in allocating resources for HEP activities. The weighted
average cost of family health services and DPC services is $4.90 and $7.40,
respectively. DPC interventions have more cost variability in their unit cost than
does the cost estimate of family health services. A study conducted in Ghana
in 2015 estimated the unit cost of providing health services at a community
level to be $5.14. The same study also estimated the cost of curative and
preventive services to be 34% and 56%, respectively. Although the total costs
largely depend on the interventions included in the community health program,
the estimated cost of family health services and DPC services are similar to the
average costs estimated in Ghana.28 This will help guide planners and decision-
makers by providing an estimate of the cost of providing family health services
through the HEP.
The societal cost of providing TB directly observed treatment (DOT) to the

patient through the HEP is $53.40, similar to the cost found in a previous study
conducted in 2007 in Ethiopia, which estimated the cost at $60.70.12 In a study
conducted in 2014 in Brazil, the cost of TB DOT by CHWs was $523.
The costs of providing diarrheal case management, oral antibiotics for

pneumonia, and malaria case management through the HEP are $2.00, $2.60,
and $1.80, respectively. Another study conducted on integrated community case
page- 967
management (iCCM) programs in 7 countries estimated the cost per treatment
to range from $2.44 to $13.71 for diarrhea case management, $2.17 to $17.54 for
malaria case management, and $1.70 to $12.94 for pneumonia treatment in the
years between 2010 and 2012.29 Another study estimated the unit cost of home
management of uncomplicated malaria to be $4.22 in Zambia.30 The variation
in the cost of service provision is attributed to the costing method, the resources
included, and the difference in the use of each type of resource, the prevalence
of the disease, and the economy of each country.
Ethiopia has also decentralized the implementation of indoor residual spraying

(IRS) from the district to the community level by incorporating the planning
and execution of the operation into the HEP. This study found that the unit
cost of providing IRS through the HEP is $3.40. A study conducted in Ethiopia
in 2013 estimated that district-based IRS costs were higher than the cost of
community-based IRS through the HEP, both in cost per district and in cost
per person protected.31 Coupled with the benefit of more structures’ now being
eligible for IRS and at a lower cost through community-based rather than
district-based IRS, the health system would offer a massive health benefit to
the community at a lower cost.
The average annualized cost of delivering services through the HEP was $9 897,
slightly lower than the cost found in a study conducted in Ghana; this difference
may be attributed to the high annualized personnel cost in the latter study.28 A
similar study conducted in India in 2014 estimated the annual cost of delivering
healthcare interventions through CHWs to be $19 381.32
The decentralization of the healthcare services from HCs to HPs has played
a key role in satisfying the need for essential healthcare services. In terms of
cost, services provided through HEWs cost 20% to 85% less than similar costs
provided at HCs. Although the content and quality of services provided at
both facilities differ somewhat, we have tried to make them comparable with
respect to service provision. Other studies have also confirmed that providing
health services through CHWs or HEWs costs much less than providing the
same services at HCs.12 It should be clear, however, that the decentralization
of healthcare services does not mean that the HEP is a stand-alone program;
rather it is a complementary program to the healthcare services provided at
HCs.
page-968
The HEP has increased communities’ access to healthcare services. The provision
of the HEP at a community level has saved clients $26 million annually for the
selected interventions, or $2.30 per capita each year. Studies indicate that cost
savings have tended to link returns to the clients’ reduced emergency department
visits and avoidance of preventable hospitalizations.33 This means that, due to
the interventions through the HEP, the clients would save a substantial amount
of money, have improved productivity by avoiding hospitalization, and improve
the overall health status of the community. One study conducted in Uganda on
the iCCM of malaria, diarrhea, and pneumonia treatment reported that the
community-based treatment of childhood illness saves money for clients when
compared with delivering the services in other health facilities.34
This study on the cost-effectiveness of the HEP indicates that the program is
very cost-effective, at an additional cost of $77.40 for every additional LYG in
the health system. Similarly, some studies and systematic reviews of the cost-
effectiveness of providing a healthcare intervention through CHWs indicate
that providing essential healthcare service through CHWs is cost-effective or
very cost-effective in their respective countries.35,36 One study conducted in
Nepal in 2003 estimated the ICER to be $211 per LYG of providing service
through women’s groups’ healthcare interventions to improve birth outcomes
in rural Nepal. Similarly, other community health programs delivered through
participatory women’s groups in 2013 estimated an ICER of $79 per DALY
averted of providing healthcare interventions in rural Malawi, which benefitted
the community with profound maternal and child health gains.37 One study
conducted in districts in Ethiopia, Kenya, and Indonesia in 2013 indicated that
community health programs (including the HEP in Ethiopia) were cost-effective,
with an ICER of $999 per LYG in Ethiopia, $82 per LYG in Kenya, and $3 397
per LYG in Indonesia. In our study, however, the HEP is a very cost-effective
program. This difference in the ICER between the previous study and our study
is because the estimation of cost was performed with data from one district
(whereas we used a national estimate), the cost of building construction type
was not considered (we considered the actual type of building construction),
and the implementation period was 3 years (we used 8 and 13 years). The
study showed that providing health services through HEWs was cost-effective
in Ethiopia, although our study suggests that the HEP is very cost-effective.10 A
study conducted in 2010 in Ethiopia on the cost-effectiveness of possible serious
bacterial infections with amoxicillin and gentamicin through HEWs and CHWs
estimated an ICER of $223 per DALY.38 Although the study was performed on
page- 969
only a single intervention and differs in its measure of health outcome, its cost-
effectiveness ratio finding is similar to that of this study.
The cost-effectiveness of providing malaria treatment through HEWs is $81

per LYG. A study in Ghana from 2012 showed the cost-effectiveness of treating
children with artesunate–amodiaquine to be $90 per DALY averted. In both
Ethiopia and Ghana, malaria treatment through HEWs or CHWs was very
cost-effective.39 A study conducted in India in 2019 on CHWs’ delivery and
facilitation of measles vaccination indicated that providing the service through
CHWs was very cost-effective, at $162 per DALY averted.40 Our study has
found pneumococcal vaccination through CHWs to be very cost-effective in
Ethiopia, similar to the findings in the Indian case, although the measures of
outcome differ.
Other study conducted on CHWs estimated the cost per DALY averted for
key child survival interventions to be $67 (with a range of $27 to $92) in
Mozambique, Rwanda, and Malawi.41 The study indicated that community
health programs represent an attractive and low-cost method of decreasing
child mortality and increasing the coverage of key child interventions for
the survival of a child. Another study conducted in Bangladesh compared
community healthcare with home-based healthcare for maternal and neonatal
interventions. The study found that implementation of the home-care strategy
was very cost-effective from both provider and societal perspectives, with
an incremental cost-effectiveness ratio of $103 per DALY averted. The HEP
services in Ethiopia are delivered through the static, outreach, and home-to-
home visit modalities. Although our study addresses the HEP as a whole, the
studies from Bangladesh and other countries provide us with the message that
the provision of maternal and child healthcare interventions through home-to-
home visits is very cost-effective.
Although making a meaningful comparison across cost-effectiveness results is

challenging, due to variations in time horizons, settings, perspectives, impact
measures, disease burdens, and range of interventions, aggregate evidence from
other research offers a detailed and genuine lesson that the HEP or CHP is a
cost-effective or very cost-effective strategy. This is similar to the findings from
our study despite the differences mentioned above. This study offers direction
to health planners in using the evidence of this cost-effectiveness analysis to
apply good “value for money” in the HEP investment across a variety of specific
page-970
healthcare interventions, with expectations that are reasonable compared to

cost-effectiveness benchmarks. Sensitivity analysis indicates the robustness of
the base case ICER value when varying the different inputs.
page- 971
7 Strengths and Limitations

A number of limitations should be considered in the interpretation of the study
findings, including the wide heterogeneity in methodologies and perspectives
adopted and in the choice of outcome measures.
Although the HEP addresses a wide range of health conditions, this study
restricted its assessment to interventions with clear costs and outcome measures
and excluded some HEP components (i.e., health education) due to the
unavailability of cost and effectiveness measures. Even if the effectiveness data
used were drawn from secondary sources, and although care was taken to
identify studies that used robust methods, different factors may play into the
change in coverage despite the HEP implementation over this period. Second,
the health outcome measure used in this study was LYG. Ideally, DALYs averted
would be the effectiveness measure of choice. In the study, the majority of an
intervention’s health benefits are from its prevention of premature mortality (not
from prevented morbidity), so a cost per LYG threshold can be used as a proxy.
For those interventions that have an effect on the prevention of morbidity as well
as mortality, however, using this approach would underestimate the number of
DALYs averted. Third, because of the LiST tool’s limitations, the calculation of
the number of lives saved does not reflect the effect of malaria or pneumonia
on adult mortality; thus, our estimates are likely to have underestimated the
actual number of lives saved. Fourth, as the service provided at an HP or HC
may differ in its quality of care, we tried to limit our cost analysis to those
similar services provided at both levels and on an outpatient basis.
Despite this limitation, our study has a number of strengths. All the important cost
centers were included in the unit cost estimation. In addition, cost-effectiveness
was estimated over the period of the HEP’s implementation (i.e., 6-11 years,
depending on when the intervention was introduced), not a 1-year outcome.
This gives us sufficient time to see the effect of interventions on health outcomes
and suggests that the difference in outcome may be due to improvements in
the coverage of the intervention.
page-972
8 Conclusion
This study found that the average unit cost of providing family health services
through the HEP is $4.90 (ranging from $0.70 for iron folate supplementation
to $15.20 for full-dose pentavalent vaccination). Similarly, the unit cost for DPC
was $7.40 (ranging from $1.80 for malaria case management to $43.10 for
TB treatment) and $0.70 for hygiene and sanitation (ranging from $0.60 for
improved water source to $0.80 for latrine use).
It is critical to observe that, across these different service packages, the major
cost driver is the cost of drugs and supplies, accounting for as much as 53-68% of
the total cost. Similarly, personnel costs, including capacity-building, account for
12-27%, equipment and furniture account for 6-15%, and infrastructure accounts
for 5-14%. For example, for pentavalent vaccine, the drug cost represents 78%
of the total cost, while for TB it is 54% and for diarrheal diseases 34%. This has
significant implications for the use of the country’s scarce resources.
Improving the staffing pattern of the HP beyond 2 HEWs also has a cost
implication, as it requires both a salary payment and a capacity-building
component that adds an additional 12-27% of the unit cost.
Moreover, the average annual cost of delivering the selected healthcare

interventions through the HEP is $9 897 per HP in the 2018 price year; this cost
compares well with the general health delivery costing estimate of the HEP in
Ethiopia. This implies that the establishment of an additional HP as a physical
facility and the provision there of the standard HEP will have a cost implication,
which requires a serious and thoughtful managerial decision-making process.
The unit cost of delivering service through HEWs is 20-85% less than the cost
of delivering similar healthcare services at higher-level health facilities. This
indicates that it costs less to provide healthcare services at the community level
than at other health facilities. The provision of healthcare services through the
HEP, however, is not a stand-alone strategy but a complementary approach to
the other mechanisms of delivering healthcare services in the country.
page- 973
This study has also found that seeking care from an HP not only is cost-effective
from the provider’s perspective but also potentially saves the client money,
amounting to $2.30 per year compared to seeking care at a higher level. The
implementation of the family health service and DPC interventions through
the HEP renders a high cost savings from the patient’s perspective in terms of
reducing their out-of-pocket expenses (direct medical and direct non-medical
expenses, as well as the expense of productivity loss).
One of the most important findings of this study was that the HEP’s cost-
effectiveness is $77.40 for each additional LYG; in other words, to gain 1
more life year through the HEP, the health system would invest $77.40 in the
interventions required. This lies within 100% of GDP per capita per LYG, making
it very cost-effective. Overall investment in the HEP provides good value for
money, and in this respect Ethiopia compares favorably to other similar studies
conducted in Ethiopia and elsewhere in Africa and Asia.
page-974
9 Recommendations
We believe that this study, which produced comprehensive CEA findings, is

useful in further and serious policy dialogue and informed decision-making.
Among the major recommendations are:
• Expanding access to primary healthcare services should take
advantage of the cost effectiveness of providing services at health
post level which, in this study, was found to be more cost effective
than providing similar services at higher levels of the health system
provided that quality of care is not compromised.
• Any decision to expand the HEP by an additional HP would have a
cost implication amounting to a unit cost of $9 897, including the cost
of construction and provision of the standard HEP. The major cost
drivers were the recurrent cost component of the investment in the
HP. The addition of new services to the HEP also has a serious cost
implication. For example, the inclusion of additional family health
services would increase the average unit cost of family health service
by 32% in additional drug and supply costs for the new intervention.
• The HEP strategies tend to offer cost savings to the client associated
with the improved coverage of essential services and support from
the health system. The provision of the program could save a huge
amount of money in terms of saving clients’ time, productivity, and
out-of-pocket expenditures.
• The CEA of the HEP may be an important first step toward expanding
the horizon of cost analysis, where managers could use evidence-
informed planning and resource-allocation decisions also to address
the high burden of disease within the context of weak administrative
data systems and severe financial constraints.
• The cost-effectiveness analysis is an important tool in helping health
managers and decision-makers argue strongly and defend their case
for additional investment in the HEP and beyond. It can also potentially
assist them in priority-setting by identifying the most cost-effective
interventions in the package. This requires careful consideration of
the need to identify interventions within the HEP that provide good
health gains at low lost.
page- 975
References
1. Federal Ministry Of Health. Essential health Service Package for

Ethiopia. Addis Ababa: Artistic printing Enterprise, 2005.
2. Federal Ministry of Health Ethiopia. Health Sector Transformation Plan
2015/16 - 2019/20 (2008-2012 EFY). In. Addis Ababa, Ethiopia,: Federal
Ministry of Health Ethiopia; August 2015.
3. Federal Ministry of Health, Ethiopia. Health and health related indicators
2009 EFY (2016/2017).
4. Admassie A, Abebaw D, Woldemichael AD. Impact evaluation of the
Ethiopian Health Services Extension Programme. Journal of Development
Effectiveness. 2009;1(4):430-449.
5. Assefa Y, Gelaw YA, Hill PS, Taye BW, Van Damme W. Community health
extension program of Ethiopia, 2003–2018: successes and challenges
toward universal coverage for primary healthcare services. Globalization
and Health. 2019;15(1):24.
6. Amare SA. The Impact of Ethiopian Health Services Extension Program
on Maternal and Child Health Outcomes:The Case of Tigray Region,
ScholarWorks @ Georgia State University; 2013.
7. Ethiopia Federal Ministry of Health. Ethiopia’s Sixth National Health
Accounts 2013/2014. Addis Ababa, Ethiopia.
8. Mann C, Ng C, Akseer N, et al. Countdown to 2015 country case
studies: what can analysis of national health financing contribute to
understanding MDG 4 and 5 progress? BMC Public Health. 2016;16
Suppl 2(Suppl 2):792-792.
9. World Health O. The world health report 2000 : health systems :
improving performance. World Health Organization. http://www.who.
int/whr2001/2001/archives/2000/en/contents.htm. Published 2000.
Accessed.
page-976
10. McPake B, Edoka I, Witter S, et al. Cost-effectiveness of community-

based practitioner programmes in Ethiopia, Indonesia and Kenya. Bull
World Health Organ. 2015;93(9):631-639A.
11. Canavan ME, Linnander E, Ahmed S, Mohammed H, Bradley EH. Unit
Costing of Health Extension Worker Activities in Ethiopia: A Model for
Managers at the District and Health Facility Level. Int J Health Policy
Manag. 2017;7(5):394-401.
12. Datiko DG, Lindtjorn B. Cost and cost-effectiveness of treating smear-
positive tuberculosis by health extension workers in Ethiopia: An ancillary
cost-effectiveness analysis of community randomized trial. PLoS ONE
PLoS ONE. 2010;5(2).
13. Federal Ministry Of Health. Cost effectiveness, costing and qualitative
analysis of the Family Health Team at the piloted health facilities in
Addis Ababa, Health Economics and Financing Analysis Case Team,
May 2018.
14. Cunnama L, Sinanovic E, Ramma L, et al. Using Top-down and Bottom-
up Costing Approaches in LMICs: The Case for Using Both to Assess
the Incremental Costs of New Technologies at Scale. Health economics.
2016;25:53-66.
15. Federal Democratic Republic of Ethiopia, Ministry of Health. Health
and Health-Related Indicators 2009 EFY (2016/2017).
16. Federal Democratic Republic of Ethiopia, Federal Ministry Of Health.
Human Resource Department, national health work force update. 2019.
17. Haddix AC, Teutsch SM, Corso PS. Prevention effectiveness : a guide to decision
analysis and economic evaluation. http://search.ebscohost.com/login.
aspx?direct=true&scope=site&db=nlebk&db=nlabk&AN=1921010.
Published 2003. Accessed.
18. Smith RD, Keogh-Brown MR, Barnett T. Estimating the economic impact
of pandemic influenza: An application of the computable general
equilibrium model to the U.K. Social science & medicine. 2011;73(2):235-
244.
page- 977
19. Adam T, Koopmanschap M, Evans D. Cost-effectiveness analysis: Can we

reduce variability in costing methods? International journal of technology
assessment in health care. 2003;19:407-420.
20. Rice DP. Estimating the cost of illness. Am J Public Health Nations
Health. 1967;57(3):424-440.
21. Rapsomanikis G. The economic lives of smallholder farmers. 2015.
22. Ethiopian Public Health Institute, Federal Ministry of Health, World
Health Organization. Ethiopia Services Availability and Readiness
Assessment 2016. In:2016.
23. Robberstad B. QALYs vs DALYs vs LYs gained: What are the differences,
and what difference do they make for health care priority setting?
NORSK EPIDEMIOLOGI. 2005;15(2):183-191.
24. Turel O, Gonen I, Acar E, Hatipoglu N. Clinical Characteristics and
Cost Burden of Children Hospitalized with Pandemic Influenza A (H1N1-
2009) in a Tertiary Care Center in Istanbul. Balkan medical journal.
2014;31(3):266-267.
25. Federal Democratic Republic of Ethiopia, Ministry of Health. HEPCAPS
II Project. 2015. Health Extension Workers Time Motion Study
Complemented by In-depth Interviews with in Primary Health Care
Units in Ethiopia. Ethiopian Federal Ministry of Health, Harvard T.H.
Chan School of Public Health, Yale Global Health Leadership Institute,
JSI Research & Training Institute, Inc.: Addis Ababa, Ethiopia, Boston,
Massachusetts, and New Haven, Connecticut.
26. Central Statistical Agency and ORC Macro. Ethiopia demographic and
health survey, 2005. [Addis Ababa]: Central Statistical Agency; 2007.
27. Central Statistical Agency and ICF International. Ethiopia : demographic
and health survey, 2011. Addis Ababa, Ethiopia; Calverton, Md.: Central
Statistical Agency ; ICF International; 2012.
28. Dalaba MA, Welaga P, Matsubara C. Cost of delivering health care
services at primary health facilities in Ghana. BMC Health Serv Res.
2017;17(1):742-742.
page-978
29. Collins D, Jarrah Z, Gilmartin C, Saya U. The costs of integrated

community case management (iCCM) programs: A multi-country
analysis. J Glob Health. 2014;4(2):020407-020407.
30. Chanda P, Hamainza B, Moonga HB, Chalwe V, Banda P, Pagnoni F.
Relative costs and effectiveness of treating uncomplicated malaria in
two rural districts in Zambia: implications for nationwide scale-up of
home-based management. Malar J. 2011;10:159-159.
31. Johns B, Yihdego Y, Kolyada L, et al. Indoor Residual Spraying Delivery
Models to Prevent Malaria: Comparison of Community- and District-
Based Approaches in Ethiopia. Global Health: Science and Practice.
2016;4.
32. Prinja S, Jeet G, Verma R, et al. Economic analysis of delivering primary
health care services through community health workers in 3 North Indian
states. PloS one. 2014;9(3):e91781.
33. Rosenthal MB, Alidina S, Friedberg MW, et al. A Difference-in-Difference
Analysis of Changes in Quality, Utilization and Cost Following the
Colorado Multi-Payer Patient-Centered Medical Home Pilot. Journal of
General Internal Medicine. 2016;31(3):289-296.
34. Soremekun S, Kasteng F, Lingam R, et al. Variation in the quality and
out-of-pocket cost of treatment for childhood malaria, diarrhoea, and
pneumonia: Community and facility based care in rural Uganda. PloS
one. 2018;13(11):e0200543.
35. Vaughan K, Kok MC, Witter S, Dieleman M. Costs and cost-effectiveness
of community health workers: evidence from a literature review. Human
resources for health. 2015;13:71.
36. Nkonki L, Tugendhaft A, Hofman K. A systematic review of economic
evaluations of CHW interventions aimed at improving child health
outcomes. Human resources for health. 2017;15(1):19.
37. Colbourn T, Pulkki-Brännström A-M, Nambiar B, et al. Cost-effectiveness
and affordability of community mobilisation through women’s groups
and quality improvement in health facilities (MaiKhanda trial) in Malawi.
Cost Effectiveness and Resource Allocation. 2015;13(1):1.
page- 979
38. Mathewos B, Owen H, Sitrin D, et al. Community-Based Interventions for

Newborns in Ethiopia (COMBINE): Cost-effectiveness analysis. Health
policy and planning. 2017;32(suppl_1):i21-i32.
39. Nonvignon J, Chinbuah MA, Gyapong M, et al. Is home management
of fevers a cost-effective way of reducing under-five mortality in Africa?
The case of a rural Ghanaian District. Tropical medicine & international
health : TM & IH. 2012;17(8):951-957.
40. Bettampadi D, Boulton ML, Power LE, Hutton DW. Are community
health workers cost-effective for childhood vaccination in India? Vaccine.
2019;37(22):2942-2951.
41. Perry H, Morrow M, Davis T, et al. Care Groups II: A Summary of the
Child Survival Outcomes Achieved Using Volunteer Community Health
Workers in Resource-Constrained Settings. Global Health: Science and
Practice. 2015;3(3):370.
page-980
Annex 1: Cost of Services at Health Centers
Intervention Literature review method Reference
Direct medical cost

A study examining out-of-pocket payments
for maternal health in rural Ethiopia, The
paradox of free healthcare un-affordability in
Antenatal care 1
Butajira, was conducted from October 2007
to May 2008 in different public and private
health facilities.
There is no available literature on the direct
medical cost of family planning. Therefore,
we used expert opinions and, based on the Expert
Family planning healthcare financing strategy, family planning opinion
is one of the exempted maternal health
services, and there is no fee at the Health
Center for any type of family planning.
medical cost of immunizations. Therefore,
we used expert opinions and, based on the
Pentavalent Expert
healthcare financing strategy, family planning
vaccine opinion
is one of the exempted child health services,
and there is no fee at Health Center for
immunizations.
medical cost of immunizations. Therefore,
we used expert opinions and, based on the Expert
Measles vaccine healthcare financing strategy, family planning opinion
is one of the exempted child health services,
and there is no fee at Health Center for
immunizations.
medical cost of vaccination. Therefore, we used
Tetanus-toxoid expert opinions and based on the healthcare Expert opinion
vaccine financing strategy, TT vaccination is one of the
exempted child health services, and there is no
fee at Health Center for vaccination.
page- 981
A national study was performed on household

Pneumonia expenditures on pneumonia and diarrhea
diagnosis and 2
treatment in Ethiopia, and a facility-based
treatment study in 5 regions were reviewed.
A national study was performed on household
Diarrhea treatment expenditures on pneumonia and diarrhea 2
treatment in Ethiopia, and a facility-based
study in 5 regions were reviewed.
A study was performed on the economic
burden of malaria and predictors of cost
variability to rural households in south-central
Malaria diagnosis Ethiopia. It estimates the direct, indirect, and 3
and treatment total malaria costs to the household at Health
Centers, Health Posts, and overall for both
levels of care. We took the direct medical costs
from this paper.
A study: Cost and Cost-Effectiveness of
Treating Smear-Positive Tuberculosis by
Tuberculosis Health Extension Workers in Ethiopia: An 4
treatment Ancillary Cost-Effectiveness Analysis of
Community Randomized Trial
Direct non-medical cost
A study examining out-of-pocket payments
for maternal health in rural Ethiopia, The
paradox of free healthcare un-affordability in
Antenatal care Butajira, was conducted from October 2007
to May 2008 in different public and private
health facilities.
Half of the direct non-medical costs of
antenatal care follow-up will be equal to that Expert
Family planning of family planning, since there will not be an opinion
attendant.
Pentavalent The direct non-medical cost of antenatal care Expert
vaccine will be equal to that of immunization. opinion
The direct non-medical cost of antenatal care Expert
Measles vaccine will be equal to that of immunization. opinion
Tetanus-toxoid The direct non-medical cost of antenatal care Expert
vaccine will be equal to that of immunization. opinion
page-982
A national study done on household

Pneumonia expenditures on pneumonia and diarrhea
diagnosis and 2
treatment in Ethiopia: a facility-based study in
treatment 5 regions was reviewed.
A national study done on household
Diarrhea treatment expenditures on pneumonia and diarrhea 2
treatment in Ethiopia: a facility-based study in
5 regions was reviewed.
A study was performed on the economic
burden of malaria and predictors of cost
variability to rural households in south-central
Malaria diagnosis Ethiopia. It estimates the direct, indirect, and 3
and treatment total malaria costs to the household at Health
Centers, Health Posts, and overall for both
levels of care. We took the direct medical costs
from this paper.
Treating Smear-Positive Tuberculosis by
Tuberculosis Health Extension Workers in Ethiopia: An 4
treatment Ancillary Cost-Effectiveness Analysis of
Community Randomized Trial.
Productivity loss
We extracted the travel time to the Health
Center, the time spent at the Health Center
(both time spent waiting for and receiving the
service) from an unpublished thesis and the
daily income of farmers from the Food and
All selected Agriculture Organization. For the tuberculosis 5, 6, 4
interventions study, we identified the client and their
attendant transport cost, travel time cost,
and food cost, and used it accordingly. For
all interventions except family planning, we
included an attendant.
page- 983
Annex 2: Cost of Services at Health Posts
Interventions Literature review method Reference

Direct medical cost
There are no studies that reveal the direct
medical costs of antenatal care service at
Antenatal care, Health Posts. Since antenatal care, family
family planning, planning, and Expanded Program of
pentavalent, Expert
Immunization services are some of the exempted
measles, and opinion
maternal and child health services and there is
tetanus-toxoid no system that allows HEWs to collect fees from
vaccines clients, we identified the direct medical cost as
zero.
HEWs follow, diagnose, and treat pneumonia
Diagnosis and and diarrhea following the iCCM/CBNC
treatment of Expert
booklet, and collect no fees from the clients.
pneumonia and opinion
Therefore, we identified the direct medical cost
diarrhea for these interventions as zero.
A study was performed on the economic burden
of malaria and predictors of cost variability to
rural households in south-central Ethiopia. It
Malaria diagnosis estimates the direct, indirect, and total malaria 1
and treatment costs to the household at Health Centers,
Health Posts, and overall for both levels of
care. We took the direct medical costs from this
paper.
Treating Smear-Positive Tuberculosis by Health
Tuberculosis Extension Workers in Ethiopia: An Ancillary 2
treatment Cost-Effectiveness Analysis of Community
Randomized Trial. We took the cost of an anti-
tuberculosis drug from this.
Direct non-medical
cost
page-984
Interventions Literature review method Reference

Antenatal care,
family planning,
pentavalent,
measles, and
tetanus-toxoid Primary data on the travel expenses to the
vaccines, and Health Post (in ETB) were used here.
treatment and
diagnosis of
pneumonia and
diarrhea
The mean direct non-medical costs
(transportation, food, and other costs) paid
while receiving the service at the Health Post
Malaria diagnosis were obtained from a study done on the 1
and treatment economic burden of malaria and predictors of
cost variability for rural households in south-
central Ethiopia.
We took the transport cost to the Health Post
Tuberculosis from a study on cost and Cost-Effectiveness of 2
treatment Treating Smear-Positive Tuberculosis by Health
Extension Workers in Ethiopia
Productivity loss
We extracted the travel time to the Health
Center, the time spent at the Health Center
(both time spent waiting for and time spent
receiving the service) from an unpublished
thesis and the daily income of farmers from
the Food and Agriculture Organization. For 5,6,4 and
All selected the tuberculosis study, we identified the client expert
interventions and their attendant transport, travel time, and opinion
food costs, and used them accordingly. For
all interventions except family planning, we
included an attendant. We made an assumption
about the number of working hours per day for
farmers.
page- 985
Annex 3: One-way Sensitive Analysis of
All Interventions
Effect. Cost Equipment Building

Cost
discount discount life years life years
Baseline ICER
Intervention 10% increase
10% decease
6% discount
no discount
decrease
3 year
5 year
7 year
6.00%
1.00%
Tetanus
toxoid vaccine 43 47 39 19 73 41 46 54 38 44
(maternal)
Iron
57 63 52 32 90 54 63 58 57 60
Supplementation
Malaria case
management 47 52 43 26 74 46 50 50 47 48
(maternal)
ITN/IRS –
Households
163 179 146 76 274 163 163 163 163 163
protected from
malaria
Pentavalent
65 71 58 29 111 64 66 70 62 65
vaccine
Pneumococcal
104 114 93 46 178 101 108 118 97 105
vaccine
Measles vaccine 31 34 28 14 53 30 33 36 29 31
Oral rehydration
81 89 73 36 139 79 84 82 80 82
solution
Diarrheal disease
78 78 78 35 134 76 82 80 78 80
management
Oral antibiotics
67 74 60 30 115 65 70 69 67 69
for pneumonia
page-986
Pre- and
Total Service
in-service
Building cost service HEW salary provided for
training life
users each user
years
10% decrease
20% decrease
10% increase
10% decrease
10% decrease
20% increase
10% increase
10% increase
cost 10%
decrease
cost 10%
increase
increase
5 year
42 42 43 43 43 44 42 42 43 39 47
56 56 59 59 56 61 54 57 58 52 63
47 47 48 48 47 49 46 47 48 43 52
163 163 163 163 163 163 163 163 163 146 179
64 65 65 65 65 65 64 65 65 58 71
103 103 104 104 103 105 103 104 104 93 114
31 31 31 31 31 32 30 31 31 28 34
80 80 82 82 80 83 79 78 83 73 89
77 77 79 80 77 81 76 76 81 70 86
66 67 68 68 66 69 66 66 69 60 74
page- 987
Effect. Cost Equipment Building
Cost
discount discount life years life years
Baseline ICER
Intervention
10% increase
10% decease
6% discount
no discount
decrease
3 year
5 year
7 year
6.00%
1.00%
ACTs 87 96 78 39 149 84 91 91 85 89
Family planning
295 325 266 160 466 291 302 302 293 298
– child impact
Tuberculosis 114 125 102 114 159 109 122 119 112 118
Total 81 87 74 37 137 79 84 86 78 82
Malaria case
81 89 73 37 137 79 85 85 79 82
management
Diarrheal disease
79 83 75 36 136 77 83 81 79 81
management
Antenatal care 47 52 42 22 79 45 51 55 44 49
Abbreviations: HEW, Health Extension Worker; ITN, insecticide-treated bed net; IRS, indoor residual
spray
page-988
Pre- and
Total Service
in-service
Building cost service HEW salary provided for
training life
users each user
years
10% decrease
20% decrease
10% increase
10% decrease
10% decrease
20% increase
10% increase
10% increase
cost 10%
decrease
cost 10%
increase
increase
5 year
86 86 88 89 86 90 85 86 88 78 96
294 294 297 297 294 300 291 294 296 266 325
111 112 116 115 113 117 111 112 116 102 125
80 80 81 81 80 82 79 80 81 72 89
80 80 82 83 80 84 79 80 82 73 89
79 79 80 81 78 82 77 77 82 71 87
46 46 48 48 47 49 46 47 48 42 52
page- 989
page-990
page- 991
MERQ Consultancy PLC has received funding for the National Assessment of the
Ethiopian Health Extension program from the Bill & Melinda Gates Foundation.

HEP Assessment Master Report Final

Uploaded by

Copyright:

Available Formats

HEP Assessment Master Report Final

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HEP Assessment Master Report Final

Uploaded by

Copyright:

Available Formats

National

Principal Dr. Alula Meressa Teklu (MD, MPH)

Technical Team Contributors to the overall assessment

Dr. Girmay Medhin, Fasil Walelign Fentaye, Tegene

Contributors to supplementary studies

Urban HEP Assessment

Assessment of the quality of Health Extension

Attrition among Health Extension Workers

Cost-Effectiveness Analysis of the Health

Analyses of HEP financing landscape

Qualitative data coding and synthesis

Database Management and Quality Assurance

Reviewers Dr. Nejmudin Kedir Billal

Programs and Frehiwot Bekele, Daniel Tadesse, Mame Gurumu

Document Dimitri and Brehanie

We envision to be a pragmatic choice for research and scientific inquiries in

Please contact us at: info@merqconsultancy.com or visit us at www.

1 MERQ Consultancy PLC, Ethiopia

The Bill and Melinda Gates Foundation’s willingness to respond to the

Our appreciation also extends to the regional health bureaus of Tigray,

Alula M. Teklu, MD, MPH

The HEP is a community-based health program that was initially designed

The program has been acknowledged as a major contributor to recent

A national assessment of the agrarian, pastoralist, and urban HEP was

The assessment had 3 components: collection and analysis of primary

In addition to the primary assessment, in-depth studies were also

Communicable diseases and maternal, neonatal, and nutritional

The current service delivery arrangement of the HEP through home-

INPUTS OF THE HEALTH EXTENSION PROGRAM

Incompetence is a major gap among the current workforce of the HEP.

Facilities and infrastructure

Financing of the HEP

HEALTH SERVICE DELIVERY THROUGH THE HEP

Information system and M&E

Coverage of HEP-related services

Disease prevention and control: Half of the households in malarious

HEP service delivery outcomes

There is a limited multi-sectoral approach to the implementation of the

Conclusion and recommendations

The HEP has been a major contributor to Ethiopia’s recent gains

The program is currently exhibiting sub-optimal quality and coverage as

The comprehensive optimization of the HEP is necessary to improve

Changes in service packages should be guided by long-term plans

Service delivery modalities should also be expanded to reach all segments

Given the vital role expected to be played by community volunteers, it is

Executive Summary --------------------------------------------------------------------------VII

PART 1: CONTEXT AND OBJECTIVES OF THE ASSESSMENT -------9

1 Context of the Ethiopian Health System ----------------------9

1.1 Health status -------------------------------------------------9

1.2 Health policy --------------------------------------------10

1.3 Organization of the health system ----------------------12

1.4 The Health Extension Program ----------------------13

1.5 Simplified logic model of the

2 Objectives of the Assessment -----------------------------------27

2.1 General objective -------------------------------------------27

2.2 Specific objectives --------------------------------------------27

SECTION I: METHODS OF THE RURAL HEALTH EXTENSION

1 General Approach -------------------------------------29

1.1 Study area and study period -------------------------29