Byskov Et Al., 2014

Byskov et al.
Health Research Policy and Systems 2014, 12:49

http://www.health-policy-systems.com/content/12/1/49
RESEARCH Open Access
The accountability for reasonableness approach

to guide priority setting in health systems within
limited resources – findings from action research
at district level in Kenya, Tanzania, and Zambia
Jens Byskov1*, Bruno Marchal2, Stephen Maluka3, Joseph M Zulu4, Salome A Bukachi5, Anna-Karin Hurtig6,
Astrid Blystad7, Peter Kamuzora3, Charles Michelo4, Lillian N Nyandieka8, Benedict Ndawi9, Paul Bloch10,
Øystein E Olsen11 and the REACT Consortium
Abstract
Background: Priority-setting decisions are based on an important, but not sufficient set of values and thus lead to
disagreement on priorities. Accountability for Reasonableness (AFR) is an ethics-based approach to a legitimate and
fair priority-setting process that builds upon four conditions: relevance, publicity, appeals, and enforcement, which
facilitate agreement on priority-setting decisions and gain support for their implementation. This paper focuses on
the assessment of AFR within the project REsponse to ACcountable priority setting for Trust in health systems
(REACT).
Methods: This intervention study applied an action research methodology to assess implementation of AFR in one
district in Kenya, Tanzania, and Zambia, respectively. The assessments focused on selected disease, program, and
managerial areas. An implementing action research team of core health team members and supporting researchers
was formed to implement, and continually assess and improve the application of the four conditions. Researchers
evaluated the intervention using qualitative and quantitative data collection and analysis methods.
Results: The values underlying the AFR approach were in all three districts well-aligned with general values expressed
by both service providers and community representatives. There was some variation in the interpretations and actual
use of the AFR in the decision-making processes in the three districts, and its effect ranged from an increase in
awareness of the importance of fairness to a broadened engagement of health team members and other stakeholders
in priority setting and other decision-making processes.
Conclusions: District stakeholders were able to take greater charge of closing the gap between nationally set planning
and the local realities and demands of the served communities within the limited resources at hand. This study thus
indicates that the operationalization of the four broadly defined and linked conditions is both possible and seems to
be responding to an actual demand. This provides arguments for the continued application and further assessment of
the potential of AFR in supporting priority-setting and other decision-making processes in health systems to achieve
better agreed and more sustainable health improvements linked to a mutual democratic learning with potential wider
implications.
Keywords: Accountability for reasonableness, Priority setting, Fairness, Decentralization, Decision making,
Democratization, Health systems, Kenya, Tanzania, Zambia
* Correspondence: jby@sund.ku.dk
1
DBL – Centre for Health Research and Development, Faculty of Health and
Medical Sciences, University of Copenhagen, Thorvaldsensvej 57, DK 1871
Frederiksberg, Denmark
Full list of author information is available at the end of the article
© 2014 Byskov et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited.
Byskov et al. Health Research Policy and Systems 2014, 12:49 Page 2 of 19
Background setting decisions and the reasons behind them are trans-
Priority setting can be defined as the distribution of re- parent and are made public. This can be done, for in-
sources among competing interests, programs, or people, stance, through open meetings, diffusion of meeting
and is one of the most prominent health care policy ques- agenda and minutes, and other communication processes.
tions [1], not least when resources are very scarce. Most Appeals and revision require in that stakeholders are given
efforts to strengthen district level priority setting in poorer an opportunity to appeal against decisions, propose revi-
countries have been using technical approaches based on sions, and receive a reasoned response. This would mean
burden of disease measures, cost effectiveness analysis, that people affected by the decision have a voice and are
capacity considerations, and other measures that claim to effectively heard, and that a procedure for revision is en-
be based on evidence. Such approaches are typically domi- sured. Enforcement must ensure that the first three cri-
nated by ‘experts’ [2,3]. However, central in priority- teria are adhered to. This final condition is commonly
setting decisions are values, which in practice are rarely referred to as leadership (of the AFR process), as arrange-
adequately recognized, made explicit, defined, discussed, ments must be made to ensure that there is one or more
and agreed upon. While technical approaches in priority legitimate bodies able to ensure procedures for continuous
setting may be useful and necessary, they have been ques- application of all four conditions among the stakeholders
tioned. First, they are based on health and economic data including the public. Improving fairness and legitimacy
and do not take social, cultural, and other values into con- also constitutes a democratic learning process within
sideration. Moreover, they have not led to the intended health systems [7-10].
sustainable improvements in addressing health needs and AFR has attracted attention among decision makers,
demands [4,5]. health care professionals, and scholars involved in em-
Making the values that underlie decisions in the pirical studies of priority setting. Applying the AFR prin-
priority-setting process explicit is important since actors ciples is not an easy and straightforward process, but in
tend to have diverging values; even the ways in which recent years, this framework has nonetheless been tested
stakeholders relate to commonly applied values such as in a number of studies in Canada, Norway, United
‘efficiency’, ‘equity’, and ‘quality’ may conflict [6]. Such con- Kingdom, and elsewhere. Results have shown that decision
flicts can only be managed meaningfully and productively makers in health care organizations have found it a useful
through open discussion. Decision-making approaches approach [11-14]. However, there has been little empirical
that do not permit discussion and choice on the basis of research of its application at the district level in low- and
relevant values tend to produce disagreement, a low sense middle-income countries. Few studies have documented
of ownership, and controversy around both the desired the challenges that the approach faces in practice. The first
outcomes and the allocation of resources. It is essential studies that have reported on the application of AFR in
that values are made explicit because they do influence low-income countries have focused on single organiza-
preferences among all concerned on health improvement tions [15,16], or on assessments of priority-setting prac-
and service options. tices across countries [17,18].
When agreement on desired priority outcomes is diffi- There has been raised criticism of AFR conditions as
cult to achieve, a mechanism of structured discussion and not being adequate to ensure that the decision-making
debate that contributes to legitimize the decisions made is process will be fair, reasonable, and legitimate, as well as
necessary. Accountability for reasonableness (AFR) is an accepted by those of a different opinion or those adversely
ethical framework for priority setting that aims at ensuring affected. Friedman has suggested, however, that profound
that the process towards setting priorities is fair, and that popular involvement and establishing criteria for avoiding
the actually decided-upon priorities are based on reasons a priori exclusion of some values could imply an improve-
that are communicated to all relevant parties involved. ment in a priority-setting and decision-making context
AFR thus provides decision makers with an approach to [19]. Resource managers may find that policy-related,
consider and jointly discuss competing values in the generally-desired, non-health effects may be more import-
priority-setting process. According to AFR, a process for ant than disease or program-specific health effects [20].
setting priorities is legitimate and fair if it meets four con- Hence, there are divergent opinions as to whether general
ditions. Relevance requires that decisions are founded in health-related or program specific and managerial argu-
the values of all concerned and considered important. In ments should be given more weight in the priority-setting
practice, this means that all relevant stakeholders have the process. A recent review illustrates the same dilemmas
chance to participate in the process, that there is respect [21]. Others again have argued that AFR and a more tech-
for differing views, and space to consider divergent opin- nical priority-setting approach may be mutually support-
ions and preferences. The debates must be based on clear ive [22]. It has been pointed out that power differences
arguments, and all actors involved must be given the can be seen as constraining the compliance with the four
chance to have a voice. Publicity demands that priority- AFR conditions, and a fifth condition of empowerment
[23] was proposed. However, this suggestion has not been countries have been published focusing on institutional
fully incorporated in the AFR approach. childbirth that well illustrate also the chosen service as-
Based on existing evidence it seemed fruitful to assess sessment indicators for quality, equity, and trust in the
the AFR approach in district health systems in African three countries [28]. We further refer to a number of
resource-poor settings with the aim to enhance existing papers that present more thematic results from other
knowledge about the relevance and usefulness of the baseline studies. They focus on HIV/AIDS and condom
AFR concept as well as about the implementation availability in all three counties [29], on malaria in
process and potential outcomes from diverse contexts. Zambia [30], on emergency obstetric care in Kenya [31],
The study “REsponse to ACcountable priority setting for on institutional childbirth in all three countries [28], on
Trust in health systems” (REACT) commenced in 2006 voluntary counseling and testing (VCT) for HIV infec-
through funding from the EU (under FP6 contact PL tion in all three countries in Additional file 1 and [32],
517709). The overall objective of the project was to on perceptions on fair decision-making and limitations
strengthen the legitimacy and fairness of priority-setting to fair decision-making processes in Tanzania [33].
processes in Tanzania, Kenya, and Zambia. The baseline studies found that involvement of com-
We applied the AFR concept to decision-making pro- munities and other relevant stakeholders featured prom-
cesses at the district level in the three countries. More inently in the official policies, but that the existing
specifically, we sought to introduce the AFR approach in mechanisms and processes for decision-making at dis-
order to assess potential changes in terms of participa- trict level did have substantial shortcomings in terms of
tion in priority-setting processes and potential influence participation in actual practice. In none of the three
of the approach on the district management, health study sites were the conditions specified by the AFR ap-
workforce management, and, eventually, provision of proach for fair decision-making and priority-setting pro-
services with a particular focus on the field of HIV/ cesses ensured [25-27]. These studies confirmed and
AIDS, malaria, emergency obstetric care, and general detailed shortcomings in the respective services which
care. seemed related to poor horizontal consensus building at
For a detailed presentation and discussion of the re- service level, and thus indicated marked gaps in the
search objectives, design, and methods, we refer to a practice of AFR conditions at all levels.
previous publication in this journal [24]. The research In this paper, we do not venture into a full discussion
study design is shown in Figure 1. of the findings from the baseline studies, nor of the
Data and analyses for gaps in AFR conditions at base- broader priority-setting debate, but we set out to iden-
line have been published for each of the three countries tify strengths and weaknesses of the AFR application as
[25-27]. The methodology, detailed data description, and further developed by the REACT project with the aim
core results of the population-based surveys in the three of adding to the debate on priority-setting processes
THE ACTION REACT
EVALUATION RESEARCH
Describe
Evaluation domains
AFR process
ACTION Management capability

Improve
RESEARCH
Human Resources
HIV/AIDS
Malaria
Evaluate
Emergency obstetric care
Generalized care
Three stage research evaluation of AFR:
1. baseline situation, 2. processes and changes, 3. consequence for quality, equity, trust - and health
Figure 1 The REACT study design.

within health systems and on provider and user Furthermore, the districts were selected as comparable
accountability. across the three countries. Being found ‘typical’ and ‘com-
The paper, thus, consists of four main parts, being first parable’ nonetheless recognizes substantial diversity be-
the references for overall AFR-based methodology and tween the districts for example pertaining to the degree
the wealth of methodological and situational details in and type of decentralization in each country. The three
published baseline papers [25-33], second the further de- countries do, however, have a similar multi-level structure
tails of the applied REACT methods and the analysis (Table 1) and have comparable decision-making structures
and the discussion of data from the final evaluation sur- and procedures at district health teams. All three study
vey, third a broadening of the discussions in relation to districts are mainly rural and include one or more urban
results from already published other thematically fo- or semi-urban centers. They have a comparable burden of
cused REACT process results papers, and fourth a dis- disease, while their populations are all larger than what
cussion of implications for current health development the WHO recommends for a health district, ranging from
policies and strategies. Conclusions finalize the paper. 241,000 to 342,000 inhabitants. A summary comparison
The subsequent appendix provides the latest version of a of the districts is shown in Table 1. Details for each
practical user guide of for application of AFR and the country are further documented in the country-specific
Additional file 2 for the REACT consortium, lists the re- general AFR gap analysis publications [25-28].
searchers who have been actively participating in
REACT as authors of one or more of the REACT publi- The baseline
cations referred to or as important contributors to pro- The baseline studies have been published as referenced.
ject results. They covered two components. The quantitative compo-
nent community-level survey administered to a represen-
Methods tative sample of 2,000 persons in each study district and
In the REACT project, we applied action research meth- the qualitative component that documented the existing
odology to assess preconditions, processes, and outcomes decision-making processes and management practices
of the application of the AFR approach in three study dis- concerning services, programs, and the health workforce,
tricts in Kenya, Tanzania, and Zambia, respectively. and secondly identified the informants’ perceptions and
definitions of ‘fairness’ of decision-making processes. The
The cases qualitative components based on documentary reviews,
We chose the districts as cases and specifically focused on in-depth interviews, focus group discussions, and observa-
decision making at the district level and in units within tions as sources of data. Informants included representa-
which district health management teams play a central tives of the community, health workers of the first line
role. The study sites included Mbarali District in Tanzania, health services, and the district hospital, health service
Malindi District in Kenya, and Kapiri Mposhi District in managers, and district officials. These studies form the
Zambia. These districts were purposively chosen because points of reference for the discussion and assessments of
they have structures, processes, and actor configurations the processes and emerging outcomes from the applica-
that emerge as typical for a district in their country. tion of AFR.
Table 1 The three districts and their relations to country health system structures
Levels Mbarali, Tanzania Malindi, Kenya Kapiri Mposhi, Zambia
International Presence of many organizations and programs – both state and non-state actors
National Ministries for Health and Social Welfare and for MOH (later divided into two) MOH
local government
Subnational Cross-sectorial region, Health Zone (Health Province – moving to a smaller unit Province
sector only) County structure
District Council Health Team District Health Team District Health Team
Council Health Services Board District Health Board District Health Board
District Hospital Board District Hospital Board District Hospital Board
Decentralization to local government, but Decentralization under MOH Decentralization under MOH
professionally overseen by MOH (devolution) (deconcentration). (deconcentration)
First line Health center/clinic and dispensary/health post; MOH, private for profit and not-private for profit.
facility
Community Local structures and committees, (CSOs), NGOs, Local structures and committees, Local structures, neighborhood committees,
users and communities CSOs, NGOs, users and communities CSOs, NGOs, users and communities
CSO, Civil Society Organization; MOH, Ministry of health.
The action research approach either university students or young researchers within or
We used a staged approach to the introduction of the associated with one of the participating university de-
AFR concept in each study district. The action research partments in each study country. They started this work
process started with informing the full District/Council in the study districts in year 3 of the project. This
Health Management Team (DHMT) and the formally in- allowed for more continuity of AFR support, but also for
volved stakeholders about the objectives of the project more in-depth observations of the ongoing process in
and about the content of the AFR principles and condi- the districts.
tions. An ART was then formed in each district, including It should be noted that the project did not provide any
3 to 4 senior members of the DHMT and 1 to 2 re- monetary input to the district or to the governmental
searchers from the national research partner institutions. employees involved with the project at district level, ex-
In a first step, initial ‘sensitization’ sessions solicited cept for provisions for meetings.
stakeholder and community views on AFR and the The project focused on health sector decision making
meanings and values these actors would associate with at district, hospital, and first line facility level. We differ-
its four conditions. Finding in all three sites substantial entiated formal decision-making through planning meet-
concurrence between the AFR approach and the ‘local’ ings, for instance for developing the annual district plan,
values, we proceeded to set up a context-adaptive con- from the routine day-to-day decision making. We define
tinuous action research approach. the changes that were introduced to attitudes and prac-
The ART met on a regular basis to facilitate, monitor, tices of actors and to the organizational culture as the
and guide the use of the AFR conditions in the district- output of the actual AFR process. The study report is
level decision-making processes. To this end, the REACT based on the analysis of the following data sources:
project used a tested cyclical review process, called De-
scribe–Evaluate–Improve (DEI) [11], aimed at stimulating Reports: the ART meeting minutes, DHMT meeting
the ARTs to continually evaluate the application of the minutes, reports and minutes from meetings of joint
AFR conditions and to address potential practical prob- district planning committees
lems that emerged in the process. The DEI cycle was Observation reports
driven by the DHMT members (in Tanzania called council Minutes of the annual round trips by the project
health management team (CHMT)) through processes coordinator.
where the researchers provided insights and advice. Minutes and reports of the joint meetings of the
Guidelines for the DEI cycle were developed in a joint ARTs during the annual REACT project meetings
process by the ARTs of the three countries. The DEI and the general reports of the annual REACT
guidelines were used in the teams’ annual priority-setting project meetings
and planning exercises, but also in daily decision-making Minutes of the monthly REACT Steering
activities. Appendix 1 presents the latest version of the Committee telephone meetings and of its meetings
guideline. Whenever new stakeholder groups or individ- during the annual project meetings
uals were included in the district decision-making process,
they were introduced to AFR and the ongoing AFR-based The second major data set consists of the in-depth in-
process. terviews carried out at the end of the study. These fo-
Beyond assessing potential change related to the AFR cused on the actual application of the AFR conditions,
process, the research also aimed at generating know- on potential changes in views on and practice of deci-
ledge about how AFR can be introduced. The learning sion making. A total of 54 interviews with 18 female and
process was thus also documented by the ARTs. They 36 male informants were carried out between July and
reported during specific AFR research meetings and dur- August 2010. The individuals interviewed during the
ing the annual REACT project meetings where discus- evaluation phase of the study were only in some cases
sions with all the involved researchers (and at times identical with the informants interviewed during the
including senior DMHT members) from all the three baseline survey, but the overall characteristics of the
countries led to the continuous adaptation of the DEI study participants were very similar. The ones included
guidelines. Additionally, the regular coordination round in the evaluation study were 24 members of the DHMT,
trips made by the project coordinator, which aimed at managers and program officers, 11 district and regional
monitoring and supporting country teams, provided op- administrators, planning and human resource officers, 7
portunities to discuss the ART intervention and the DEI peripheral facilities’ staff (including one from a mission
process. A focal person representing the researchers was health center), 4 hospital senior staff (including one from
agreed upon for each country in order to improve the a mission hospital), 4 NGO representatives, and 4 dis-
data collection and AFR advisory support for the district health board members. Table 2 shows the respond-
tricts. The focal persons were selected based on being ent by district.
Table 2 Respondent distribution by study district, organization, and gender

Organization DHMT Higher authority, Manag, Admin Facility Hospital NGO Board
Gender/District M F M F M F M F M F M F
Mbarali 5 2 7 0 3 0 0 1 1 0 2 1
Malindi 4 2 1 0 1 1 0 2 0 2 0 1
Kapiri Mposhi 8 3 3 0 1 1 0 1 0 1 0 0
All 17 7 11 0 5 2 0 4 1 3 2 2
A graphic overview of the REACT project elements, and interview notes were entered in NVIVO7, and rigor-
processes, and emerging output has been presented in ous thematic analysis was carried out by the country re-
several conferences and was recently displayed as part of search teams.
a REACT project parallel session in the Global Sympo-
sium on Health Systems Research in Beijing in 2012, as Ethical issues
shown in Figure 2. Ethical clearance was obtained in the three countries. In
Kenya, scientific and ethical approval was obtained from
Data analysis the Kenya Medical Research Institute and the National
All interviews were recorded, translated into English Ethical Review Committee (KEMRI SSC, number 1096).
where necessary, and transcribed verbatim. The transcripts In Tanzania, research clearance was obtained from the
Figure 2 REACT project overview.

Medical Research Coordinating Committee of the Na- acknowledged as relevant and accommodated into the
tional Institute of Medical Research number (NIMR/ existing set of organizational values. Our analysis
HQ/R.8a/Vol.IX/416), and in Zambia research permis- shows that this was facilitated by the team’s shared
sion was obtained from the University of Zambia Re- values – values ultimately based in the Primary Health
search Ethics Committee (assurance No. FWA00000338, Care (PHC) strategy which had been implemented some
IRB00001131 of IOR0000774). Written informed con- 30 years earlier [34], and which was reflected in their
sent was obtained from all informants. Confidentiality motto “provision of health services in partnership with
and anonymity of the study informants was emphasized the community”. The AFR intervention provided the
and maintained throughout the study. team with an additional stimulus to further apply notions
of participation and transparency in their decision-making
Results practice. Interviewees, indeed, said that the DHMT re-
We present the results as follows: the introduction of AFR vived prior existing consultative mechanisms, such as
in decision-making processes, the application of the AFR health centre committees, as a consequence of the REACT
conditions, and the effects of the introduction of AFR on intervention.
decision making.
“Following the dissolution of the Central Board of
The introduction of AFR in the decision-making processes Health, some committees had stopped functioning and
The AFR concept was introduced in all three study dis- only started operating after the REACT program
tricts in 2006, but the intensity and timing differed be- commenced.” (DHMT member).
tween the sites.
In Mbarali (Tanzania), the preparations for application In all study districts we found that the greatest challenge
of AFR started in 2007. It became clear that repeated was the initial difficulty to grasp and to apply the concept
sensitization sessions with the stakeholders were needed. of AFR in daily practice. Inherent assumptions underlying
This delayed the start of the actual application of the AFR, e.g., the notion of no decisions being absolute right
framework. Furthermore, the ART decided to introduce or wrong, or the potential of individuals’ taking a respon-
the AFR process step-wise, and to gradually build up ex- sibility for choice, seemed to be new to most of those in-
perience and expertise in handling each of the conditions. volved. Study participants noted that in this process the
In practice, this meant that a start was made by the intro- personal attributes of the DHMT members and especially
duction of the relevance condition in the priority-setting of those in management positions had a direct impact on
process in 2008. The publicity and appeals conditions how AFR was adopted and implemented. A comparison
were introduced from 2010 onwards, followed by working of the districts processes is shown in Table 3.
on the full application of the leadership condition in 2011.
In Malindi (Kenya), the AFR concept was introduced to Application of the AFR conditions
the DHMT in May 2008. All conditions of AFR were in- In this section, we make an attempt to present how each
troduced in the decision-making process from 2009 on- of the four conditions of AFR – relevance, publicity, ap-
wards. A high staff turnover at senior level, including the peal and revision, and leadership – were perceived and
District Medical Officer, led to irregular DHMT meetings dealt with in each of the study sites.
during the first years of the project. Further turbulence re-
sulted from the serious post-election social unrest in 2007 Putting the relevance condition into practice
and 2008. The post-election split of the Ministry of Health The relevance condition consists of two main compo-
meant that meeting routines were undermined in the nents: i) involvement of all relevant stakeholders and ii)
Kenyan districts, which denied DHMT the opportunity identification of all values that play out in priority set-
for continuous application and follow-up of the AFR ting. The aim is to ensure that content of what is dis-
process except for one DHMT member and the Hospital cussed in decision- and priority-making processes is
Director, who both remained in post during the project perceived as relevant in the context where it is dis-
and continued their efforts to align their own area of work cussed. Our analysis shows that, within the general DEI
to the conditions. The District Medical Officer who sanc- guidelines, each ART developed its own approach.
tioned the introduction of AFR continued to apply the In Mbarali, Tanzania, the ART introduced the rele-
conditions in the next posting in the regional office. vance condition in 2008. Most members of the CHMT
In Kapiri Mposhi (Zambia), the DHMT was introduced felt that the planning guidelines from the Prime Minis-
to AFR in 2007. The District Director of Health imme- ters’ Office, the Regional Administration and Local Gov-
diately saw the potential for district decision making. ernment and from the Ministry of Health and Social
AFR application started in 2008, increasingly influen- Welfare did not allow for broad stakeholder involvement
cing the DHMT management. All AFR principles were in decision-making processes. This was considered as
Table 3 Comparison of the process of AFR introduction in the three districts

Steps in establishment Mbarali, Tanzania Malindi, Kenya Kapiri Mposhi, Zambia
Sensitization of stakeholders. From 2007 and including From 2008, not going beyond From 2007 including already coopted NGO and
Number of recorded sessions community members; 6 the hospital team; 3 joint increasing to others including representatives of
sessions sessions communities
Effective start of application 2008 2009 2008
Number of ART meetings and Total 18 ART and 4 planning Total 3 ART meetings. Report Monthly meetings for AFR associated with
of planning meetings meetings. Report to CHMT. to DHMT. No record of other DHMT meetings and thus also taken up in plans
referring to AFR Other meetings. meetings.
ART members regularly 2 researchers, 5 CHMT members 1 researcher, 3 DHMT 4 DHMT members. Researcher presence irregular
involved members
Focal person/observer Junior scientist 2009-10 Scientist for irregular periods Only as ad hoc visits by a researcher.
AFR, Accountability for Reasonableness; ART, Action Research Team; CHMT, Council Health Management Team; DHMT, District Health Management Team.
limiting the legitimacy of the decision-making process people could not be invited since there was no provision
made at district level. Drawing upon the AFR relevance for their reimbursement.
condition, the CHMT made the initiative to identify In an effort to identify views from the communities
women and youth representatives, as well as representa- and to respond to the failure to involve more stake-
tives of disadvantaged groups and people living with holders in their meetings, in 2008, the CHMT members
HIV/AIDS, and invited them to the district planning visited villages to collect information that would enable
meetings. Six medical professionals from the hospital them to improve the priorities based on the views from
and from NGOs were also included in the process taking the community.
place in the planning meetings. It was said that this in- In late 2010, the CHMT members revived the idea of
clusion increased the legitimacy of the decisions made involving specific groups from the communities in the
through better representation of diverse stakeholders, planning meetings. The CHMT invited representatives
and broadened the set of values considered in the from the communities; women, youth, elderly, disabled,
priority-setting process. and people living with HIV/AIDS were invited to partici-
Interviewees noted that, with this inclusion, the pate in the preparation of the district annual budget and
CHMT chair did to a lesser extent dominate the health plans. These community representatives were,
decision-making process. To a greater extent the team moreover, trained in participatory planning, priority set-
members made the decisions jointly during the meet- ting, and AFR, and did participate during the 2011/2012
ings. Discussing decisions made in the CHMT manage- planning session.
ment meetings and communicating them onwards to Concerning the second element of relevance – identi-
relevant staff and other actors moreover led to greater fication of core values – our analysis shows that the ma-
transparency. One of the study participants indicated jority of the study participants from the CHMT indicate
that some degree of change in the bureaucratic culture that, prior to the REACT project, CHMT decisions were
occurred during this process. largely based on priorities set by the national ‘essential
health package’ with little consideration of local values.
“Contrary to the past, this time you will find the The interference from the central government and by
agenda in the meetings being discussed with a lot of vertical programs made local priority-setting processes
openness.” (CHMT member) difficult. The analysis shows that the involvement in the
AFR process increased the awareness of CHMT mem-
The analysis of meeting records, however, demonstrates bers of such top down ‘interference’ in what is supposed
that the groups representing the community were actually to be locally driven processes.
not invited to the planning meetings. The reason given by
the CHMT members during the interviews were a per- “Another challenge is that there is frequent
ceived lack of authority to invite new stakeholders to par- interference from the Ministry by giving us directives
ticipate in the planning meetings without the approval and guidelines. This affects our performance and
and funding from the District Executive Director, who is implementation of various health programs. And very
the overall local government in charge of all devoluted often, the guidelines and directives from the Ministry
sectors including health. Another issue brought up during come to the district level very late, and that affects the
the interviews was related to the practice to pay for ‘vol- whole process of preparing district health plans. Not
untary work’; since the REACT project did not reimburse only that, but also the ministerial guidelines always
people for attending committee meetings, it was felt that change, and that leads to spoiling of the whole process
of setting priorities.” (Interview with a member of managers and staff outside the District Health Team cov-
CHMT). ering district health planning, malaria control, clinical ser-
vices, human resources, and mother and child health were
The visits of the CHMT to villages in 2008 demon- to a greater extent involved in coordinated priority-setting
strated to the CHMT that people’s needs, preferences, processes.
and underlying values could be identified through meet-
ings at community level. For instance, in the discussions “The number has grown, … there are people from
during these visits, important issues related to service DHMT, also the hospital staff attend, the nursing
delivery at the district hospital were raised by the vil- sisters from health centers, representatives from
lagers. The knowledge generated during these visits was neighborhood health committees, we are quite many
incorporated in the 2010–2011 Comprehensive Council this time. For a long time, what used to happen was
Health Plan (CCHP). that just a few officers would attend and then they
would come to tell us whatever was discussed.” (Staff,
“The visit conducted in twelve villages enabled the DHMT).
district to solicit useful information. Members
acknowledged that there were some important health The AFR process was moreover extended to the health
needs identified by communities during the visit that centres through the establishment of their own AFR ac-
did not receive enough consideration in the 2009/2010 tion teams.
CCHP.” (Minutes of ART meeting) The study informants indicated that, as a result, deci-
sions were now to a greater extent made collectively and
Identified priorities included the need for construction not merely by one person, and that the new arrange-
of new health facilities, solving problems with procure- ments allowed problems in the community to be dealt
ment of drugs, supplies, and equipment, and shortage of with at the health center level. Interviewees also
health staff. Most importantly, the CHMT seemed to expressed the view that the sense of community owner-
learn that through such a community oriented process ship of the health programs increased.
important information may be obtained from the com-
munities. Analysis of the interviews and documents “I think the greatest thing that we have improved is
moreover indicate that the CHMT subsequently started community involvement, because now, we don’t just sit
using local data more frequently for priority setting and and agree with the district commissioner. We don’t do
in other decision-making contexts. anything without being convinced that, actually, the
As mentioned above, in Malindi (Kenya), the initiation community is happy about it and they have made an
and implementation of AFR was accompanied by a num- input in it.” (DHMT member)
ber of difficulties. Against the background of many
changes in the DHMT during the project period, the Publicity
hospital director remained in post and became central in The second AFR condition is publicity, which implies
the introduction and application of AFR conditions that decision makers are to publicize their priorities and
within the hospital decision-making processes. A DHMT the reasons for their decisions, so that stakeholders, in-
member who remained in the same post during the cluding the public, can understand the values involved
REACT period moreover reported to have adopted the in the choices made, and assess whether the processes
AFR concept in decision-making processes. Both indi- decided upon are actually implemented.
cated that they had observed positive changes in terms The CHMT in Mbarali, Tanzania, initiated efforts to
of broader involvement as the result of their attempts at disseminate district health priorities in 2009 by starting
applying AFR conditions, however limited these would a communication of the CCHP priorities to all its mem-
be in the larger picture. Although the use of the AFR bers and to the district health program managers and
conditions was not involving all DHMT members in district hospital workers. Later, the CHMT translated
Malindi, the concept was broadly perceived as accept- the priorities from English to Kiswahili (the national lan-
able and useful. guage in Tanzania) and disseminated them further
In Kapiri Mposhi (Zambia), increased stakeholder in- through the notice boards of the district hospital, the
volvement was actively pursued after the initiation of the District Council Offices, the Village Council Offices, the
AFR intervention. More mid-level managers were in- Ward Executive Offices, and the health centers and dis-
volved in the decision making and management of the pensaries. The analysis of interviews and documents
health services, including the management team of the indicates that publicizing district health priorities saw a
Rural Health Centres and the district hospital manage- notable change in the management culture of the
ment team. Existing cross cutting teams also including CHMT:
“Although the existing mechanisms for publicity are ad In summary, we found that the strengthening of
hoc, people were impressed with the decision to make the publicity efforts took place at all three study sites,
priorities known to them. For example, some women from and that the increased dissemination of the priority-
Isitu village indicated this when they met one of the ART setting activity was found to be both an acceptable
members. The district hospital workers were also and feasible practice. Our analysis indicates that in all
surprised by the decision to publicize the priorities, which three sites, the AFR approach made people aware that
was not the culture in the previous years.” (CHMT the existing channels of information sharing were
member) inadequate, and that it for various reasons was benefi-
cial to become more interactive and to ensure en-
However, the CHMT members mentioned that unpre- hanced exchange between providers and users.
dictable and late disbursement of funds from the national However, despite the efforts to improve the com-
level was often a barrier to fully implementing the publi- munication channels, the analysis showed that none
city condition, not primarily because this decreased oper- of the districts systematically communicated the rea-
ational funding for dissemination, but because of the sons and criteria that were used to arrive at their de-
uncertainty it induced. cisions. Actual discussion of such criteria occurred
only when knowledgeable staff happened to be avail-
“It is very difficult to publicize priorities in advance, able at the site of posting and thus able to inform
because we are not sure beforehand how much funds staff or other actors.
will come from the Ministry. Since we are not sure of
the amount we will get, we are afraid of announcing Appeal and revision
publicly our priorities to avoid problems and The CHMT in Mbarali, Tanzania, took initiative to
complaints from the citizens after failing to implement develop the appeal and revision mechanism for the
what we announced.” (CHMT member). district. This began with the creation of the appeal
and revision procedures for the district hospital
Additionally, the CHMT members recognized that the workers in early 2010, informing them through a staff
publicity mechanism did not take into account those meeting.
who can neither read nor write.
In Malindi, Kenya, there were also increased efforts to “The CHMT developed an appeal mechanism at
publicize decisions during the project period. The Dis- the district hospital level through which hospital
trict Medical Officer of Health (DMOH) launched a staff would be able to voice their concerns, views,
health service newsletter at the start of the project, an and opinions concerning not only health priorities
initiative later taken up by the provincial health team set during planning processes but also concerning
when the DMOH was promoted to direct its public day-to-day management of the hospital.” (Minutes
health activities. The hospital management team rein- of ART meeting)
forced its posting of decisions and priorities at notice
boards. However, the response to the request for comments
In Kapiri Mposhi, Zambia, the DHMT increased its and appeals relating to district health priorities was
use of existing ways to make decisions and reasons pub- low. Some members reported that out of ten health
lic to the community. This included the use of drama facilities that publicized health priorities, only one re-
groups, neighborhood health committees, traditional ceived appeals from the community.
birth attendants, community meetings, information ses- In Malindi, Kenya, the suggestion box was found to
sions at the clinics, posters, suggestion boxes, and the be a prominent feature in health facilities, but its use
development committees. Increased use was also made remained very limited. However, other improvements
of memos, meetings, and reports to communicate with to communication occurred after repeated introduc-
members of staff. Our analysis shows that these efforts tions of AFR.
contributed to higher inclusion in meetings of represen-
tatives from the churches, traditional leaders, healers, “I find people have kind of become free. People have
community-based organizations, NGOs, and of ordinary been interacting freely. They have been exchanging
community members. their ideas rather freely, starting with the DHMT
members themselves, even to the people in the rural
“As I said, there’s been a lot of publicity and, you health facilities. The interaction to me has been good.
know, program officers are actually coming up and People have really taken into the account what others
everyone will bring their issues and you look at them are saying and there is some consideration of what
now fairly.” (Health center staff member). people are airing.” (DHMT member)
However, we found that in Malindi, no further efforts In Kapiri Mposhi, the District Director of Health was
were undertaken to specifically improve the modalities well established in his function from the onset and
for appeal during the project period. remained in post throughout the project period. The
Informants in Kapiri Mposhi described how appeal analysis shows that there was no formal decision taken
channels existed before the REACT project was initiated, to vest the leadership function in the DHMT, but most
but they qualified that these were neither well defined nor stakeholders appreciated the role played by the leaders
functional. In practice, contested issues, including those at the district level in promoting AFR. The informants
that could be sorted out at local level, were usually re- stated that the leaders were willing to consider or seek
ferred to the Ministry. Analysis of documents and obser- advice from other members of staff. Informants were
vations indicates that over the period of three years of the generally said to have confidence in the leadership of the
initiation of AFR, there was an increase in appeal mecha- DHMT, because it consisted of people who were willing
nisms at all levels and in their actual use by health center to involve all stakeholders, provide guidance, and gener-
staff and community representatives. ally take responsibility in situations of disagreement,
while maintaining inclusiveness in decision-making pro-
“In the old days, once management made a decision, it cesses. The management and personal leadership style of
was final, unlike now: if you make a decision and the the district director was thus noted as important.
decision it’s not favored by the people, they come and
they try to appeal, then you revisit or probably change “Apart from REACT, I think that good leadership skills
it. This has been a result of REACT.” (DHMT and attributes in the District have contributed to
member) improved fair priority-setting processes. REACT has
just supplemented the efforts”. (Member of provincial
In summary, the analysis showed that the informants health team)
considered the appeal and revision mechanism to be
very important. In all three districts it was felt that to In several instances where decisions of the DHMT
allow the members of the planning team to establish a were misunderstood, the appeal procedure clarified the
fair process for change to contested priority-setting deci- issue, confirmed it, or led to revision. This change was
sions, would enable them to also defend such change to mainly attributed to applying the AFR conditions, but
the higher authorities as well reasoned commitments also to the management approach of the District Med-
made to the locally involved other stakeholders and to ical Officer of Health in Kapiri Mposhi. It was felt that
the communities. his managerial abilities and competencies played a major
role in opening up channels of appeal in the district.
Leadership and enforcement In summary, effective leadership was considered by
The AFR approach to priority setting requires that public the interviewees in all three study districts as a pre-
or voluntary regulation of the decision process is put in requisite for enforcement of all rules and regulations, as
place to ensure that the relevance, publicity, and appeal well as the practices required by AFR. The DHMT as-
and revision conditions are met. sumed the enforcement function by default in all cases.
In Mbarali, Tanzania, the CHMT ensured enforcement. There was, however, little evidence of concrete mecha-
The District Medical Officer played a key role in this nisms or aims for public ‘take over’ as guarantor for
process. He was present during the whole project. When AFR.
he was transferred towards the end of the project, his suc-
cessor was introduced to the ongoing AFR process during Outcomes of the introduction of AFR-guided
the transition period. The incorporation of new members decision-making processes
in the decision-making team, primarily program managers In this section, we present the findings regarding changes
and hospital staff, was said to increase the legitimacy of that seem to be related to the implementation of AFR on
the CHMT as the leader for AFR. decision making where we focus on the type of decisions,
In Malindi, the leadership condition was hardly ful- levels addressed, and changes it seems to have triggered
filled. To a large degree, this was due to frequent (see Methodology). Although we are strongly aware of the
changes in personnel in the DHMT. During the imple- shortcomings of the intervention, and we recognize that
mentation period, there were three transfers of DMOH through this research endeavor we cannot attribute changes
as well as of two other officers in the ART. Each transfer in service output or outcome directly to AFR, we argue that
resulted in two to three months of waiting for a replace- certain changes in attitudes and approaches to decision-
ment, and next waiting for the new entrants to settle making processes seem to be strongly associated with the
into required office routines before their involvement REACT project’s introduction of the AFR conditions. In
with the AFR process. Mbarali, Tanzania, as presented above, AFR conditions
were introduced sequentially. This process was mostly ap- thus be seen as examples of managerial outcomes from
plied in the formal planning cycle meetings, and was re- the practice of AFR:
ported to have introduced a more inclusive and transparent
routine for making decisions, including stimulating the “With the REACT program, you have to analyze:
CHMT to actively identify priorities at the village level. An- ‘Surely, if we are to do this, what benefit does it have
other practical example was the issue of more speedy pro- to the community or how is it going to help improve
cesses of decision making due to AFR mentioned by several maybe the, the health status of the community?’ […]
health staff: AFR talks of priority setting. Even when the resources
are limited, we are able to see what could be done
“Another change which I have noted is washing with the same limited resources that we have, and as I
machine. We had no washing machine at this mentioned that where sometimes we have completely
hospital. Medical attendants used to wash clothes. We failed to see what we can do. That is how we came to
repeatedly asked for a washing machine. This issue the stakeholders and see if they can come to our aid.”
was included in the last year’s comprehensive council (District Health Officer)
health plan.” (Senior hospital staff ).
A number of examples showed that AFR can usefully
In Malindi, the AFR intervention was introduced in far support districts in dealing with divergent interpreta-
more difficult circumstances, and had less chance to take tions and can prevent crises by resolving conflicts.
off and, not surprisingly, the document review and analysis The analysis identified several critical incidents that
of the interviews indicate that the effects of the AFR inter- show how, in Kapiri Mposhi, the DHMT was stimulated
vention were very limited. When applied at all, it was dur- on the basis of AFR principles to engage with other
ing the formal planning meetings, but the continuity stakeholders, and facilitated the identification and dis-
among key staff was limited, thus reducing its potential. On cussion of local priorities. For instance, when a major
the personal initiative of the hospital director, the AFR prin- NGO, which had scaled up an anti-retroviral treatment
ciples were introduced in decision-making processes at the program without formal coordination with the DHMT
hospital, but its implications did not seem to trickle down plans suddenly stopped its activities in the district, the
to the first-line services. The interventions seemed to be DHMT acted upon it, realizing that the sudden stop
too unsystematic and lack the necessary follow-up to have would lead to a major gap in the service delivery. The
substantial impact upon attitudes and organizational cul- DHMT sought contact with other NGOs, which jointly
ture, even if, as we have seen, the approach resulted in with the DHMT maintained the service delivery through
change of decision making practice by individual members a broader than usual redistribution of tasks, responsibil-
in the DHMT and the hospital. ities and, thus, resource use between government and
The AFR approach was most comprehensively imple- Civil Society Organization/NGO services.
mented in Kapiri Mposhi. Here, the interviewees explicitly Another example was presented describing how the
describe that priority-setting and other decision-making DHMT found that several members of teams doing in-
processes had changed as a result of the introduction of secticide spraying to control malaria had not performed
the AFR process. The AFR approach thus seems to have well. The DHMT identified unclear selection criteria
assisted the DHMT in better exploiting their actual deci- and procedures as the cause, and in response developed
sion spaces. transparent selection criteria and engaged other stake-
holders in overseeing the recruitment process of the
“What has changed, as I said, is [that] we are able now sprayers. This process ensured that all applicants (and
to sit down as a team and make a decision together at their communities) accepted the employment decision.
every level, which was not the case before. […] It has As a result, the teams became motivated and team per-
been gradual, actually, in the last three years. It’s been formance improved.
gradual and even ourselves, we have just found that we
can’t do certain things without involving as many people “We decided to be open to external stakeholders to
as possible.” (District ART member) help in the recruitment of spray providers. This helped
getting the right people and, subsequently, in providing
Such noted changes are not attributable to particular quality services to a lot of people.” (District ART
AFR conditions, but are examples of improved manage- member)
ment that, according to the DHMT itself, came about
because of the practice of AFR. One might see those ex- A summary assessment of the AFR implementation
amples as a first outcome of the broader leadership con- process, focusing on the uptake and use of the AFR con-
cept inherent in the AFR process. The following may cept, is presented in Table 4.
Table 4 Summary assessment of the AFR implementation

What worked well Challenges
AFR was regarded as a concrete and workable approach to strengthen AFR principles of legitimacy and fairness as supported by the
the influence of values and context on decision making conditionschange ways of thinking and acting which is only consolidated
after a relatively long joint practice
The AFR conditions were accepted as process guidance for use of Stakeholders, including communities, were used to be included in
criteria for priority setting decision-making processes on an ad hoc basis, and had some trouble see-
ing AFR as a change from ‘business as usual’
AFR increased the stakeholder and public understanding of their Action research methods were not well recognized by all involved
opportunities to influence local health action researchers and their institutions to be as valid as other research
The AFR process guidance facilitated the coordination between current
decision makers and expanded their inclusion of others in support of
the implementation of national policies in local contexts
AFR conditions influenced priority setting and other decisions in some
of the sites
Elements facilitating the application of AFR Elements constraining the application of AFR
Fairness and other AFR-related values of transparency, accountability, Concerns for managerial consequences and risks to existing agendas and
and equity were already recognized as desirable aims by respondents power relations were likely to be the reason for a limited national and
donor interest in the approach
AFR principles of inclusiveness and accountability corresponded well The lack of focus on predetermined outcomes may not have been seen as
with existing policy guidelines and planning aims a procedural support, but rather as a challenge to the strong international
and national priority setting and programming
Formal structures in place for boards and committees Limited organizational, leadership, communication, and advocacy skills
may have been among reasons for poor stakeholder and public awareness
The action research approach with continuous researcher support of options for health action
bridged the research into practice gap for AFR from the onset
Limitations governments from roughly the same group of aid agen-

This study had a number of limitations. First, as already cies. This similarity is reflected in the results of the
mentioned, the actual project duration was too short to baseline studies from all three countries, which do not
demonstrate effects in terms of changes in the ultimate differ substantially in the identified gaps and expressed
outcomes of AFR for quality, equity, and trust and for concerns.
health outcomes. The late start of the application of Fourth, the private ‘for profit’ sector was not specific-
AFR was due to the need to establish a shared interpret- ally targeted, although if private actors are stakeholders,
ation of the AFR approach, and of the consequences of they would be ‘captured’ by the AFR approach.
its introduction in the study sites as well as for the quali- A final limitation arises from the focus on the district
tative research to complete the baseline study. health team as the main unit of analysis for change of
Second, there is a risk that interviewees might have decision making processes. We chose the DHMT be-
expressed what they believed the interviewers wanted to cause we consider it as an important hub at the level of
hear, rather than their experience with the intervention local health systems, where constructive engagement of
(social desirability). We attempted to reduce this risk by users and their communities as well as organizational
carefully formulating open ended questions in interview management of health services occurs.
guides, and through checking concurrence of statements
with process progress, records, observations, and con- Discussion
crete examples. Because the project did not introduce This section further summarizes and reflects on the as-
any new resources, the risk of social desirability bias sessment of AFR in the REACT project and on oppor-
seemed to have been reduced. tunities for continued practice. In general, the principles
Third, research results from studies in several coun- of AFR were deemed relevant and useful by the district
tries, which differ in their organizational and managerial level decision makers in all three study settings. While it
set-up at district level, must be interpreted with caution. took time to reach a shared understanding of the terms,
However, our purposive selection of countries and dis- it was also clear that fair decision making is considered
tricts allowed us to have three broadly similar settings. to be important by district level stakeholders.
The three study countries have differently organized Strengthening specific capacities including leadership
health systems, but priority setting, planning, and man- and communication skills of managers, as well as know-
agement values and approaches were all heavily influ- ledge and involvement of representatives on the demand
enced by decades of support to the health sector and to side emerge as important for the implementation and
consolidation of AFR. These observations concur with the REACT baseline publications [25-28]. In such cases,
those made in a realist evaluation of AFR in Tanzania [35]. the AFR conditions were found to provide an alternative
We also learned that it is important to fully involve user procedure for coping with unexpected, but in reality ex-
and community representatives as participants in the ac- pected change.
tion research teams rather than mainly as informants. A Involving the stakeholders – both health and non-health
more standardized and controlled approach, for instance actors and increasingly those beyond the district health
coordinated by a project manager, was considered, but management teams in decision making influenced deci-
was also found to possibly restrict the ‘open ended’ nature sions, shared responsibility for them, and better supported
of the action research. Instead the addition of a focal per- their implementation. Publicizing and sharing information
son in each district did strengthen the advice on AFR and contributes to improved transparency, and thus to in-
improved on observation reports. The arrangement was creased perceptions of fairness which seemed to emerge
most stable in the Tanzanian district. In the other two in a number of the examples seen in the study. The uni-
study districts the focal person was changed once or twice versalistic application of specific principles – as opposed
or was not as regularly available to the district team. to managerial decisions based on personal relations – has
This study seems to add to earlier findings suggesting been shown to lead to higher levels of trust and commit-
that the AFR concept provides a governance tool that ment of staff [36,37]. The importance of procedural justice
contributes to balancing major opposed objectives such for trust in decision making has also been shown [38].
as improvement of population health and well-being on The involvement of stakeholders for increased mutual
the one hand, and providing the highest possible quality trust is located at the core of AFR.
response to the illness of any individual on the other. The capacity and stability of the district health leadership
This can be said with reference to the concept that AFR was found to be particularly important for the enforcement
contributes to identifying priorities, and attempts to en- function that facilitates the AFR conditions. However, the
hance the fairness of the priority-setting process, not AFR leadership function may be better consolidated if dele-
only by addressing technical, managerial, and contextual gated to reasonably permanent and committed staff and
factors, but by allowing the involvement of all relevant possibly other local actors. Optimally, the leadership func-
actors and by structuring their meaningful participation tion is to be adopted and enforced by local actors and com-
in the process. It is the latter principle that gives AFR munities who demand fair decisions and accountability to
the potential to enhance perceived fairness and legitim- themselves. However, to transgress actual power relations
acy of the priority-setting process, in our case, at the dis- (authoritative or technical insight-based) would need ac-
trict level in the chosen study countries. This study ceptance that communities and individuals are the experts
showed that AFR can be applied as an ethical and demo- on their own values, and that technical values only lead to
cratic value-based approach to enhance mutual responsi- sustainable progress if weighed against the social and indi-
bility for health for all. vidual values that facilitate engagement. This is relevant
However, as we have stated, a number of challenges whether the concerned communities and the individuals in-
emerged. It took time to reach a shared understanding of volved are managers, health workers, service users, or com-
the core concepts, perhaps less because of fundamental munity leaders. In an AFR process, many values and other
philosophical differences than operational constraints that contectual factors may remain implicit, but attempts to
made it difficult to see how such a new approach could be tease out such context in each setting may consume all the
implemented. However, sustained support and the use of energy of change, lose sight of potential mutual benefits
the DEI cycle were instrumental to meet these challenges. across groups and knowledge differences, and may prevent
Initially, concerns about the feasibility of involving non- the AFR process from strengthening the legitimacy and
professionals in priority setting were expressed by a few thus relevance of decisions across power divides. Likewise
DHMT members. Discussions with the ARTs addressed the term empowerment has connotation of a battlefield,
this concern, but when practical experiences demonstrated and may in a longer term democratic development process
the feasibility and acceptability of the approach, this initial be better termed as development of capacity for responsive
concern was overcome. We also found that while the exist- leadership, workforce communication skills, community ac-
ing priority-setting, planning, and budgeting mechanisms countability, and joint monitoring.
and policies seemed well suited for formal planning and al- The health sector of a country is one of the most com-
location of funds, they emerged as ineffective in dealing plex sectors and the one that most directly and regularly
with unanticipated changes during implementation, such as affects all individuals as well as being of high awareness,
decreased or delayed resource allocation and inclusion of opinion, and concern of the whole population. Long term
new programs and policy changes. The elaborate district planning as well as daily decisions within actual rather
plans thus became increasingly less realistic for guidance as than expected resource limits are necessary at all levels of
the financial year progressed. This was also documented by the health system. AFR may assist in managing this
complexity, and emerges as highly relevant as it is cur- Implications

rently targeted to the health system. The strategies and methods that the REACT project has
The challenges of implementing AFR, as seen in the employed in its application of the AFR conditions add to
present study, indicate that fairness in priority setting may a vast number of other efforts to support good govern-
be an ideal that can be approached though probably never ance and democratic learning. AFR constitutes a frame-
be fully achieved. However, unfairness can be mitigated work of conditions that in principle are easy to overview
through a continuous effort. The argument is thus related and use in capacity development for democratic practice
to a similar one for good governance and democratic shaped by specific contexts.
practice, which are never fully achieved, but where efforts The implementation of AFR in the three districts dem-
counter emerging undemocratic practices. State level rep- onstrates that stakeholders are concerned with com-
resentative democracy in various forms is recognized as monly shared, seemingly universal values pertaining to
indispensable for stabile development, but more debate is fairness. A broad range of globally shared values, stra-
needed as to why or why not to more strongly apply tegic approaches, broader provider, user, and community
democratic approaches at sub-national levels and in the involvement, were integrated in the PHC strategy. They
management of fields such as health. were much valued by developing countries which gained
The implementation and adaptation of AFR in all three increased influence on their own health systems. Both
countries could be further facilitated by stronger involve- PHC and newer systems approaches have received
ment and support of the provincial or regional offices as recent attention by the WHO [43,44]. These meet new
well as the relevant ministries. It could be part of the efforts at community monitoring and accountability evi-
process through which district administrations and polit- dent through networks such as Equinet (Additional file
ical bodies, health service boards, and facility governing 4) and the Community of Practitioners on Accountabil-
committees hold district health managers accountable for ity and Social Action in Health (COPASAH; Additional
doing what they have agreed to do. Additionally, stake- file 5). However, many of these initiatives have difficul-
holders should be more continually informed, and their ties in being brought to scale within national health
opportunities for appeal and for influencing decisions systems. The importance of complementary roles of all
should be improved. This would make the leadership con- stakeholders to achieve scale up of relevant, acceptable,
dition less dependent on individual managerial leaders and operationally feasible interventions has been empha-
and closer to the broader enforcement condition of the sized in other studies [45]. The procedural guidance
AFR. Two later publications elaborate how the relevance through AFR could be a new way for the national health
and the application of AFR was ensured and led to better sector to become more inclusive of such civil society ini-
understanding of community participation in Tanzania tiatives. AFR might also assist in the necessary coordin-
[39], and on the challenges of involving decision makers ation with a generally fragmented ‘private for profit’
and researchers in action research in Tanzania [40]. health sector.
At the end of the project and its funding, researchers Current global attempts of introducing equitable, Uni-
in Tanzania and Zambia made commitments to continue versal Health Coverage (UHC) [46] could in this context
their involvement and support to study district AFR pro- benefit from harvesting knowledge from this experience.
cesses. In Tanzania, AFR was absorbed into guidance for PHC, as a defined strategy, has been misunderstood as a
national health research priority setting from 2006 [41] focus on the first contact level of services for too long to
see additional file 3. Moreover, since 2009 AFR has been again become a sufficiently unifying term. UHC may be
applied on a pilot basis in four other districts in an answer to this if defined not as a providers’ view to
Tanzania within the area served by the participating health care, but as a comprehensive systems approach bal-
Zonal Resource Centre in Iringa. Training packages for ancing provision by all providers with use and action by
AFR and for supportive capacity-building in this setting all user communities. Introducing AFR-based fair and le-
have been developed and implemented [42]. gitimate priority-setting processes has the potential to
The expansion of a country knowledge base can sup- guide UHC towards a sustainable, long-term impact for
port a continued assessment of AFR. In Kenya, a health all. To this end, a concurrent health systems research and
systems research unit was formed in 2011 by REACT re- capacity development can best be achieved through les-
searchers in one institution. In all study countries, AFR- sons from various forms of action research [47]. Inspir-
associated research is continuing within PhD studies. ation for action based on a focus on people and their
While capacity development and influence of both the preferences exists in several recent debates and in publica-
supply and demand sides are important and should be tions such as When people come first [48].
promoted, the most important at any level would be the AFR is expected to contribute to good governance and
continuous promotion of fairness principles using an democratic learning by guiding priority setting in the
AFR approach linked to decentralization policies. health sector, as this is probably the service sector which
is most directly involving all individuals and is of high Step 2. Clarification of decision-making procedures
importance for their trust in government, private, volun- ■ Decision-makers and stakeholders need to know
tary health systems, and their relation to other sectors and understand
and to national governance. The AFR conditions are not a) how decisions will be made;
formulated to be specific for the health sector only and b) on what basis will these decisions be
could, if agreed, be modeled and tested as a practical made.
approach to strengthen other sectors and multilevel in
country democratization. Step 3. Start the dialogue
Stakeholders now can start the discussion focusing on
identification of priorities.
Conclusions During the dialogue, the following issues should be
The AFR concept was accepted as a tool for improved considered:
fairness and legitimacy of priority setting in health care ■ Provide a statement of rationale for each
in the three study districts. Although differences be- decision;
tween the study districts were observed, implementing ■ The discussion should focus on values of the
stakeholders took with the implementation of AFR stakeholders involved in the decision-making
greater charge of closing the gap between nationally set process;
planning and the local realities and demands of the ■ To ensure the reliability, validity, and completeness
served communities within the resources actually re- of the data made available and presented in priority
ceived or possible to mobilize locally. Participation and setting;
contextual relevance of decisions and transparency of ■ To capture values of each stakeholder;
the priority-setting process contributed to the improve- ■ Criteria should be developed for each identified
ments of decision making beyond priority setting. priority.
Supportive policies, leadership capacity, and commitment
are key determinants for effective uptake of AFR. Capacity Step 4. Consensus building and decision making
development for using the AFR conditions in decision mak- ■ Develop list of ranked priorities;
ing, including the associated communication and leadership ■ Rating and scoring mechanism may assist to
skills, can accelerate their application. The study results guide the criteria to reach the consensus;
imply that AFR can be applied to health systems and pos- ■ Provide reasons for each identified priority.
sibly also to other service-providing organizations and so-
cial systems. Further research should, however, be of at Step 5. Review the stakeholder’s involvement
least 5 years duration, be based on additional adaptation of
the guidance, facilitation in specific settings, on support 2. Publicizing the priorities
from higher levels, and on integration of the monitoring of It is important to disseminate the decisions on the prior-
AFR with routine service monitoring. ities identified and their reasons in order to legitimate
them to the public. Publicizing of the priorities can be con-
ducted through the following steps:
Appendix ■ Decide who is the target group;
Application of accountability for reasonableness (AFR) ■ Identify the ways/methods by which the decided
Guidance for the Action Research Team (ART) priorities will be disseminated (e.g., letters, notice
boards, meetings, newsletters, etc.);
1. Relevance ■ Disseminate the priorities and their reasons.
In the relevance condition, the goal is to make reason-
able decisions that are inclusive, transparent, and fair.
The following steps should be followed in applying this 3. Appeals/revision
condition. Some things to consider:
Step 1. Stakeholder identification ■ Depending on the subject in question, appeals may
Identify your stakeholders and include them in decision- come from different groups including:
making in order to ensure that decision-making includes a 1. health workers;
broad range of ideas and stakeholder perspectives. 2. community, etc.
This can increase gradually from a start of the condi- ■ Appeals can be directed towards health planning
tions within your own team to increasingly involve others priorities or daily management decisions;
aiming at good representation of communities and their ■ Stakeholders should be given enough time to
representatives. appeal and feedback should be given promptly.
Actions to be taken ■ The ART should report to all other health

■ Set up a committee responsible for receiving and management team members on a quarterly basis;
compiling appeals to be submitted to the District ■ The ART should promote AFR in other
Health Management Team (DHMT); meetings as well, e.g., DHMT and others.
■ Set up ways for appeals in different situations for (v)Agenda for conducting ART meetings
stakeholders, e.g., through meetings, letters, face-to- The mandate for the meeting is on the implementa-
face conversation, etc.; tion of the four conditions of AFR
■ DHMT to process and discuss appeals for ■ Opening of the meeting;
decisions on whether to reject, accept, or revise; ■ Approval of the agenda;
■ DHMT to give reasons for accepting or rejecting ■ Confirmation of the previous meeting;
appeals and revising decisions; ■ Matters arising using DEI approach.
■ Set up mechanisms for providing feedback to
stakeholders. Feedback can be provided through 1. Describing
meetings, letters, face-to-face conversation, etc.; Good practices
■ DHMT to ensure that the process of feedback Challenges
provision does not exceed one month from 2. Evaluating
receiving appeals. Lessons learnt
Opportunities for improvement
4. Leadership 3. Improving
To spearhead and drive the process, a team comprised Way forward
of the District Medical Officer (DMO) and another three New action points
members from the Health Management Team should be AOB
formed to lead the application of AFR in the district. Closing of the meeting
(i) The team leader (DMO) should ensure the
following: The DEI process was applied for AFR conditions in ART
■ Meetings are transparent and fair; and in some other meetings:
■ The members have equal rights;
■ Roles and responsibility of the members are 1. Describe
known; Ensure a focus on values as the starting point for dia-
■ There is a mechanism for reaching consensus logue. It is the value base that determines the criteria
during decision-making process. used, the information needed, the priorities set, and, fi-
(ii) The criteria for selecting Action Research Team nally, the decisions made
(ART) members from the district: ■ What decision makers and priority-setting
■ They must understand the application of four members actually do concerning the four condi-
conditions of AFR and the Describe–Evaluate– tions (relevance, publicity, appeals/revision and
Improve (DEI) process; leadership);
■ They must have interest in improving health ■ Review/revise current priorities in relation to
systems through AFR implementation; values and criteria;
■ They must have a mandate to bring change in ■ Align priorities with the values and criteria and
the district; make them public;
■ Committed members; ■ Management to invite and respond to comments
■ Consider issue of gender. and appeals to priorities
(iii)Terms of reference for the ART
■ Create awareness by conducting capacity building 2. Evaluate
on AFR concepts to health staff, community, and How successfully the decision-making process met the
other stakeholders; conditions of ‘Accountability for Reasonableness’. The
■ Facilitate the application, implementation, and gaps between ‘what you do’ and ‘what you should be
sustainability of AFR intervention in the district; doing’ must be identified. To close this gap, you need to
■ Monitor progress of the process in the be able to evaluate your success.
implementation of AFR conditions. ■ Confirm progress in applying AFR conditions for
(iv) Other issues to be considered priority setting;
■ The ART should meet every second month ■ Identify gaps in application of AFR conditions;
with an agenda using a discussion guide, i.e., DEI ■ Opportunities for improvement;
approach; ■ Other lessons learnt.
3. Improve All REACT consortium researchers, as shown in the attached consortium

The decision-making process to make it more ethical, the description, have contributed to make this paper possible through their
continued participation in the overall study.
gaps you identify are areas of improvement for subsequent
iterations of decision-making. Author details
■ Plan and implement change to close the gaps in
1
DBL – Centre for Health Research and Development, Faculty of Health and
Medical Sciences, University of Copenhagen, Thorvaldsensvej 57, DK 1871
AFR during priority setting; Frederiksberg, Denmark. 2Department of Public Health, Institute of Tropical
■ Develop new action points in relation to AFR Medicine, Nationalestraat 155, B 2000 Antwerpen, Belgium. 3Institute of
concept. Development Studies, University of Dar Es Salaam, PO Box 35169, Dar Es
Salaam, Tanzania. 4Department of Public Health, School of Medicine,
University of Zambia, PO Box 50110, Lusaka, Zambia. 5Institute of
Anthropology, Gender and African Studies University of Nairobi, PO Box
Additional files 30197, Nairobi 00100, Kenya. 6Umeå International School of Public Health,
Umeå University, SE 90185 Umea, Sweden. 7Department of Public Health
Additional file 1: Njeru MK. HIV testing services in Kenya, Tanzania and and Primary Health Care, University of Bergen, PO Box 7804, 5020 Bergen,
Zambia: Determinants, experiences and responsiveness. PhD Dissertation. Norway. 8Centre for Public Health Research, Kenya Medical Research Institute
Additional file 2: The REACT Consortium. (KEMRI), PO Box 20752, Nairobi 00202, Kenya. 9Primary Health Care Institute,
PO Box 235, Iringa, Tanzania. 10Steno Health Promotion Center, Steno
Additional file 3: Tanzania National Health Research Priorities, 2006-2010. Diabetes Center, Niels Steensens Vej 8, DK-2820 Gentofte, Denmark.
11
Additional file 4: Equinet. Affiliated to Centre for International Health, University of Bergen,
Additional file 5: Community of Practitioners on Accountability and Årstadveien 21 5th floor, N-5009 Bergen, Norway.
Social Action for Health - COPASAH.
Received: 1 April 2013 Accepted: 3 August 2014
Published: 20 August 2014
Abbreviations
AFR: Accountability for Reasonableness; ART: Action Research Team; References
CHMT: Council Health Management Team; CCHP: Comprehensive Council 1. Martin D, Singer PA: A strategy to improve priority setting in health care
Health Plan; DEI: Describe–Evaluate–Improve; DHMT: District Health institutions. Health Care Anal 2000, 11(1):59–68.
Management Team; DMOH: District Medical Officers of Health; NGO: Non- 2. Murray CJL, Lopez A: Quantifying the burden of disease and injury
Governmental Organization; PHC: Primary Health Care; REACT: REsponse to attributable to ten major risk factors. In The Global Burden of Disease: A
ACccountable priority setting for Trust in health systems; UHC: Universal Comprehensive Assessment of Mortality and Disability from Diseases, Injuries,
health coverage. and Risk Factors in 1990 and Projected to 2020. Edited by Murray CJL, Lopez
AD. Cambridge, MA: Harvard University Press; 1996.
Competing interests 3. Hoedemaekkers R, Dekkers W: Justice and solidarity in priority setting in
The authors declare that they have no competing interests. health care. Health Care Anal 2003, 11(4):325–343.
4. Maundy E, Kapiriri L, Norheim OF: Combining evidence and values in
priority setting: testing the balance sheet method in a low-income
Authors’ contributions
country. BMC Health Serv Res 2007, 7:152.
JB conceived the study and developed its main methods with ØO,
5. Mshana S, Shemilu H, Ndawi B, Momburi R, Olsen OE, Byskov J, Martin DK:
coordinated the study implementation, participated in the intervention, data
What do district health planners in Tanzania think about improving
collection and analysis, and drafted the manuscript. BM contributed to
priority setting using ‘accountability for Reasonableness’? BMC Health
methods development, monitored data use, participated in data analysis,
Serv Res 2007, 7:180.
and provided major contributions to the manuscript. SM participated in data
6. Olsen ØE, Ndeki S, Norheim OF: Human resources for emergency
collection and analysis in Tanzania and contributed to the manuscript. JZ
obstetric care in northern Tanzania: distribution of quantity or quality?
participated in data analysis in Zambia and contributed to the manuscript.
Hum Resour Health 2005, 3:5.
SB participated in data collection and analysis in Kenya and contributed to
7. Daniels N, Sabin J: Limits to health care: fair procedures, democratic
the manuscript. AKH contributed to methods development, managed data,
deliberation, and the legitimacy problem for ensurers. Philosophy Public
and contributed to the manuscript. AB contributed to methods
Affairs 1997, 26(4):303–350.
development, data analysis, and the manuscript. PK facilitated the
8. Daniels N, Sabin J: The ethics of accountability in managed care reform.
intervention, led Tanzanian teams, participated in data analysis, and
Health Affairs (Millwood) 1998, 17:50–64.
contributed to the manuscript. CM facilitated the intervention, led Zambian
9. Daniels N, Sabin J: Setting Limits Fairly: Can we Learn to Share Medical
field teams, participated in data analysis, and contributed to the manuscript.
Resources?. New York: Oxford University Press; 2002.
LN facilitated the intervention, collected data, participated in their analysis,
10. Daniels N: Just Health: Meeting Health Needs Fairly. Cambridge: Cambridge
and contributed to the manuscript. BN facilitated the intervention, collected
University Press; 2008.
process data, assisted in their analysis, and contributed to the manuscript. PB
11. Martin DK, Giacomini M, Singer PA: Fairness, accountability for
assisted in methods development, managed data, and contributed to the
reasonableness, and the views of priority setting decision-makers.
manuscript. ØO conceived the study and developed methods, participated
Health Policy 2002, 61:279–290.
in data analysis, and contributed to the manuscript. All authors read and
12. Martin DK, Reeleder D, Keresztes C, Singer PA: What do hospital decision
approved the final manuscript.
makers in Ontario, Canada, have to say about their fairness of priority
setting in their institutions? BMC Health Serv Res 2005, 5:8.
Acknowledgements 13. Walton NA, Martin DK, Peter EH, Pingle DM, Singer PA: Priority setting and
This study was made possible through funding and support from the cardiac surgery: a qualitative case study. Health Policy 2007, 80(3):444–458.
European Union Sixth Framework Programme (INCO-2003-A.1.2, contract 14. Jansson S: Implementing accountability for reasonableness-the case of
PL517709) for the Specific Targeted Research and Innovation Project pharmaceutical reimbursement in Sweden. Health Economics Policy Law
REACT – REsponse to ACcountable priority setting for Trust in health 2007, 2:153–171.
systems. Full title: Strengthening of fairness and accountability in priority 15. Kapiriri L, Martin DK: Priority setting in developing countries health care
setting for improving equity and access to quality health care at district institutions: the case of a Ugandan hospital. BMC Health Serv Res 2006, 6:127.
level in Tanzania, Kenya and Zambia. The contributions and supportive 16. Kapiriri L, Martin DK: Bedsides rationing by health practitioners in a
environment by the whole REACT consortium, the country authorities and context of extreme resource constraints: the case of Uganda. Med Decis
all the respondents have been essential. Mak 2007, 27:44–52.
17. Johansson KA, Jerene D, Norheim OF: National HIV treatment guidelines 36. Pearce J, Branyiczki I, Bigley G: Insufficient bureaucracy: trust and
in Tanzania and Ethiopia: are they legitimate rationing tools? J Med Ethics commitment in particularistic organisations. Organ Sci 2000, 11:148–162.
2008, 34:478–483. 37. Gould-Williams J: The importance of HR practices and workplace trust in
18. Kapiriri L, Norheim F, Martin D: Priority setting at the micro-, meso- and macro- achieving superior performance: a study of public sector institutions.
levels in Canada, Norway and Uganda. Health Policy 2007, 82(1):78–94. Int J Human Res Manag 2003, 1:28–54.
19. Friedman A: Beyond accountability for reasonableness. Bioethics 2008, 38. Gilson L, Palmer N, Schneider H: Trust and health worker performance:
22:101–112. exploring a conceptual framework using South African evidence. Soc Sci
20. Lippert-Rasmussen K, Lauridsen S: Justice and the allocation of healthcare Med 2005, 61(7):1418–1429.
resources: should indirect, non-health effects count? Med Healthcare 39. Kamuzora P, Maluka S, Ndawi B, Byskov J, Hurtig A-K: Promoting community
Philosophy 2010, 13(3):237–246. participation in priority setting in district health systems: experiences from
21. Hipgrave DB, Alderman KB, Anderson I, Soto EJ: Health sector priority Mbarali district, Tanzania. Glob Health Action 2013, 6:22669.
setting at meso-level in lower and middle income countries: lessons 40. Maluka S, Kamuzora P, Ndawi B, Hurtig A-K: Involving decision-makers in
learned, available options and suggested steps. Soc Sci Med 2014, the research process: challenges of implementing the accountability for
102:190e200. reasonableness approach to priority setting at the district level in
22. Baltussen R, Mikkelsen E, Tromp N, Hurtig A-K, Byskov J, Olsen OE, Bærøe K, Tanzania. Glob Public Health 2014, 9(7):760–772.
Hontelez JA, Singh J, Norheim OF: Balancing efficiency, equity and 41. National Institute for Medical Research: Tanzania Health Research Priorities,
feasibility of HIV treatment in South Africa – development of 2006-2010. Tanzania: NIMR; 2006 [http://www.nimr.or.tz/wp-content/
programmatic guidance. Cost Effect Res Allocation 2013, 11:26. uploads/2013/07/TANZANIA-HEALTH-RESEARCH-PRIORITIES-2006.pdf]
23. Gibson JL, Martin DK, Singer PA: Priority setting in hospitals: fairness, 42. Primary Health Care Institute: AFR Training Packages. Iringa, Tanzania: PHCI;
inclusiveness, and the problem of institutional power differences. Soc Sci 2010.
Med 2005, 61:2355–2362. 43. WHO: The World Health Report 2008: Primary Health Care (Now More Than
24. Byskov J, Bloch P, Blystad A, Hurtig A-K, Fylkesnes K, Kamuzora P, Kombe Y, Ever). Geneva: WHO; 2008:2008.
Kvåle G, Marchal B, Martin DK, Michelo C, Ndawi B, Ngulube TJ, Nyamongo 44. WHO, Alliance for Health Policy and Systems Research: Systems Thinking for
I, Olsen ØE, Onyango-Ouma W, Sandøy IF, Shayo EH, Silwamba G, Songstad Health Systems Strengthening. WHO: Geneva; 2009.
NG, Tuba M: Accountable priority setting for trust in health systems – the 45. Milat AJ, King L, Newson R, Wolfenden L, Rissel C, Bauman A, Redman S:
need for research into a new approach for strengthening sustainable Increasing the scale and adoption of population health interventions:
health action in developing countries. Health Res Policy Systems 2009, 7:7. experiences and perspectives of policy makers, practitioners, and
25. Bukachi SA, Onyango-Ouma W, Siso JM, Nyamongo IK, Mutai JK, Hurtig A-K, researchers. Health Res Policy Systems 2014, 12:18.
Olsen ØE, Byskov J: Healthcare priority setting in Kenya: a gap analysis 46. WHO: The World Health Report: Health Systems Financing: The Path to
applying the accountability for reasonableness framework. Int J Health Universal Coverage. Geneva: WHO; 2010.
Plann Manag 2013, In press. 47. Loewenson R, Flores W, Shukla A, Kagis M, Baba A, Ryklief A, Mbwili-Muleya
26. Zulu JM, Michelo C, Msoni C, Hurtig A-K, Byskov J, Blystad A: Increased C, Kakde D: Raising the profile of participatory action research at the
fairness in priority setting processes within the health sector: the case of 2010 global symposium on health systems research. MEDICC Rev 2011,
Kapiri-Mposhi District, Zambia. BMC Health Serv Res 2014, 14:75. 13(3):35–38.
27. Maluka S, Kamuzora P, San Sebastiån M, Byskov J, Olsen ØE, Shayo E, Ndawi 48. Biehl S, Petryna A: When People Come First, Critical Studies in Global Health.
B, Hurtig AK: Decentralized health care priority-setting in Tanzania: Princeton, NJ: Princeton University Press; 2013.
evaluating against the accountability for reasonableness framework. Soc
Sci Med 2010, 71(4):751–759. doi:10.1186/1478-4505-12-49
28. Ng’anjo Phiri S, Kiserud T, Kvåle G, Byskov J, Evjen-Olsen B, Michelo C, Cite this article as: Byskov et al.: The accountability for reasonableness
Echoka E, Fylkesnes K: Factors associated with health facility childbirth in approach to guide priority setting in health systems within limited
districts of Kenya, Tanzania and Zambia: a population based survey. resources – findings from action research at district level in Kenya,
Tanzania, and Zambia. Health Research Policy and Systems 2014 12:49.
BMC Pregnancy Childbirth 2014, 14:219.
29. Sandøy IF, Blystad A, Shayo EH, Maundy E, Michelo C, Zulu J, Byskov J:
Condom availability in high risk places and condom use: a study at
district level in Kenya, Tanzania and Zambia. BMC Public Health 2012,
12:1030.
30. Tuba M, Sandoy IF, Bloch P, Byskov J: Fairness and legitimacy of decisions
during delivery of malaria services and ITN interventions in Zambia.
Malar J 2010, 9:309.
31. Echoka E, Kombe Y, Dubourg D, Makokha A, Evjen-Olsen B, Mwangi M,
Byskov J, Olsen OE, Mutisya R: Existence and functionality of emergency
obstetric care services at district level in Kenya: theoretical coverage
versus reality. BMC Health Serv Res 2013, 13:113.
32. Njeru MK, Blystad A, Shayo EH, Nyamongo IK, Fylkesnes K: Practicing
provider-initiated HIV testing in high prevalence settings: Consent concerns
and missed preventive opportunities. BMC Health Services Research 2011,
11:87. doi:10.1186/1472-6963-11-87. Also included and supplemented in:
Njeru MK: Determinants, Experiences and Responsiveness of HIV Testing Submit your next manuscript to BioMed Central
Services in Kenya, Tanzania and Zambia. PhD Thesis. University in Bergen; and take full advantage of:
2011 [https://bora.uib.no/bitstream/handle/1956/4837/Dr.thesis_Mercy%20K.
%1666%2020Njeru.pdf;jsessionid=F5CC2289DA52585E7CB06E31CC161E5B.
• Convenient online submission
bora-uib_worker?sequence=1]
33. Shayo EH, Norheim OF, Mboera LEG, Byskov J, Maluka S, Kamuzora P, • Thorough peer review
Blystad A: Challenges to fair decision-making processes in the context of • No space constraints or color ﬁgure charges
health careservices: a qualitative assessment from Tanzania. Int J Equity
Health 2012, 11:30. • Immediate publication on acceptance
34. WHO: Declaration of Alma-Ata International Conference on Primary Health • Inclusion in PubMed, CAS, Scopus and Google Scholar
Care. USSR: Alma-Ata; 1978.
• Research which is freely available for redistribution
35. Maluka S, Kamuzora P, San Sebastian M, Byskov J, Ndawi B, Olsen ØE, Hurtig
A-K: Implementing accountability for reasonableness framework at
district level in Tanzania: a realist evaluation. Implement Sci 2011, 6:11. Submit your manuscript at
www.biomedcentral.com/submit

Byskov Et Al., 2014

Uploaded by

Copyright:

Available Formats

Byskov Et Al., 2014

Uploaded by

Document Information

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

Byskov Et Al., 2014

Uploaded by

Copyright:

Available Formats

Byskov et al.

Health Research Policy and Systems 2014, 12:49

RESEARCH Open Access

The accountability for reasonableness approach

THE ACTION REACT

ACTION Management capability

Three stage research evaluation of AFR:

Figure 1 The REACT study design.

Table 2 Respondent distribution by study district, organization, and gender

Figure 2 REACT project overview.

Table 3 Comparison of the process of AFR introduction in the three districts

Table 4 Summary assessment of the AFR implementation

Limitations governments from roughly the same group of aid agen-

complexity, and emerges as highly relevant as it is cur- Implications

Actions to be taken ■ The ART should report to all other health

3. Improve All REACT consortium researchers, as shown in the attached consortium

You might also like