Bb9f9 Hawaii Pathways Final Report

HAWAI‘I
PATHWAYS
PROJECT
FINAL REPORT
August 2014–September 2017
1
Acknowledgements
The Hawai‘i Pathways Project is funded by the Cooperative Agreement to Benefit Homeless Individuals for
States (CABHI-States) Grant, No.: TI025340, awarded by the Substance Abuse and Mental Health Services
administration’s Center for Substance Abuse Treatment (SAMHSA-CSAT), under the U.S. Department of Health
and Human Services.
Data and information presented in this report were provided by Hawai‘i Pathways Project Team and
participating organizations, including Hawai‘i State Department of Health’s Alcohol and Drug Abuse Division’s
Treatment Branch, Helping Hands Hawai‘i, Catholic Charities Hawai‘i, and Hawai‘i Interagency Council on
Homelessness. We especially appreciate the Pathways clients who voluntarily participated in various data
collection activities, which contributed greatly to this report.
The photographs featured on the cover page and throughout the report are products from the PhotoVoice
Project taken by Hawai‘i Pathways Project clients.
Report Citation
Yuan, S., Azuma, J., & Gauci, K. T. (2018). Hawai‘i Pathways Project: Final Report, August 2014‒September 2017.
Honolulu: University of Hawai‘i, Center on the Family.
Contact Information
Center on the Family, University of Hawai‘i at Mānoa
2515 Campus Road, Miller Hall 103, Honolulu, HI 96822
Phone: 808-956-4132
Email: cof@ctahr.hawaii.edu
Website: www.uhfamily.hawaii.edu
2 Hawai‘i Pathways Project: Final Report

Table of Contents
Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
About the Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Programmatic Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1. Program Fidelity and Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2. Referral, Admission and Discharge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3. Clients’ Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4. Service Team. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5. Treatment and Support Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
6. Permanent Housing Placement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Program Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
7. Housing Stability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
8. Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
9. Involvement in the Criminal Justice System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
10. Progress in Personal Recovery Goals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
11. Health Status and Health Care Service Utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
System Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
12. Cost Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
13. Impacts on Programs and Policies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Discussion and Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Accomplishments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Barriers and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Hawai‘i Pathways Project: Final Report 3

List of Tables
Table 1: Evaluation Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Table 2: Data Collection Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Table 3: Referral Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Table 4: Demographic Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Table 5: VI-SPDAT Score and Homeless Service Utilization History. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 6: Length of Homelessness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 7: Substance Use and Mental Health Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Table 8: Substance Use and Mental Health Disorders among Veterans. . . . . . . . . . . . . . . . . . . . . . . . . 21
Table 9: Service Team Composition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Table 10: Average Number of Service Encounters per Client per Month. . . . . . . . . . . . . . . . . . . . . . . . 25
Table 11: Number of Clients Served and Service Utilization by Type of Service. . . . . . . . . . . . . . . . . . . 26
Table 12: Source of Housing Voucher. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Table 13: Housing Availability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Table 14: Housing Status at the End of the Grant Period. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Table 15: Length of Housing Stability among Clients who Remained in Housing. . . . . . . . . . . . . . . . . . 32
Table 16: Housing Relocation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Table 17: Current Employment and Education Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Table 18: Had been Hit, Kicked, Slapped, or Otherwise Physically Hurt in the Past 30 Days. . . . . . . . . 35
Table 19: Social Support for Recovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Table 20: Crime and Criminal Justice Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Table 21: Drug and Alcohol Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Table 22: Illegal Drug Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Table 23: Health, Behavioral and Social Consequences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Table 24: Health Status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Table 25: Health Care Utilization in the Past 30 Days. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Table 26: Frequency of Health Care Utilization in the Past 30 Days . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Table 27: Estimated Changes in Health Care Cost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

List of Figures
Figure 1. Logic Model of Hawai‘i Pathways Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 2: Numbers of Referrals, Admissions, and Target Enrollment . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Figure 3: Length of Admission Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 4: Type and Location of Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 5: Reasons for Denial. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Figure 6: Service Team Positions Available and Filled. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 7: Length of Enrollment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figure 8: Service Utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Figure 9: Housing Placement by Month. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Figure 10: Housing Location. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Figure 11: Estimated Cost Savings for Housed Clients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Hawai‘i Pathways Project: Final Report 5

EXECUTIVE SUMMARY
Overview mental health, or co-occurring disorders. The Project

aimed to deliver a program that focused on:
In the spring of 2013, the homeless rate in Hawai‘i was • connecting clients to sustainable, permanent
45.1 per 100,000 people in the general population, housing;
second only to the District of Columbia, and 1.3 times • connecting clients to mainstream benefits
higher than the national average (19.3). Hawai‘i’s and services for low-income people, such as
chronically homeless population reached 1,031, Supplemental Security Income and Medicaid;
exceeding 1,000 persons for the first time ever; this • providing community-based evidence-based
increase represented a 13.3% growth, a stark contrast treatment for substance use and psychiatric
to the 7.3% decline in the national rate of chronic disorders that is client driven and recovery
homelessness. O‘ahu had the state’s largest share oriented; and
of chronically homeless population (67%), with the • providing a range of recovery resources and
majority (73%) living unsheltered, such as on the support including peer navigation services.
street, in parks, encampments, or other places not
meant for human habitation. Among the unsheltered, It also aimed to develop short- and long-term
regardless of their length of homelessness, 429 strategies to expand or enhance the collaboration of
people reported being diagnosed with serious mental various public and private agencies to address system
illnesses, 299 were substance users, 22 had HIV/AIDS, barriers for accessing housing, treatment and recovery
and 191 were veterans. services among the chronically homeless population.
From FFY 14 to FFY 17, the Hawai‘i Pathways The Hawai‘i Pathways Project provided treatment
Project was initiated and administered by Hawai‘i and support services through a hybrid model of
State Department of Health, Alcohol and Drug Assertive Community Treatment (ACT) and Intensive
Abuse Division (DOH-ADAD) with funding from Case Management (ICM). A multi-disciplinary team of
the “Cooperative Agreement to Benefit Homeless housing specialists, mental health counselors, nurses,
Individuals for States” (CABHI-States) grant received peer support specialists, psychiatrists, social workers,
from the Substance Abuse and Mental Health Services substance abuse counselors, and vocational specialists
Administration (SAMHSA). The Project was the first delivered services to clients. Program services were
in the state of Hawai‘i to adopt the evidence-based provided on the island of O‘ahu by the Helping Hands
Pathways to Housing program, which is designed to Hawai‘i and Catholic Charities Hawai‘i. Program
promptly connect chronically homeless people to evaluation was conducted by the University of Hawai‘i
permanent housing, without precondition. Addressing Center on the Family.
any underlying issues then followed around mental
health, addiction, medical care, income and education This report describes the experience of implementing
using a client-driven harm reduction approach. the Pathways Housing First model in Hawai‘i, evaluates
the changes this program has made on the lives
The goal of the Hawai‘i Pathways Project was to of people and on the service delivery system, and
address the gaps in supportive housing services at discusses lessons learned and recommendations for
program and system levels to achieve sustainable future efforts in serving the most vulnerable people
outcomes in housing stability and recovery among among the homeless. The following are highlights of
chronically homeless persons with substance use, the report.

Clients’ Demographic Characteristics and Permanent Housing Placement
Experiences
• The Hawai‘i Pathways Project placed clients in
• There were 134 clients who participated in the scattered-site private market housing, where
Hawai‘i Pathways Project. rental subsidies were provided through existing
• At program enrollment, 7% were in hospitals Permanent Supportive Housing Programs in the
or residential treatment facilities, 24% were state. Clients were required to sign standard
in emergency shelters or Safe Haven, and 69% leases with landlords and pay 30% of their
were unsheltered. income towards the rent.
• 40% of the clients were Hawaiian or Other • 99 clients moved into permanent housing. The
Pacific Islander (HOPI) or part HOPI, 34% were housing placement rate of 79% was due to the
Caucasian, 8% were Asian, 2% were Black, and limited availability of housing vouchers.
16% were another single race or two or more • 38% of clients were housed within four months
races. of program enrollment—a benchmark of the
• Prior to program enrollment, about half of the Pathways Housing First model for programs
clients had been homeless for a continuous without their own supply of housing vouchers.
period of six years or more, with a similar share • Clients waited an average of 8.5 months to
of clients in the range of 6 to 9 years (23%) move into a permanent housing unit because
versus 10 years or more (26%). some vouchers were not available until the
• One-fourth of the Pathways clients were second or the third year of the Project.
diagnosed with a substance use disorder, 15% • The locations of housing units rented by
had a serious mental illness (SMI), and 60% had Pathways clients were spread across 15 ZIP
a co-occurring substance use and SMI disorders code areas on O‘ahu. About three-quarters of
(57%) or co-occurring substance use and clients (73%) lived in urban Honolulu.
mental health disorders (3%).
Program Outcomes
Treatment and Support Services
• The housing retention rate was 90%, with 7%
• During the 38 months of Project returning to homelessness, and 3% leaving the
implementation, clients enrolled in the program program to unknown destinations.
for an average of 22 months. • The average length of housing at the end of the
• The service team reported a total of 10,549 grant was 13.9 months (range: 0.1 to 34.2).
service contacts, of which 80% were successful • 13 housed clients were relocated one to three
contacts and 20% were attempted contacts. times due to lease violations (5), illegal or drug
• The average number of service encounters related activities (4), client’s choice (3), or poor
(successful contacts) per client per month was physical housing conditions (1).
2.9 times, with more frequent encounters • Follow-up interviews reported increases in
during the post-housing period than the pre- housed clients’ reports of good to excellent
housing period (at 4.2 and 2.8, respectively). health (+42%); and not being physically hurt
• The Pathways team delivered services to clients (+38%), abstinence from alcohol or illegal
via home visits or in places where they were drugs (+41%), and not having experienced
needed, without time limits. Service planning drug-related health, behavioral, or social
was based on a client-centered approach where consequences (+41%).
the client’s choice drove the decision on the • Housed clients also reported decreases in
type, frequency and intensity of services. psychological or emotional problems not due
• Clients most frequently used housing support to substance use (–16% to ‒26% on various
services (32%), followed by treatment services problems); alcohol and drug use (alcohol users
(27%), case management services (18%), ‒3%, drug users ‒34%, frequency of alcohol
and peer support services (16%). Other less use ‒32%, frequency of drug use ‒47%); and
frequently used services were medical services health, behavioral and social consequences
(3%), health education (3%), and employment due to substance use (‒50% to ‒62% on various
services (2%). consequences).
Executive Summary 7
System Outcomes Discussion and recommendations
• From baseline to follow-up, health care costs The Hawai‘i Pathways Project demonstrated a
per client per month decreased by 76%. The successful model in housing the hardest-to-serve
average housing costs averaged to $1,100 per population among the homeless—chronically
month for each client housed by the Hawai‘i homeless adults with mental illness, addiction or
Pathways Project, and the average cost for co-occurring disorders. The Project was implemented
providing supportive services by Pathways was with high fidelity, achieved a high housing retention
$850 per month per client. After considering rate, transformed clients’ lives, reduced costs in health
these costs, the net savings equaled $4,247 per care utilization, filled the service gap by helping the
month per client. hard-to-serve homeless population, and accelerated
• Trainings helped to build the system’s system change. However, the pilot project faced
capacity by providing housing-focused case barriers and challenges related to grant administration,
management for health plans, enabling health workforce availability, housing placement, and clients’
plans to examine service gaps in the system, treatment and recovery.
and assisting Community Care Services (CCS)
workers to engage more directly with the The report provides several recommendations
homeless service sector. that include expanding the Housing First program;
• Collaboration among state agencies—for prioritizing the needs of chronically homeless
example, Departments of Human Services, individuals and allocating appropriate resources
Health, Public Safety, Transportation, Land and for services; developing a Housing First learning
Natural Resources—through HICH strengthened community; and addressing the needs for positive
as they worked together on the housing social inclusion.
homeless individuals and providing them with
resources.
• Medicaid 1115 Waiver Amendment was
submitted, which, if approved, would allow
for Medicaid to provide supportive housing
services for chronically homeless individuals
with a behavioral or physical illness, or a
substance abuse diagnosis.
• The technical assistance from the Corporation
for Support of Housing (CSH) through the
Pathways Project’s funding provided Hawai‘i
with housing plans and a financing model. A
snapshot of current housing options, as well as
proposal on how to finance the development of
more affordable housing to meet the needs of
the homeless population based on the Point-in-
Time Count estimation, were developed.

INTRODUCTION
Background
The Hawai‘i Pathways Project was the first in the to the 7.3% decline in the national rate of chronic
state of Hawai‘i to adopt the evidence-based homelessness.4 O‘ahu had the state’s largest share
Pathways to Housing1 program, which is designed of chronically homeless population (67%), with the
to promptly connect chronically homeless2 people majority (73%) living unsheltered, such as on the
to permanent housing, without precondition, and street, in parks, encampments or other places not
then to address any underlying issues around mental meant for human habitation.5 Among the unsheltered,
health, addiction, medical care, income and education regardless of their length of homelessness, 429
using a client-driven harm reduction approach. From people reported being diagnosed with serious mental
FFY 14 to FFY 17, this Project was administered by illnesses, 299 were substance users, 22 had HIV/AIDS,
Hawai‘i State Department of Health, Alcohol and and 191 were veterans.
Drug Abuse Division (DOH-ADAD) with funding from
the “Cooperative Agreement to Benefit Homeless These challenges were met with a new synergy that
Individuals for States” (CABHI-States) grant received focused on improving the coordination of care in the
from the Substance Abuse and Mental Health Services service delivery system and adopting a Housing First
Administration (SAMHSA). On the island of O‘ahu, approach in programming.6 The leading organizations
Helping Hands Hawai‘i and Catholic Charities Hawai‘i of these efforts were the Hawai‘i Interagency Council
provided program services. Program evaluation was on Homelessness (HICH), the State’s Homeless
conducted by the University of Hawai‘i Center on the Programs Office, the City’s Housing Office, and two
Family. continuum-of-care organizations—Partners In Care
on O‘ahu and Bridging the Gap on other islands. The
In the spring of 2013 when the DOH applied for the public/private partnership worked together to build a
CABHI-States grant, the homeless rate in Hawai‘i was coordinated homeless service entry system, starting
45.1 per 100,000 people in the general population, from the implementation of a standard prescreening
second only to the District of Columbia, and 1.3 times tool called Vulnerability Index—Service Prioritization
higher than the national average (19.3).3 Hawai‘i’s Decision Assistance Tool (VI-SPDAT). In addition,
chronically homeless population reached 1,031, a centralized process was piloted for submitting
exceeding 1,000 persons for the first time ever; this prescreening data, prioritizing cases with high level
increase represented a 13.3% growth, a stark contrast of medical and social vulnerability, and connecting
1
Tsemberis, S., & Eisenberg, R. F. (2000). Pathways to housing: Supported housing for street-dwelling homeless individuals with psychiatric
disabilities. Psychiatric Services. 51 (4): 487–493.
2
According to Housing & Urban Development’s definition issued in December 2015, a homeless person is considered “chronically
homeless” when he or she has a disability, and has been living unsheltered, in an emergency shelter, or Safe Haven either for a continuous
period of at least 12 months, or on at least four separate occasions in the past three years provided that the combined length of time of
those occasions is 12 months or more.
3
National Alliance to End Homelessness. (2014). The state of homelessness in America 2014. Retrieved from Alliance to End Homelessness
http://endhomelessness.org/wp-content/uploads/2015/04/2014-state-of-homelessness.pdf
4
Ibid.
5
C. Peraro Consulting. (2013). Statewide homeless point-in-time count: 2013 methodology and results. Retrieved from http://
humanservices.hawaii.gov/wp-content/uploads/2014/05/2013-Statewide-PIT-Report-5.15.13pdf.pdf
6
Yuan, S., Vo. H., & Gleason, K. (2014). Homeless service utilization report: Hawai‘i 2014. Retrieved from http://uhfamily.hawaii.edu/
publications/brochures/60c33_HomelessServiceUtilization2014.pdf
Introduction 9
clients with appropriate housing services through case are system barriers and gaps in resources that add to
conference. From 2012 to 2014, another significant the difficulty in providing adequate mental health and
development of Hawai‘i’s homeless service system substance abuse services to the chronically homeless
was the establishment of permanent supportive population. Building on the Housing First initiative,
housing programs that used a Housing First approach.7 the piloting of the Hawai‘i Pathways Project offered a
The first program was piloted by the state on O‘ahu timely opportunity for stakeholders to work together
from 2012 to 2013,8 and the second program was to address these issues. In this report, we describe the
undergoing the process of planning by the City experience of implementing the Pathways Housing
and County of Honolulu in 2013 and 2014. These First model in Hawai‘i, evaluate the changes this
programs provided rental support for private-market program has made on the lives of people and on the
apartments and offered case management services service delivery system, and discuss lessons learned
that emphasized pre-housing support, landlord liaison, and recommendations for future efforts in serving the
and housing tenure support. most vulnerable people among the homeless.
While the adoption of a Housing First approach to

reduce chronic homelessness is a significant milestone
for Hawai‘i’s homeless service system, there are still
issues that remain unaddressed. Some of these issues
About the Project
The goal of the Hawai‘i Pathways Project was to housing placement. The remainder was provided for
address the gaps in supportive housing services at program administration and system enhancement
program and system levels to achieve sustainable (13%), Pathways Housing First program training (5%),
outcomes in housing stability and recovery among and Project evaluation (3%). Rental support was not
chronically homeless persons with substance use or part of CABHI-States and Supplement grants. With the
co-occurring disorders. assistance of HICH, this Project obtained commitments
from permanent supportive housing programs to
The Project was initially funded for three years from provide housing vouchers for Pathways clients. The
October 2013 to September 2016. Due to delays serving capacity of the program was 155 clients, with
in executing the service contracts, the Project did the enrollment of new clients targeted at 40 to 60 per
not begin its implementation until after 10 months year.
into the first year (August 2014). A one-year no-cost
extension was approved by SAMHSA, extending the Single adults who met all of the following criteria
Project’s implementation period to September 2017. were eligible for this program: (1) scored 10 or
The total amount of funding for the Project was $3.1 higher in VI-SPDAT (version 1); (2) were experiencing
million, of which 68% was from the CABHI-States grant chronic homelessness; and (3) were diagnosed with a
awarded in 2013 and 32% was from the CABHI-States substance use disorder, serious mental illness, or co-
Supplement grant awarded in 2014. The majority occurring substance use and mental health disorders.
(79%) of the funding was budgeted for the provision
of direct treatment and recovery services, outreach,
case management, vocational and peer support, and
7
According to the U.S. Department of Housing & Urban Development (HUD), “Housing First is an approach to quickly and successfully
connect individuals and families experiencing homelessness to permanent housing without preconditions and barriers to entry, such
as sobriety, treatment or service participation requirements. Supportive services are offered to maximize housing stability and prevent
returns to homelessness as opposed to addressing predetermined treatment goals prior to permanent housing entry.” Housing first in
permanent supportive housing [Housing brief]. Retrieved from https://www.hudexchange.info/resources/documents/Housing-First-
Permanent-Supportive-Housing-Brief.pdf
8
Yuan, S., Vo, H., & Garcia, L. (2015). Permanent supportive housing: Pilot implementation report, Hawai‘i 2012–2013. Retrieved from:
http://uhfamily.hawaii.edu/publications/brochures/5d3e2_PSHP_Pilot_FY13.pdf

The Project aimed to deliver a program that focused The project team, including DOH-ADAD, Helping
on the following four aspects: Hands Hawai‘i, and Catholic Charities Hawai‘i,
• Connecting clients to sustainable, permanent received training and technical assistance from
housing; Pathways to Housing to build its capacity to implement
• Connecting clients to mainstream benefits the program with fidelity. Developed by Dr. Sam
and services for low-income people, such as Tsemberis in the early 1990s, Pathways Housing First
Supplemental Security Income and Medicaid; model has shown strong evidence in improving the
• Providing community-based evidence-based health and quality of life of clients with chronically
treatment for substance use and psychiatric homeless experience and mental health/substance
disorders that is client driven and recovery use disorders. Pathways Housing First model reported
oriented; and 85%‒90% retention rate across many cities and
• Providing a range of recovery resources and programs in the U.S.9 Partnering with HICH through
support including peer navigation services. the Governor’s Homeless Coordinator, this Project
facilitated planning and policy development to
It also aimed to develop short- and long-term address system-level solutions for ending chronic
strategies to expand or enhance the collaboration of homelessness.
various public and private agencies to address system
barriers for accessing housing, treatment and recovery
services among the chronically homeless population.
Evaluation
The University of Hawai‘i Center on the Family A program logic model was created based on the
developed a program evaluation plan with input original proposals submitted with the CABHI-States
from the program development team at DOH’s Office and Supplement grant applications (Figure 1). It served
of Program Improvement and Excellence. The plan as a blueprint for the Hawai‘i Pathways Project and
also incorporated SAMHSA’s program reporting guided the evaluation. Some program activities were
requirements and performance measures. The not implemented and some system-level issues were
evaluation period covered the entire Project from not addressed due to low priority and other external
August 2014 to September 2017. The first year refers readiness factors; therefore, those areas were not
to the first 14 months of implementation, ending evaluated (as indicated by an asterisk in the logic
September 2015; the second year refers to the next 12 model).
months, ending September 2016; and the third year
refers to the no-cost extension period of the grant,
ending September 2017.
9
Tsemberis, S., & Eisenberg, R. F. (2000). Pathways to housing: Supported housing for street-dwelling homeless individuals with psychiatric
disabilities. Psychiatric Services. 51 (4): 487–493.; Pathways to Housing. (2012). Pathways to Housing 2012 Annual Report.
Introduction 11
Figure 1: Logic Model of Hawai‘i Pathways Project
OBJECTIVES INPUTS ACTIVITIES OUTPUTS OUTCOMES
• Place chronically • Administration: • Implement • Offered a Client Level System Level

homeless DOH –Alcohol a Pathways Housing First
• Improved the • Developed a
individuals who and Drug Abuse Housing First program with
living conditions statewide plan
meet required Division (ADAD) program high fidelity
and housing to address
disabling • Sponsor: SAMH- • Provide services • Served 155
stability the needs for
conditions into SA’s CABHI-States via Assertive clients who
• Improved interim housing
permanent & Supplement Community met one of
quality of life and the long-
housing grants Treatment (ACT) the following
(e.g. increased term financing
• Link clients to • Program Devel- team: conditions:
education levels, strategies to
health insurance opment: DOH – - Harm reduction - Chronically
employment, sustain Housing
and other Office of program and trauma homeless with
income, social First approach
entitlement Improvement and informed substance use,
connectedness) • Expedited access
programs Excellence approaches or co-occurring
• Reduced to mainstream
• Provide clients - Treatment substance use
• Consultant: Path- utilization of benefits and
with substance services and mental
ways to Housing emergency services*
abuse and - Case management disorders
• Service provider: facilities and • Developed
mental health services - Homeless
Helping Hands increased policies/
treatment as - Education veterans with
Hawai‘i, Catholic utilization of procedures
well as recovery employment severe mental
Charities Hawai‘i services in to streamline
and independent services illness (SMI) or
primary care access to third
living support • Key Partner: Ha- - Peer-to-peer co-occurring
settings party network
services including wai‘i Interagency services disorders
• Reduced payments*
peer navigation Council on Home- - Medical services - Chronically
involvement • Increased in
• Reduce chronic lessness (HICH) - Housing support homeless
with the criminal services to
homelessness • Program Evalua- • Offer Pathways individuals with
justice system clients being
sustainably tion: University of Housing First SMI
• Made progress paid for by
• Implement Hawai‘i Center on program • Organized
toward personal Medicaid
evidence-based the Family training & social Housing First
treatment and • Reduced costs
program with marketing community
recovery goals to serve this
fidelity • Establish training
• Satisfied with population
and facilitate • Convened CAB
services received
Consumer meetings*
Advisory Board • Developed
(CAB)* policies and
• Work with HICH procedures
to develop policy • Provided training
and expand to staff from
partnership 15 agencies in
• Train agencies in how to become
how to become Medicaid Long-Term Impacts
Medicaid providers*
providers*
• End chronic homelessness
• Enhance capacity of the
homeless service system
Note: The original logic model is presented above. An asterisk (*) denotes a program area that was not implemented due
to changes in project's priorities and other external readiness factors.

The evaluation of the Hawai‘i Pathways Project focused on 13 key areas, including six programmatic areas (1–6),
five program outcomes (7–11), and two system outcomes (12–13), which are listed in Table 1.
Table 1: Evaluation Areas

Programmatic Areas Program Outcomes System Outcomes
1. Program fidelity and training 7. Housing stability 12. Cost reduction

2. Referral, admission, and discharge
8. Quality of life 13. Impacts on Programs and Policies
processes
9. Health and health care service
3. Clients' characteristics
utilization
10. Involvement in the criminal justice
4. Housing First service team
system
5. Treatment and supportive services 11. Progress in personal recovery goals
6. Permanent housing placement
Evaluation data was collected via various different methods, including SAMHSA’s Government Performance and
Results Act (GPRA) and other tools. A summary of the data collection methods that were used is presented in
Table 2.
Table 2: Data Collection Methods

Tool Sources Schedule
1. Government Performance and Face-to-face structured interviews Complete interview at program intake,
Results Act (GPRA) questionnaire with clients, by service team 6-month, 12-month, and discharge.
2. PhotoVoice Housed clients told stories of Conduct in the 1st and 2nd years
photographs they took, in a one-on-
one or small group setting with the
evaluation team
3. Key informant interview: Service Semi-structured interviews by the Conduct at the end of the 2nd & 3rd
team, ADAD project coordinator, evaluation team years
HICH chair
4. Program’s referral, admission, and Online forms completed by service Enter data on a continuous basis, due
discharge information team by the 5th day of the following month
5. Service log: Services provided to Spreadsheet completed by service Enter data on a continuous basis, due
clients team by the 5th day of the following month
6. Housing status of clients Spreadsheet completed by service Enter data on a continuous basis, due
team by the end of program implementation
Introduction 13
GPRA interviews were completed by all 134 clients at reference was the Program Fidelity Assessment report
program intake. Follow-up interviews were conducted completed for the Hawai‘i Pathways Project by Dr.
with 116 clients, representing a follow-up rate of Sam Tsemberis and Juliana Walker of the Pathways to
87%. When multiple follow-up interviews were Housing Institute on February 28, 2017.
completed, only one was used in the baseline follow-
up analysis: For the housed clients, it was the first To maintain client confidentiality, pseudonyms
follow-up interview after housing placement; for the are used throughout the report. Due to rounding,
un-housed clients, it was the last follow-up interview. percentages may not always add up to 100 in this
Twenty clients participated in the PhotoVoice project report. This evaluation study focused on describing
and were interviewed individually (15) or in small the Project’s implementation and short-term impacts
groups of 2‒3 persons (5). Semi-structured interviews on the lives of the participants and the homeless
were conducted with 11 Project staff members and service system. Significance tests were performed on
other key stakeholders. Participation in the GPRA changes from baseline to follow-up. In this study, the
interviews and PhotoVoice were both voluntary and minimum significance level for a given test is a p-value
not a condition for services. Tokens of appreciation of less than 0.1 (*), which means the probability for
in the form of supermarket gift cards were provided the observed change to occur by chance is less than
to clients who participated in the GPRA follow-up 1%. Stronger evidence is indicated by a p-value of less
interview ($10), GPRA discharge interview ($30), than 0.05 (**) and less than 0.01 (***). It should be
and PhotoVoice interview ($30). Institutional Review noted that the strength of evidence increases with
Board (IRB) approvals for various study protocols were the increase in sample size, the size of change, and
obtained from the University of Hawai‘i. the data variance. Due to the small number of Project
participants, it is expected that strong evidence on
Other data used in this report were gathered from program impacts would not be found on behaviors or
the program’s referral, admission and discharge events of few occurrence.
information; reports of service delivery; and changes
in housing status of the clients. An additional
These are my slippers. They symbolize for me for

other people to live in my shoes. To just be more
open and not so closed-minded or looking down at
us [who] are homeless. Not everybody chooses to be
homeless, you know? I didn’t; it just happened. I’m
taking it one day at a time [now]….I try to say, don't
judge a book by its cover but so many people [do].
It's sad that society puts a title on your head that
you're homeless so you must be bad, on drugs or an
alcoholic. Instead of just asking.
—Kapena, client

PROGRAMMATIC AREAS
1. Program Fidelity and Training

The Hawai‘i Pathways Project was modeled after the evidence-based Pathways Housing First program developed
by Dr. Sam Tsemberis in New York City during the 1990s. This program was adopted by many cities across the
United States, Canada, and other countries to eradicate chronic homelessness with great success. The Project
received training and technical assistance from Dr. Tsemberis and Julian Walker of the Pathways Housing First
Institute to guide the implementation. In January 2017, the Institute conducted a fidelity assessment of the
Hawai‘i Pathways Project. The fidelity assessment measures cover five domains: (1) housing choice and structure,
(2) separation of housing and services, (3) service philosophy, (4) service array, and (5) program structure. Each
domain has 6 to 10 items, with each item rated on a scale from 1 (low fidelity) to 4 (high fidelity), for a total of
38 items and 152 maximum score possible.10 The Hawai‘i Pathways Project received a score of 134, which was
88% of the total possible score. The Program Fidelity Report11 results showed that the Hawai‘i Pathways Project
implemented the Housing First model with high fidelity.
• Team philosophy, practice and operations are consistent with the Pathways Housing First
model.
• The team has several well-trained and experienced staff members who have a clear understanding
of the Housing First model and operate the program in a manner that is consistent with the
program’s core principles and values.
• The team has…consistently [offered] clients housing of their choice, on their own terms, without
prerequisites for treatment or sobriety that has created the success for many who had remained
homeless for years.
• The team understands and practices the principle of keeping housing separate but coordinated
with support and treatment.
• The housing provided by the team generally meets the fidelity standards for Housing First.
• The apartments are rented from community landlords and are integrated into the building and
community so that participants are living in “their home” not “in a program.” …This is helpful in
facilitating a normative lifestyle and social interaction with other members of the community.
—Program Fidelity Report, p.3‒4
The Hawai‘i Pathways Project also engaged in training and social marketing of Pathways Housing First
program in the state. In February 2014, Dr. Tsemberis and Ms. Walker were invited to provide training and to
educate stakeholders about the Pathways Housing First model. The community training was well attended by
stakeholders including state agencies, policymakers and service providers in the homeless, health and human
service fields. Moreover, the project team participated in several “boot camps“ organized by the Interagency
Council to discuss strategies and challenges in implementing Housing First. The team members also presented at
the annual Statewide Homeless Awareness Conference (2015‒2017) and the Harm Reduction Conference (2017)
to share their experience about implementing Housing First and harm reduction approaches. Furthermore,
the project team reached out to health plans and case management programs to get them on board with
incorporating the Housing First approach and applying Motivational Interviewing techniques in serving clients
with chronically homeless experiences.
10
Tsemberis, S. & Stefancic, A. (2011). Pathways Housing First fidelity scale (ACT version) [unpublished].
11
Tsemberis, S., & Walker, J. (2017). Program fidelity report. Prepared for Helping Hands Hawaii Pathways Housing First Program
[unpublished].
Programmatic Areas 15
2. Referral, Admission and Discharge
Referral and Admission

The Hawai‘i Pathways Project received 262 referred cases and admitted 134 (51%) clients, meeting 86% of the
enrollment target. Due to a 10-month delay in executing service contracts, the Project was under great pressure
to catch up with the target number, which was reached at the sixth month of program implementation in January
2015. However, the enrollment of new clients fell short again after April 2015. The last client was admitted to the
program in September 2016.
Figure 2: Numbers of Referrals, Admissions, and Target Enrollment

300
Referrals Admitted 262
250
Referrals Received
Target Number
200
Number of People
155
150
134
100
50
0
Aug Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr Jun Aug
2014 2015 2016
Nearly three-quarters of referrals came from the coordinated entry system (CES) for homeless services where
people with a VI-SPDAT score of 10 or higher were identified as needing permanent supportive housing. Through
the Project’s outreach efforts, the referral network was expanded to other service providers who didn’t have
access to the CES. Other referral sources included substance abuse treatment service providers (11%), hospitals
(9%), community mental health service providers (5%), and others (2%). Referrals from CES had the highest
admission rate, at 57%, and referrals from substance abuse treatment service providers and “other” sources had
the lowest rates, at 27% and 33%, respectively.
Table 3: Referral Sources

CES Hospital SAT CMH Other Total
Number of referrals 191 23 30 12 6 262
% of total referrals 73% 9% 11% 5% 2% 100%
% of referrals admitted 57% 43% 27% 50% 33% 51%
Note: CES—coordinated entry system for homeless services; SAT—substance abuse treatment providers; CMH—community-
based mental health service providers; other—state and other community agencies.

Half of the clients were admitted within 20 days upon being referred to the Project, while the mean was 46 days
(range: 0–380). People referred from hospitals took an average of 77 days, probably due to health conditions
requiring extended periods of care.
Figure 3: Length of Admission Process

90
80
John was living in a tent on the
70 77
Kapālama Canal and had no
Number of Days
60 interest in being housed, but

50 we kept showing up…. Then, we
40 46
started to uncover some of his
43 45
grief and trauma. [Now] he's been
30
housed for two and a half years.
20 —Service team member
20
10
0
All Referrals All CES Hospital Other
Median Mean
Note: CES stands for Coordinated Entry System for homeless services; other referral sources included
community-based mental health and substance abuse treatment providers and others.
The admission process involved establishing contacts with the referrals, engaging the referrals for eligibility
screening, and obtaining consent from those eligible for participating in the program. On average, the service
team contacted each referral six times before an admission decision could be made. About 10% of referrals were
contacted 15 times or more. Overall, one-third of the contacts during the admission process were face-to-face
meetings with referrals at places in which they were located, such as: streets (53%); hospitals, clinics, or other
organizations (31%); or shelters or other temporary locations (16%). About two-thirds of the contacts were made
to clients by telephone or through various types of collateral contacts.
Average number of times
Figure 4: Type and Location
5.9
of Contacts
7.0 Other cont Face‐to‐face contacts

Face‐to‐face 16%
Total Other contacts4.0 2.0
6.0
2.2 53%
5.0 2.0
Number of Times
1.7 31%
4.0
3.0
2.0 4.2 4.0

3.6 Streets
Hospitals, clinics, offices
1.0
Shelter, other temporary places
0.0
Not admitted Admitted All referrals
32% 68%
Note: Other contacts include those made to clients by telephone or through various types of collateral contacts.
Of the 128 referred cases that were closed or denied, the main reasons were: no contacts were established with
referrals before the close of Project enrollment period by September 2016 (27%); inability to contact or locate
the referrals (25%); referrals did not meet program entry criteria such as low VI-SPDAT score, not homeless, or no
disabilities (22%); referrals were withdrawn due to other housing opportunities, hospitalization or death (16%);
and referrals refused services or other reasons (10%). People referred by hospitals were more likely to have
problems in re-establishing contacts by the team after they left the hospital, whereas those referred by other
community organizations were more likely to be not meeting the chronically homeless and behavioral health
criteria for utilizing this program.
Figure 5: Reasons for Denial

70%
CES Hospital Other Total
60%
% of Referrals Not Admitted
59%
50% 54%
40%
39%
30%
31%
27%
20% 25% 25%
22% 22% 20%
10% 16%
13%
10%
8% 3% 6% 0% 6% 8% 6%
0%
Withdrawn Not eligible Unable to contact Refused/other No contact
established
Note: CES stands for Coordinated Entry System for homeless services; other referral sources included community-based
mental health and substance abuse treatment providers and others.
Discharge
Of the 134 clients enrolled, 19% were discharged due Twenty-three clients (21%) were discharged without
to the client’s death (9); client becoming ineligible for a transition plan due to lack of contact or inability to
services (9); or client requiring long-term nursing care locate the client (14), client refusing services (7), or
or hospitalization, or incarceration (8). staff’s safety concerns (2).
The remaining 108 clients (81%) were all discharged

by September 2017 when the CABHI-States and
Supplement grants ended. Of these clients, 85 (79%)
were successfully discharged to a program with an
appropriate level of continuing support. Of these
One of our graduates just came back over here to
clients, the majority (51) was connected to the
say hello. He wanted to say, “Hey, I'm still here and
Comprehensive Community Services (CCS) program
just want you guys to know I'm doing okay.”
or the case management services by Helping Hands
—Service team member
Hawai‘i. Some (29) were transferred back to their
permanent supportive housing programs for case
management, and a few un-housed clients (5) were
admitted to another housing placement program.

3. Clients’ Characteristics
Demographics
The average age of the 134 clients was 50 years (median: 52; range: 24‒76). Nearly three-quarters (72%) of the
clients were men. Most clients self-identified as Hawaiian or other Pacific Islander (HOPI) or Part HOPI (40%),
or white (34%). Regarding their educational attainment, most clients reported having obtained a high school
diploma (38%) or higher level of education (44%). More than half (57%) had children. At the start of Project
enrollment, the majority of clients (69%) was living in places not meant for human habitation while one-fourth
(24%) were in an emergency shelter or Safe Haven, and the remaining clients were in an institution (hospital 3%,
or residential treatment facility 4%).
Table 4: Demographic Characteristics
Number Percent
Gender
Male 96 72%
Female 35 26%
Transgender 3 2%
Race
Hawaiian or Other Pacific Islander (HOPI)/
51 40%
Part HOPI
Caucasian/White 44 34%
Asian 10 8%
Black 2 2%
Other single race/2 or more racesa 21 16%
Educational Attainment
Less than high school diploma 25 19%
High school diploma 50 38%
Some college 43 33%
Bachelor's degree or higher 14 11%
Have children
Yes 75 57%
No 56 43%
Living Situation at Program Enrollment
Unsheltered 92 69%
Emergency shelter or Safe Haven 32 24%
Hospital or residential treatment facility 10 7%
Total 134 100%
Note: Some totals are less than 134 due to missing: race (6), education attainment (2), have children (3).
a
The “2 or more races” category does not include Part HOPI.
VI-SPDAT Score
All clients were screened by the VI-SPDAT (single person, version 1) where a score between 10 and 20 signified
the need for permanent supportive housing placement. Among the Pathways clients, 57% had a higher score,
between 13 and 16. Those who had participated in any programs in the homeless service system prior to
enrolling in Pathways tended to have a higher vulnerability score (13 and above), compared to the new homeless
service users (62% vs. 41%).
Table 5: VI-SPDAT Score and Homeless Service Utilization History

New Homeless Existing Homeless
VI-SPDAT Score Number of Clients Percent of Clients
Service Users Service Users
10 15 18 33 25%
11–12 5 20 25 19%
13–14 12 37 49 37%
15–16 2 25 27 20%
Total 34 100 134 100%
Homeless Experience
All Pathways clients are chronically homeless by HUD’s definition. About half of them enrolled in the Project
had been homeless for a continuous period of six years or more, with a similar share of clients in the range of 6
to 9 years (23%) vs. 10 years or more (26%). Three-quarters of clients (100 out of 134) have accessed homeless
services before being referred to Pathways.
Table 6: Length of Homelessness

Length of Continuous Homelessness Percent of Clients
1‒2 years 23%

3‒5 years 28%
6‒9 years 23%
10 years or more 26%
Behavioral Health Status

The CABHI-States and Supplement grants focus on the chronically homeless population with substance use,
co-occurring disorders, or serious mental illness. One-fourth of the Pathways clients were diagnosed with a
substance use disorder, 15% had a serious mental illness (SMI), and 60% had a co-occurring substance use
and SMI (57%) or co-occurring substance use and mental illness (3%). The majority of clients (72%) had two or
more diagnoses. The top six diagnoses that affected 12% or more clients were: affective psychoses (43%), drug
dependence (43%), adjustment reaction (39%), alcohol dependence syndrome (37%), nondependent abuse of
drugs (20%), and schizophrenic psychoses (12%).

Table 7: Substance Use and Mental Health Disorders
Diagnosis Number Percent
Category
Substance abuse only 34 25%
Serious mental Illness (SMI) only 20 15%
Co-occurring substance use and mental health disorders/SMI 80 60%
Co-occurring Substance use and...
Mental 4 3%
SMI 76 57%
Most Common Diagnosis
296. Affective psychoses 57 43%
304. Drug dependence 57 43%
309. Adjustment reaction 52 39%
303. Alcohol dependence syndrome 49 37%
305. Nondependent abuse of drugs 27 20%
294. Schizophrenic psychoses 16 12%
Note: Diagnosis was reported in ICD-9 codes.
Veteran Status
Twenty-five clients reported that they had previously served in the armed forces, the National Guard, or the
reserves, representing 19% of all clients. Two-thirds of military veterans (68%) who participated in Pathways
were diagnosed with SMI.
Table 8: Substance Use and Mental Health Disorders among Veterans

Diagnosis Category Number of Veterans Percent
Substance abuse only 8 32%
Serious mental illness (SMI) only 5 20%
Co-occurring substance use and SMI 12 48%
Total 25 100%
Here’s the bus stop on my next station where every

time I get up, this is where I go. Tripler, bus stop, and
the VA….I used to go [to Tripler] every day. Now they
got me going only on my appointment time.
—Mike, client
4. Service Team
The Hawai‘i Pathways Project provided treatment and The service team grew from three to eight members
support services through an Assertive Community in the first year and from eight to fourteen members
Treatment (ACT) model. Pathways’ ACT team was in the second year. However, only eight members
a multi-disciplinary team that consisted of housing remained on the team in the third (extension) year
specialists, mental health counselors, nurses, peer because some staff members decided to leave due
support specialists, psychiatrists, social workers, to the uncertainty of whether the Project would be
substance abuse counselors, and vocational specialists. extended. The ICM-to-clients ratio was about 1 to 20,
The service team held meetings 2‒3 times a week and the overall staff-to-clients ratio was maintained
to discuss cases, and to deliver services based on at around 1 to 10 until the third year when it became
clients’ changing needs. For example, the Certified as high as 1 to 18. To address the staff shortage, the
Substance Abuse Counselor (CSAC) would visit a client Project contracted Mental Health Kokua to provide
who expressed interest in entering a rehabilitation peer coaching and the CHOW Project to provide
program; a nurse would visit a client to treat his housing navigation and outreach services in the
wound; and a peer support specialist—someone who last year of the Project. In addition, there were
had lived the recovery experience—would visit a client administrators at the Helping Hands Hawai‘i and
to provide support. While team members had their Catholic Charities Hawai‘i who oversaw the team’s
own specific roles, they worked together to provide effort and a Project assistant who provided clerical
the needed supports for clients to be successful in support.
housing and recovery. For instance, an Intensive
Case Manager (ICM) would take a client to a medical
There was one day where I was out doing
appointment if the team nurse was not available or to outreach, and all I did was to help her clean her
a housing appointment when the housing specialist tent--that was what she needed. Her tent had
could not. roaches all over it, and old food.
–Service team member
Table 9: Service Team Composition
Role Start of Program Year 1 Year 2 Year 3
Psychiatrist/APRN-RX 1 1 1 1
Project Coordinator 1 1 1 0
Team Leader 0 1 1 1
RN-ICM 1 1 2 1
ICM – Veteran 0 1 1 1
ICM – CSAC, other 0 1 3 2
Housing Specialist 0 1 2 1
Peer Navigator 0 1 2 1
Vocational Specialist 0 0 1 0
Total 3 8 14 8
Note: Service team composition data included all positions that were filled at any length or full-time equivalent (FTE) units
in the respective reporting year. APRN-RX—advanced practice registered nurse with prescriptive authority; RN—registered
nurse; ICM—intensive case manager; CSAC—Certified Substance Abuse Counselor. Year 1: August 2014–September 2015;
Year 2: October 2015–September 2016; Year 3: October 2016–September 2017.

One of the main barriers in reaching the Project’s two years of experience serving the homeless or
enrollment target was the inability to fully staff substance use populations (55%).
the service team. The service team had a total of
18 positions or 16 full-time equivalent (FTE) units Other positions had better recruitment results at
available, but only 50% of all available FTEs were 70%‒80% of the available FTEs having been filled.
filled over the three-year period. The state’s low The Project coordinator position (filled 70% FTE)
unemployment rates (e.g., 4.1% in October 2014 and was responsible for coordinating Project’s activities,
3.3% in October 2015) could have made hiring more connecting with partner agencies, and engaging
difficult, and certain job positions were particularly community stakeholders. The team leader position
hard to fill due to required work experience or unique (filled 80% FTE) required a Certified Substance Abuse
qualifications. The most difficult to fill was the peer Counselor or equivalent credential and was mainly
navigator positions that required a Hawai‘i Certified responsible for running daily meetings, coordinating
Peer Specialist credential (20% of available FTEs filled), services, supervising staff, and providing direct services
followed by the vocational specialist (30%), housing to clients. The psychiatrist or advanced practice
specialists (55%), and positions with specializations registered nurse with prescriptive authority (APRN-
in intensive case management (including ICM-Nurse, RX) position (filled 80% FTE) provided assessment and
ICM-Veteran, ICM-CSAC, and other ICM) that required medical care in the community where the clients were
located.
Figure 6: Service Team Positions Available and Filled

6.0
FTE available
5.0 FTE filled (averaged for 3 years)
5.0
4.0
3.0 3.5
FTE
2.8
2.0
2.0
1.0 1.5 1.0 1.0
1.0 1.0
1.1
0.8 0.8 0.7 0.8 0.7 0.3
0.0
Team ICM RN– Peer Psychiatrist Project Housing Vocational
Leader ICM Navigator Coordinator Specialist Specialist
Note: Data for the available full-time equivalent (FTE) units for each position is presented. Filled positions were calculated
based on a three-year average to represent the extent to which each position was filled for the implementation period.
Due to staffing issues, the Hawai‘i Pathways Project

team adopted an individual caseload approach, with The morning meeting is very much like
each client assigned to a case manager while other an ACT team morning meeting…. Also,
team members served as a backup and provided individual caseloads are managed flexibly
specialized services as needed. This deviated from the to make geographic coverage for home
Pathways Housing First recommended approach to visits more efficient.
shared caseloads. —Program Fidelity Report, p. 7
5. Treatment and Support Services
The Hawai‘i Pathways Project focused on providing

harm reduction and trauma-informed care for people [This project is about] harm reduction, housing
experiencing chronic homelessness and behavioral first…. If you are going to tell them [to stop], they’re
health issues. It also offered a broad scope of going to be turned off by you. We’re already working
treatment services directly and coordinated treatment with people that nobody else wants to work with,
with community providers if the clients were already that nobody else wants to house…. So if you want to
connected to their own providers. Intensive case get results, you got to meet them where they’re at
management was available for everyone, which and provide them with education and build rapport.
included the coordination of medical care with And you’re not going to see anything happen
providers such as hospitals, primary care providers, overnight, but this is the steps that it’s going to take.
Waikiki Health Center, and Kalihi-Palama Health —Service team member
Center. Other main categories of services were housing
support, peer support, employment support, and
health education. Crisis response was available 24/7. on strategies and support needed to reach those goals.
As a Housing First program, Pathways clients were
The Pathways team delivered services to clients via not required to participate in psychiatric treatment or
home visits or in places where they were needed, obtain sobriety as a condition to housing; however,
without time limits. Service planning was based they had to agree to weekly home visits by the service
on a client-centered approach where the client’s team.
choice drove the decision on the type, frequency and
intensity of services to be provided. The service team During this 38-month Project, clients were enrolled for
recognized that clients have their own set of needs, an average of 22 months (or 648 days, ranging from 51
experiences, characteristics and strengths, and that to 1,110 days). One in five clients enrolled for 30‒36
support should be flexible, accepting and adaptive to months, 25% enrolled for 24‒29 months, 14% enrolled
specific needs of individual clients. The team utilized for 18‒23 months, 26% enrolled for 12‒17 months,
Motivational Interviewing techniques to help clients and 14% for 11 months or less.
set their own goals and worked with clients in deciding
Figure 7: Length of Enrollment
30%
25% 26% 25%

Percent of Clients
20%
20%
15%
14% 14%
10%
5%
0%
11 months 12‒17 18‒23 24‒29 30‒36
or less months months months months

The service team reported a total of 10,549 service contacts, of which 80% were successful contacts and 20%
were attempted contacts. The majority of the unsuccessful contacts were outreach attempts to clients at
unsheltered locations before they were housed. The average number of service encounters (successful contacts)
per client per month was 2.9 times, ranging from 0.1 to 7.7 times. The intensity was higher after, rather than
before, the clients were housed, at 4.2 versus 2.8 times per month per client.
Table 10: Average Number of Service Encounters per Client per Month
Number of Service
Pre-housing Period Post-housing Period Overall
Encounters
Mean 2.8 4.2 2.9
Median 2.5 3.7 2.9
Range 0.1‒15.2 0.5‒22.8 0.1‒7.7
Most service encounters involved multiple types of services, with the average being 1.7 service units per
encounter. While the amount of time spent for each service unit was not reported, a typical service encounter
was 30 minutes (median; mean: 33.78 minutes; 90% range: 5‒120 minutes). The category of services that
were most frequently used was housing support services (32%), followed by treatment services (27%), case
management services (18%), and peer support services (16%). Other less frequently used services were medical
services (3%), health education (3%), and employment services (2%).
Figure 8: Service Utilization
35%
30% 32%
25% 27%
Percent of Services
20%
15% 18%
16%
10%
5% 3% 3% 2%
0%
Housing Treatment Case Peer Support Medical Health Employment
Support Management Education
Note: A total of 14,676 service units were reported from 8,432 service encounters.
During the project period, housing support services were received by 130 clients (97%) and each of these clients
received an average of 35.9 service units. Housing support services were mainly provided by case managers
and housing specialists, and included pre-housing services, services to support housing tenure, and re-housing
services.
Table 11: Number of Clients Served and Service Utilization by Type of Service
Number of Service Average Units Per
Service Type Number of Clients
Units Client
TOTAL 14,676 134 109.5
Housing Support Services 4,663 130 35.9
Pre-housing Services 1,705 130 13.1
Services to Support Housing Tenure 2,672 99 27.0
Re-housing Services 286 22 13.0
Treatment Services 3,974 121 32.8
Assessment 1,458 109 13.4
Individual Counseling 748 95 7.9
Brief Intervention 584 91 6.4
Treatment or Recovery Planning 363 84 4.3
Mental Health Services 232 62 3.7
Pharmacological Interventions 142 53 2.7
Screening 129 50 2.6
Community Integration & Recovery Suppl. Service 142 46 3.1
Co-occurring Treatment/Recovery Services 106 36 2.9
Referral to Treatment 52 24 2.2
Brief Treatment 11 10 1.1
Other Counseling 7 7 1.0
Case Management Services 2,628 116 22.7
Care Coordination 1,768 111 15.9
Transportation 310 87 3.6
Individual Services Coordination 535 75 7.1
Other Case Management Services 15 10 1.5
Peer-to-Peer Services 2,299 92 25.0
Peer Coaching or Mentoring 920 79 11.6
Peer Housing Support 851 76 11.2
Peer Navigation Services 492 56 8.8
Information and Referral 30 17 1.8
Other Peer Services 6 6 1.0
Medical Services 444 69 6.4
Health Education: Substance Abuse & Other 378 67 5.6
Employment 290 28 10.4
Note: Data is based on unduplicated types of services reported for each encounter. A total of 14,676 service units were
reported from 8,432 service encounters, averaged at 1.7 service units per encounter. The length of each service unit was
not reported. A typical service encounter was 30 minutes (median; mean = 33.78 minutes; 90% range = 5‒120 minutes).

Treatment services were received by 121 clients (90%) Medical services were received by 69 clients (51%) at
at an average of 32.8 service units per client. Major an average of 6.4 service units per client. Pathways’
treatment services included assessment, received by psychiatrists and nurses provided community-based
109 clients, followed by individual counseling (95), treatment for illness or injury, and other medical
brief intervention (91), treatment or recovery planning services for clients who did not have an established
(84), and mental health services (62). relationship with any medical service providers.
Case management services were received by 116 General health education and substance abuse
clients (87%) at an average of 22.7 service units per education were received by 67 clients (50%) at
client. Major case management services were care an average of 5.6 service units per client. Pre-
coordination (111 clients), transportation (87), and employment and employment coaching services were
individual services coordination (75). These services received by 28 clients (21%) at an average of 10.4
focused on supporting housing stability of clients. service units per client.
Peer support services were received by 92 clients

(69%) at an average of 25.0 service units per clients.
We would show up every single day, post-housing, for
More than half of clients used peer coaching or
a week or two, and give the clients the consistency
monitoring services (79), housing support services
of care.... [We would] say, “Does your shower work?
(76), and four in ten used peer navigation services
Do you have food in your fridge? Did you finish your
(56). These services focused on supporting clients after
application for your benefits? Did you collect your
they were housed. The majority of the services was
allowance check…?” Things like that.
provided by the peer specialists but some were offered
by other service team members.
I'm looking [at] where I came from to get

here ….[It wasn’t] too long [ago] or too
bad. I could have probably sped [things]
up if I got my head out of my rear—excuse
me—and kept doing what I wanted or
needed to do. [It] just took time, nothing
else, to get me, to get my head, [and to] get
one with myself. I didn't feel I was together
as a whole. And if I didn't get with myself,
I can't help myself, [and] how could I help
anybody else?...The worker Camille had a
lot of influence [in] getting me to wake up…
She would be here [and] come to wherever
I was. I could be doing whatever, and she
would come to me and [say], “Hey, we got
to do this. We got to do that.” Basically,
[she was] getting me out of my shell and
into society.
—Bailey, client
6. Permanent Housing Placement
The Hawai‘i Pathways Project placed clients in clients’ preferences, and set up unit showings to let
scattered-site private market housing, where rental the clients decide if they wanted to apply. Once the
subsidies were provided through existing Permanent application was submitted and approved, the clients
Supportive Housing Programs in the state. Clients would sign the lease with the landlord and set up
were required to sign standard leases with landlords electricity or other utilities as needed. The Hawai‘i
and pay 30% of their income towards the rent. They Pathways Project would also furnish the apartment,
had to follow the conditions of their lease in the provide move-in kits, and assist clients with moving
same ways as any other renters. Housing was not their personal items.
contingent on abstinence or treatment, but intensive
support services helped clients maintain stable The Project secured a total of 80 permanent
housing. One of the core principles of Housing First is supportive housing vouchers—52% of the enrollment
that housing and support services are separated. As target— through the assistance of the Hawai‘i
a couple of examples, clients who have to move out Interagency Council on Homelessness (HICH). The
of their housing by choice or due to involuntary lease scarcity of rental assistance was the Project’s main
termination will receive re-housing assistance rather barrier in offering Housing First to all clients. Sources
than being discharged from the Project; clients who of the vouchers were the newly-funded Housing First
no longer require intensive support services will not programs by the state (20) and the City and County of
be asked to leave their home or give up their housing Honolulu (10 from Increment I and 20 from Increment
vouchers. II), as well as other continuing programs funded by
HUD such as Shelter Plus Care program (20), Veterans
Another core principle of Housing First is client choice. Affairs Supportive Housing (VASH, 6), and permanent
The housing specialist searched for housing based on supportive housing program for persons with AIDS (4).
Table 12: Source of Housing Voucher

Funding Source Organization Number of Vouchers Percent
City Housing First Increment I I.H.S. 10 13%
City Housing First Increment II U.S. Vets 20 25%

Kalihi-Palama Health Center (10)
HUD Shelter Plus Care 20 25%
Steadfast Housing Devel. Cor. (10)
HUD VASH U.S. Vets 6 8%
HUD HOPWA Gregory House 4 5%
State Housing First U.S. Vets 20 25%
Total 80 100%
Note: I.H.S.—Institute of Human Services; U.S. Vets—United States Veteran Initiatives; HUD—U.S. Department of Housing and
Urban Development; VASH—Veterans Affairs Supportive Housing; HOPWA—Housing Opportunities for Persons with AIDS.
Before the Project started, 40 vouchers were promised; however, about half of them were not available until
the second half of the first year. An additional 20 vouchers were secured in the second year and another 20
were obtained in the third year. The Project was able to place a total of 99 clients with 80 vouchers due to re-
allocation of vouchers following program discharge of housed clients. Reasons for discharge are discussed in the

Housing Stability Section of this report. Overall, the housing placement rate was 74%. About 2.6 clients were
placed into housing per month (range: 0‒13), and the months with few or no housing placements reflected the
lack of availability of housing vouchers in those periods.
Figure 9: Housing Placement by Month
14
12
Number of Clients
10
8
6
4
2
0
8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9 101112 1 2 3 4 5 6 7 8 9
2014 2015 2016 2017
The Hawai‘i Pathways Project helped 38 clients move quickly into a permanent housing unit of their choosing,
defined as moving-in within four months upon entering the Project by the Pathways Housing First model.
However, due to the limited availability of housing vouchers and given that about only one-fourth of the
vouchers became available in the third year, half of the housed clients waited 6.1 months or longer to move into
a permanent housing unit (mean: 8.5 months; range: 0.0‒33.1).
Table 13: Housing Availability

Number of Months from Enrollment to Number of Housed
Percent
Housing Placement Clients
0‒4 38 38%
5‒8 26 26%
9‒12 18 18%
13 or more 17 17%
Total 99 100%
Note: Mean = 8.5 months; median = 6.1 months; range = 0.1‒33.1 months.
Another major barrier to housing placement was finding landlords who accepted housing vouchers and were
willing to rent to Pathways clients. More landlords became willing to rent after hearing about some initial
successes of housing placements. In addition, the availability of the city’s housing complex—Winston Hale—
made it possible for the Project to house 18 remaining clients in the final year.
The locations of housing units rented by Pathways clients were spread across 15 ZIP code areas on O‘ahu (see
Figure 10). About three-quarters of clients (73%) lived in urban Honolulu with the top three highly concentrated
ZIP codes being 96817 (23 clients), 96815 (11 clients) and 96822 (10 clients).
Figure 10: Housing Location
a. Island of O’ahu
b. Part of urban Honolulu c. Zip Code 96786

PROGRAM OUTCOMES
7. Housing Stability
The top intervention priority of the Hawai‘i Pathways permanent supportive housing programs that provided
Project was to assist chronically homeless persons the vouchers took over the case management of
with addictions and mental health challenges to obtain the clients. Some clients were also enrolled in the
and retain housing. Of the 99 clients who moved into Medicaid’s Community Care Services or received the
permanent housing, 11 withdrew from the Project extended case management services from the Helping
due to death (7) or requiring nursing care or long-term Hands Hawai‘i.
hospitalization for non-psychiatric reasons (4). Of the
remaining 88 clients, nine clients left for an unknown
[Pathways clients] are the guys with the highest
destination (3) or returned to homelessness (6), VI-SPDAT scores. [The kind of guys that some
but the rest—a total of 79—were still in permanent outreach] workers would say, “honestly, I don't
housing at the end of the grant period, representing know how this person ever going to get into
a housing retention rate of 90%. Hawai’i’s outcome housing.” Now, to see Pathways figure out a
was consistent with nationwide results: Pathways way where they can provide the right levels of
Housing First model reported 85%‒90% retention support that are tailored to each person to get
rate across many cities and programs in the U.S.12 them in housing—I think that's huge.
Housing vouchers of the clients were not affected — Governor’s Coordinator of Homelessness
by the ending of the CABHI grant. At discharge, the
Table 14: Housing Status at the End of the Grant Period

Housing Status Number of Clients
Permanent Housing 79
Subsidized housing with supportive services 77
Non-subsidized permanent housing 1
Moved in with family/friend—permanent 1
Homeless 6
Unsheltered settings 5
Emergency/transitional shelter (including hostel) 1
Unknown 3
Other 11
Deceased 7
Long-term care facility or care home 3
Hospitalization (non-psychiatric) 1
All Housed Clients 99
12
Tsemberis, S., & Eisenberg, R. F. (2000). Pathways to housing: Supported housing for street-dwelling homeless individuals with
psychiatric disabilities. Psychiatric Services. 51 (4): 487–493.; Pathways to Housing. (2012). Pathways to Housing 2012 Annual Report.
Program Outcomes 31
For the clients who were housed and remained housed (79) at the end of the grant, the mean length of housing
was 13.9 months and the median was 11.1 (range: 0.1–34.2). A fifth (20%) were stably housed for 24 to 34
months, more than a quarter (28%) between 12 to 23 months, another quarter (24%) between 6 to 11 months,
and the remaining (28%) were housed for less than six months. The shorter housing periods were attributed to
housing placement in the third year of the Project.
Table 15: Length of Housing Stability among Clients who Remained in Housing
Length Number of Housed Clients Percent
Less than 6 months 22 28%

6 to 11 months 19 24%
12 to 23 months 22 28%
24 to 34 months 16 20%
Total 79 100%
Note: Mean = 13.9 months, median = 11.1 months, and range = 0.1–34.2 months.
Of the 99 clients housed, 13 were relocated one to three times, and 10 of them remained stably housed at the
end of the grant while three returned to living on the street. The reasons for relocation were lease violations (5),
illegal or drug-related activities (4), client’s choice (3), and poor physical housing conditions (1). All clients who
relocated due to lease violations, choice and poor physical housing conditions remained housed at the end of
the grant. However, only one of the four clients who relocated due to illegal or drug-related activity remained
housed at the end of the Project.
Table 16: Housing Relocation

Number of Clients Remained
Reasons for Moving Out Number of Clients Relocated
Housed at the End of the Project
Illegal or drug-related activity 4 1
Other lease violation 5 5
Client's choice 3 3
Poor physical housing conditions 1 1
Total 13 10
Of the clients who were housed, many expressed a

sense of accomplishment, pride and independence
Now I haven’t been to the psych ward or the when they obtained a place of their own. Having
emergency [room] because I have a place I can housing provided them with several positive life
try to get my head together. It makes so much improvements such as the ability to more easily
of a difference. You’ve got a lock on the door. access public benefit programs, reconnect with their
You can lock the door. I don’t know if you guys families, and care for their physical and mental health.
have ever been homeless, but can you imagine Without a place of residence, making mental health
trying to sleep on the street? You wouldn’t feel and welfare appointments, for example, were difficult.
safe. Everybody’s trying to steal everything. Prior to being housed, many clients’ prioritized their
—Terrance, client immediate safety first. Being housed provided safety
and shifted their priorities, allowing clients to focus
on other things like their appointments. They also-

became more motivated to do things for themselves,
and many of them became sober enough to do so.
In addition, the Pathways team members were more
[You need to learn] how to live again. When you
easily able to work closely with housed clients to
are homeless you have freedom.… You can do
provide them reminders about appointments and even
anything you like. No one can tell you nothing.…
transport them to appointments when needed.
When you live in a house you have freedom, but
you have responsibilities. When you are homeless
Reconnecting with family members was possible when
you don’t really have responsibilities.
clients had a place of their own made them easier
—Danny, client
to locate. As with Bill’s case, his son was able to visit,
drop off food, and leave notes—things his son couldn’t
do when Bill was on the streets. Not only did housing
stability allow for clients to reestablish relationships the efforts that the support services team provided.
with others, it provided clients with opportunities to Others’ opinions about the support staff echoed his
focus on themselves in ways that were not prioritized thankfulness and appreciation of the team’s critical
when they were homeless. No longer worried about roles in helping the clients view options, secure
where they would sleep or if they were safe, they were housing and cope with challenges.
able to reengage in activities that made them happy
such as creating art, caring for plants and crafting an Acquiring stable housing required adjustments
exercise routine. in routines and interpersonal interactions. One
client highlighted how his routine consisted of him
With regards to the housing process, clients had constantly moving about to survive when he was
varying experiences. Some were able to get housed homeless. He would visit specific agencies at set
quickly, and others enjoyed taking time to look at times for meals and events. But after securing stable
different places. One participant spoke of his negative housing, his routine mainly revolved around organizing
experience with signing a lease for an apartment that his belongings, feeding himself and taking care of his
he later found was not well maintained. While he was hygiene. Some other adjustments clients had to make
unhappy with the apartment itself, he commended were changing their sleeping habits, interacting with
others—such as their neighbors—and personalizing
their apartments.
When you’re homeless, here are my little rules: in There were challenges that clients faced throughout
the housing process including breaking from old
after dark, out before dawn, don’t leave a mess,
routines; creating boundaries with friends who were
and don’t leave a trace… just keep moving.… 5:00
still homeless; integrating with the community;
I’m at the storage locker. 6:00 I’m at the Vietnam
abiding by restrictions tied to housing arrangements;
Vets. 8:00 I’m at the State Capitol.… What I’ve
and coping with loneliness. A client spoke of her
done since I’ve [gotten housing]—it should be
need to break routines that she had with people with
simple to most people—but what I’ve really
whom she once used drugs. She emphasized finding
tried to concentrate on doing is organizing my
other things to do that would not drive her back to
belongings, keep everything clean. Little things homelessness. Several clients faced restrictions while
like eat breakfast, take a shower, brush your teeth. housed. It was common for landlords to restrict house
Eat lunch, take a shower, brush your teeth. Eat guests, which posed a problem for Mary who hurt
dinner, take a shower, brush your teeth.… I’ve just her leg and needed her partner to come over to help
tried to concentrate on those elements of hygiene care for her. Some clients experienced loneliness
and eating because when you’re homeless it’s real after being housed, and one client in particular
difficult to eat regularly and to bathe regularly. felt that having housing further contributed to her
Those things are challenging. depression. It was likely for clients to still feel a sense
—Richard, client of community with their homeless friends and to help
them with supplies, money or food helped to ease
their loneliness.
Program Outcomes 33
8. Quality of Life
Employment and Education

Employment and education provide economic security
and more easily allow for social integration. Volunteer
work also provides social and psychological benefits. Interestingly enough, we met Jack at 3B2 at
Clients were asked about their employment and Tripler which is the psychiatric floor. He threw
education statuses during baseline and follow-up something at one of our staff members when
assessments. Employment status referred to whether she tried to do the assessment. He went from
they had current employment in a full-time (35 or that to housing.… He's still in the same house
more hours per week) or part-time job, and education that he was in, and he now works full time…
status was defined as being enrolled in school, a GED manages his mental health. He's working on
class, or a job training program. paying his rent.
At baseline, 17 clients were working, volunteering or
taking classes, with the number dropping by 15% at
follow-up. Even though 13 clients found a job at some
point during the project period, their employment
tended to be volatile to changes in health and job
conditions. At follow-up, the number of those looking
for work (8) increased slightly and those who reported
that they were disabled, retired, or not looking for
work (91) dropped slightly. Overall, the changes in
employment and education status from baseline to
follow-up were found to be not statistically significant
(at p < 0.1) for all clients, regardless of their housing
status at follow-up.
Table 17: Current Employment and Education Status

All Clients: All Clients: All Clients: Housed: Housed: Housed:
Status
Baseline Follow-up % Change Baseline Follow-up % Change
Working, volunteering, taking classes 17 15 ‒12% 12 9 ‒25%
Looking for work 8 11 38% 3 9 200%
Disabled, retired, not looking for work 91 90 ‒1% 60 57 ‒5%
Note: Data included all 116 clients who completed the follow-up interview; 75 of them were housed at follow-up. No
statistically significant difference (at p < 0.1) was found between baseline and follow-up data among all clients as well as the
housed clients.
I was a waiter. I had a business of my own at one point. I've had quite a bit of good [work]
experiences before [my mental illness crisis] started. Those I can't really forget, so just
not really having that level of respect [working as a busboy at my age], I was not able to
keep that job because it became stressful in another way. Ever since then, the team is just
recommending that I just really not work right now and focus on myself, to just make sure
that I'm really stronger from not drinking.
—Steve, client

Violence Victimization
The number of clients that stated they had not been
hit, kicked, slapped or hurt physically in the past 30 For the last two and a half, three years, I slept
days increased by 23% from baseline to follow-up. For underneath that truck. After living on the streets
those that stated that they were physically hurt a few for many years, your defense mechanisms
times or more than a few times, the number of clients become sharpened and you know when it’s
decreased (60% and 33%, respectively). Positive and peaceful. You can almost feel things coming
statistically significant changes (p < 0.01) were seen before they come…. Sleep an hour at a time if
for those who were housed, with more of them never you could, and then just wake up and then look
having been physically hurt (38% increase) and less around. If it’s okay go back to sleep for a while.
of them being hurt (a few times: 75% decrease; more You never really sleep deeply.
than a few times: 75% decrease). —Richard, client
Table 18: Had been Hit, Kicked, Slapped, or Otherwise Physically Hurt in the Past 30 Days
Status
Baseline Follow-up % Change Baseline Follow-up % Change ***
Never 74 91 23% 48 66 38%
A few times 25 10 ‒60% 20 5 ‒75%
More than a few times 6 4 ‒33% 4 1 ‒75%
Note: Data included 105 clients who completed the follow-up interview; 72 of them were housed at follow-up. Eleven
clients (including three housed clients) were excluded due to missing data. Tests of significance were conducted on the
changes from baseline to follow-up for all clients and housed clients. Three asterisks (***) denotes p < 0.01, meaning that the
probability for the change to occur by chance is less 1%. A significant level of p >= 0.1 is not presented.
Social Connectedness
Being connected socially and positively could provide

individuals with productive support that influences
their progress. Clients were asked several questions
that aimed to determine their social connectedness.
Social connectedness comprised of their involvement
within the past 30 days in religious- or faith-affiliated
recovery self-help groups, other self-help groups,
other organizations that support recovery, and
interactions with family and/or friends who were
supportive of their recovery. Overall, 74 clients
reported that they were socially connected at
baseline, but that number dropped by 7% at follow-
up. The observed changes for all clients, as well as
the housed clients, were not found to be statistically
My kids are my backbone in life. They are the significant (p >= 0.1).
reason I am still alive. I don't want to be a
deadbeat dad, so I try to be a part of my kid's lives.
—Kale, Client
Program Outcomes 35
With regards to attending self-help groups or recovery
meetings, 56 clients reported at baseline that they
Fortunately, I have a psychiatrist. I have a
participated in these activities in the past month, but
therapist. I have these home visits. Normally,
that number dropped by 29% (p < 0.05) by follow-
Sao [case manager] is here at 9:30 every
up. A drastic and statistically significant decrease was
Monday, and then Wednesday I have my
observed among the housed clients (‒38%, p < 0.05),
therapist visit. There's a food bank that I
with the largest decrease in their participation of
go to on Thursday. I'm working on getting
religious- or faith-affiliated recovery self-help groups
my knee fixed…. I could just stay [in the
(‒63%), followed by other recovery meetings (‒47%).
apartment] all day and all night every day,
but that's not good either. So what I'm trying
A total of 60 clients stated that they had interactions
to do is figure out a new routine. And it
with family and/or friends that were supportive of their
doesn't happen overnight. Sometimes I think
recovery at baseline, and the same number stated the
I've done really well for the last two months,
same at follow-up (60). The number of clients who
but I still have a long way to go.
were housed and interacted with friends and/or family
—Richard, client
increased slightly (by 7%) from baseline to follow-up.
Table 19: Social Support for Recovery

In the Past 30 Days...
Socially connected 74 69 ‒7% 51 47 ‒8%
Had attended any self-help and recovery
56 40 ‒29% ** 39 24 ‒38% **
groups or meetings
…Number of times 550 494 ‒10% 402 236 ‒41% **
Voluntary self-help groups 33 27 ‒18% 22 16 ‒27%
…Number of times 397 291 ‒27% ** 292 142 ‒51% ***
Religious/faith-affiliated recovery
26 16 ‒38% * 16 6 ‒63% **
groups
…Number of times 81 105 30% 61 48 ‒21%
Meetings of organizations 22 15 ‒32% 17 9 ‒47% *
…Number of times 72 68 ‒6% 49 16 ‒67% *
Had have interaction with family and/or
friends that are supportive of [client’s] 60 60 0% 41 44 7%
recovery
Note: Data included 116 clients who completed the follow-up interview, with 12‒15 missing responses in each question.
The number of housed clients was 75. Tests of significance were conducted on the changes from baseline to follow-up for all
clients and housed clients. An asterisk (*) denotes p < 0.1, meaning that the probability for the change to occur by chance is
less 1%; ** denotes p < 0.05; and *** denotes p < 0.01. A significant level of p >= 0.1 is not presented.

9. Involvement in the Criminal Justice System
Upon entering the Project, nine clients reported that The number of clients awaiting charges, trial, or
they had been arrested in the last 30 days, with a sentencing (16) decreased by 38% from baseline to
cumulative total of 13 arrests. The number of clients follow-up. A 75% decrease (statistically significant at
and arrests decreased 33% and 54%, respectively, at p < 0.05) was observed among the housed clients.
follow-up. Among the housed clients, the number
of clients reported being arrested in the last 30 days By follow-up, there was a 14% increase in those who
remained unchanged at follow-up while the number of were currently on parole or probation (from 14 at
total arrests decreased. baseline), though not statistically significant; and
there was no change among the housed clients.
The number of clients (7) who had been in jail or
prison in the past 30 days did not change from
baseline to follow-up, but the total number of nights
spent in jail or prison (40) increased by 50%. Of the
housed clients, the numbers of clients and nights
spent in jail or prison increased 33% and decreased
55%, respectively.
Table 20: Crime and Criminal Justice Status

Status
Had no involvement in the system in the
93 96 3% 63 63 0%
past 30 days
Had been arrested in the past 30 days 9 6 ‒33% 5 5 0%
…Number of arrests 13 6 ‒54% 9 5 ‒44%
Had been in jail/prison 7 7 0% 3 4 33%
…Number of nights in jail 40 60 50% 29 13 ‒55%
Currently awaiting charges, trial, or
16 10 ‒38% 12 3 ‒75% **
sentencing
Currently on parole or probation 14 16 14% 10 10 0%
Note: Data included 116 clients who completed the follow-up interview; 75 of them were housed at follow-up. The number
of missing responses for each question was: 14 for arrests, 20 for in jail or prison, 12 for awaiting charges, and 9 for parole.
Tests of significance were conducted on the changes from baseline to follow-up for all clients and housed clients. Two
asterisks (**) denote p < 0.05, meaning that the probability for the change to occur by chance is less 5%. A significant level of
p >= 0.1 is not presented.
Program Outcomes 37
10. Progress in Personal Recovery Goals
Abstinence
From baseline to follow-up, the number of clients who
reported that they abstained from alcohol or illegal
drugs in the past 30 days (23) increased by 48%. For [Pathways is] keeping tabs on me. That way
those who continued to use alcohol, the number of every week I can kind of tell them what I'm doing
days where alcohol was consumed decreased (‒33%). this week, what are my goals.… What this is all
Similarly, the total number of clients who used illegal about is trying to build a new reality. The other
drugs and the number of days used decreased 37% one [homeless reality] I had down to a science,
and 52%, respectively. All the aforementioned changes but it was tiny and it wasn't very fulfilling. What
were statistically significant (p < 0.1) among housed I'm trying to do is figure out fulfillment.
clients. In addition, the number of clients who used —Richard, client
alcohol to intoxication with more than five drinks in
one sitting, as well as the number of days intoxicated,
decreased by 25% and 40%, respectively, even though
the strength of evidence was low (p >= 0.1).
Table 21: Drug and Alcohol Use

Had abstained from alcohol or illegal
23 34 48% * 17 24 41%
drugs
Used any alcohol or drugs a 74 63 ‒15% * 48 41 ‒15%
Used alcohol 51 45 ‒12% 30 29 ‒3%
…Number of days 788 526 ‒33% *** 425 290 ‒32% *
Intoxicated (5+ drinks in one sitting) 24 18 ‒25% 15 10 ‒33%
…Number of days 434 259 ‒40% 261 109 ‒58%
Used illegal drugs 52 33 ‒37% ***
35 23 ‒34% **
…Number of days 771 367 ‒52% *** 530 280 ‒47% ***
Note: Data included 98 clients who completed the follow-up interview and answered this set of questions; 65 of them were
housed at follow-up. Eighteen clients didn't respond and 10 of them were housed clients. a Nine clients used both alcohol
and drugs for a total of 97 days. Tests of significance were conducted on the changes from baseline to follow-up for all
clients and housed clients. An asterisk (*) denotes p < 0.1, meaning that the probability for the change to occur by chance is
less 1%; ** denotes p < 0.05; and *** denotes p < 0.01. A significant level of p >= 0.1 is not presented.
Of the 52 clients who used illicit drugs, the most commonly used drugs were marijuana/hashish (65%) and
methamphetamine or other amphetamines (63%), followed by OxyContin/oxycodone (13%), heroin (12%) and
cocaine/crack (10%). From baseline to follow-up, all five top drugs saw a decrease in the number of users and
the frequency of use. Clients who were housed at follow-up reported a decrease across all individual drugs and
frequency of use.

Table 22: Illegal Drug Use
In the Past 30 Days, Clients Used...
Baseline Follow-up Change Baseline Follow-up Change
Marijuana/hashish 34 19 decreased 23 12 decreased
…Number of days 294 214 decreased 154 140 decreased
Methamphetamine or other
33 14 decreased 20 7 decreased
amphetamines
OxyContin/oxycodone 7 6 decreased 6 5 decreased
Heroin 6 3 decreased 5 3 decreased
Cocaine/crack 5 0 decreased 4 0 decreased
Note: Data included clients who reported using illicit drugs: 52 clients at baseline and 33 clients at follow-up. The housed
group consisted of 35 at baseline and 23 at follow-up. Drugs used by less than five clients are not presented in the table.
Health, Behavioral and Social Consequences

At the follow-up interview, there was a 31% increase in
There really isn't a place to go if you're not ready
the total number of clients who reported that they did
to go through detox…. you have to be ready, and
not experience any alcohol or illicit drug-related health,
you have to show up… Steve is a great example
behavioral, or social consequences. From baseline, there
of that. After two and a half years of us talking
were fewer clients who said that their use of alcohol or
about it, he's finally showing up at IOP [Intensive
drugs contributed to their stress, caused them to reduce
Outpatient Program] treatment center three
or give up important activities, and caused emotional
days a week…. he's doing it!
problems. This pattern of change was statistically
significant (p < 0.01) among all clients and housed clients.
Table 23: Health, Behavioral and Social Consequences

Had experienced no alcohol or illegal
drug-related health, behavioral or social 62 81 31% *** 42 57 36% ***
consequences
Had stress due to alcohol or drug use 61 33 ‒46% *** 40 17 ‒58% ***
Use of alcohol or drugs caused clients to
46 23 ‒50% *** 29 11 ‒62% ***
reduce or give up important activities
Use of alcohol or drugs has caused
51 27 ‒47% *** 34 17 ‒50% ***
emotional problems
Note: Data included 98 clients who completed the follow-up interview and answered this set of questions; 65 of them were
housed at follow-up. Eighteen clients didn't respond and 10 of them were housed clients. Tests of significance were conducted
on the changes from baseline to follow-up for all clients and housed clients. Three asterisk (***) denotes p < 0.01, meaning that
the probability for the change to occur by chance is less 1%. A significant level of p >= 0.1 is not presented.
Program Outcomes 39
11. Health Status and Health Care Service Utilization
Health Status
At baseline, 70 clients rated their general heath as being fair or poor, but by their follow-up interviews, that
number decreased significantly by 26%. The same positive change was observed and found to be statistically
significant among those who were housed at the follow-up interviews.
The number of clients who experienced mental, emotional and behavioral hardships not due to substance abuse
in the past 30 days decreased from baseline to follow-up. Drastic and significant drops (18%‒38%, p < 0.1) were
seen for those who experienced serious depression, serious anxiety or tension, cognitive difficulties, and trouble
controlling violent behavior. There was also a significant decrease (47%, p < 0.01) in those who were considerably
to extremely bothered by psychological or emotional problems not related to non-substance abuse in the past
30 days. Most of these positive changes were observed in among the housed clients, except the improvement in
controlling violent behavior, which was not statistically significant.
Table 24: Health Status

Status
General Health Status * *
Excellent, very good, good 34 52 53% 24 34 42%

Fair, poor 70 52 ‒26% 45 35 ‒22%
In the Past 30 Days, not Due to
Substance Use, Experienced…
Serious depression 85 67 ‒21% *** 55 46 ‒16% *
Serious anxiety or tension 82 67 ‒18% ** 53 44 ‒17% *
Trouble understanding,
76 54 ‒29% *** 47 35 ‒26% **
concentrating, or remembering
Trouble controlling violent behavior 26 16 ‒38% * 17 10 ‒41%
Hallucinations 27 19 ‒30% 16 11 ‒31%
Attempted suicide 6 2 ‒67% 4 2 ‒50%
Been prescribed medication for
43 38 ‒12% 31 24 ‒23%
psychological or emotional problem
Bothered by Non-substance-use-
related Psychological or Emotional *** ***
Problems in the Past 30 Days

Not at all, slightly, moderately 27 52 93% 19 35 84%
Considerably, extremely 53 28 ‒47% 34 18 ‒47%
Note: Data included 116 clients who completed the follow-up interview; 75 of them were housed at follow-up. Missing
data for each question were: general health (12), various psychological or emotional problems (14‒24), bothered by these
problems (12). Tests of significance were conducted on the changes from baseline to follow-up for all clients and housed
clients. An asterisk (*) denotes p < 0.1, meaning that the probability for the change to occur by chance is less 1%; ** denotes
p < 0.05; and *** denotes p < 0.01. A significant level of p >= 0.1 is not presented.

Health Care Service Utilization
The number of clients who reported using inpatient

treatment and emergency room treatment in the
past 30 days decreased 50% and 19%, respectively, I'm trying to get back into drawing. I like to
from baseline to follow-up. The number of clients do design. I couldn’t when I was on the street;
utilizing outpatient treatment increased slightly (4%) things always get lost or wet or stolen. I wasn’t
at follow-up. For the different treatment services, feeling very creative. When I feel like I’m
clients were asked if they received care for physical creative, I feel better, especially [with] plants
complaints, mental or emotional difficulties, or alcohol because they are alive. I can’t have a pet, so
or substance abuse. Except for those who sought [the plant] is my pet [in my apartment]. They’re
outpatient treatment for physical issues, the number doing really good too.
of clients dropped who sought care across all settings. —Terrance, client
A statistically significant decline was found in the use
of any inpatient treatment for all clients (p < 0.05).
The same decline applied to inpatient treatment for
alcohol or substance abuse for all clients (p < 0.01) and
for those who were housed (p < 0.1).
Table 25: Health Care Utilization in the Past 30 Days

Utilization
Any Emergency Room Treatment 31 25 -19% 23 15 -35%
For physical complaint 27 22 -19% 21 13 -38%
For mental or emotional
7 3 -57% 4 2 -50%
difficulties
For alcohol or substance abuse 4 1 -75% 3 1 -67%
Any Inpatient Treatment 24 12 -50% ** 14 8 -43%
For physical complaint 12 9 -25% 9 7 -22%
6 3 -50% 2 1 -50%
ifficulties
For alcohol or substance abuse 10 1 -90% *** 6 0 -100%*
Any Outpatient Treatment 50 52 4% 35 38 9%
For physical complaint 29 31 7% 23 23 0%
33 26 -21% 21 20 -5%
difficulties
For alcohol or substance abuse 12 9 -25% 7 6 -14%
Data included 116 clients who completed the follow-up interview; 75 of them were housed at follow-up. There were about
9‒13 missing responses for this set of quesions. Tests of significance were conducted on the changes from baseline to
follow-up for all clients and housed clients. An asterisk (*) denotes p < 0.1, meaning that the probability for the change to
occur by chance is less 1%; ** denotes p < 0.05; and *** denotes p < 0.01. A significant level of p >= 0.1 is not presented.
Program Outcomes 41
The number of visits for emergency room treatments, outpatient visits, and nights for inpatient treatments, all
dropped from baseline to follow-up for an average of 53% decrease across settings. The decline was statistically
significant for all clients, as well as the housed group (‒63%, p < 0.05), but insignificant for the non-housed
group across all settings (‒33%, p >= 0.1).
Table 26: Frequency of Health Care Utilization in the Past 30 Days

Type of Setting Baseline Follow-up % Change
All: ER, number of visits 94 43 –54% **

All: Inpatient, number of nights 355 126 –65% ***
All: Outpatient, number of visits 402 231 –43% **
All: Total 851 400 –53% **
All: Average per client 8.0 3.7 –53%
Housed: ER, number of visits 74 26 –65% **
Housed: Inpatient, number of nights 231 42 –82% ***
Housed: Outpatient, number of visits 267 145 –46% *
Housed: Total 572 213 –63% **
Housed: Average per client 7.9 3.0 –63%
Un-housed: ER, number of visits 20 17 –15%
Un-housed: Inpatient, number of nights 124 84 –32%
Un-housed: Outpatient, number of visits 135 86 –36%
Un-housed: Total 279 187 –33%
Un-housed: Average per client 8.0 5.3 –33%
Note: Data included 107 clients who completed the follow-up interview and answered this set of questions; 72 of them
were housed and 35 were un-housed at follow-up. Nine clients didn't respond and three of them were housed clients. Tests
of significance were conducted on the changes from baseline to follow-up for all clients, as well as housed clients and un-
housed clients. An asterisk (*) denotes p < 0.1, meaning that the probability for the change to occur by chance is less 1%; **
denotes p < 0.05; and *** denotes p < 0.01. A significant level of p >= 0.1 is not presented.
It's a real blessing

now, that's a peace
of mind. Before I felt
more [of an] outcast in
society. [This] is just a
step toward normalcy.
—Danny, Client

SYSTEM OUTCOMES
12. Cost Reduction

The public costs of managing the service needs of the chronically homeless people with behavioral health
disorders created a significant financial burden to the criminal justice and health care systems.13 One of the
objectives of the Hawai‘i Pathways Project was to reduce public costs by providing stable housing and supportive
services for this population based on the Housing First approach. Due to the small number of Pathways clients
reporting involvement with the criminal justice system and no significant changes found from baseline to follow-
up interviews, this section focuses on analyzing cost reduction through the health care system rather than the
criminal justice system.
Table 27: Estimated Health Care Cost

Estimated Cost: Estimated Cost: Estimated Cost:
Type of Setting
Baseline Follow-up Difference
All: ER, number of visits $115,902 $53,019 –$62,883
All: Inpatient, number of nights $681,245 $241,794 –$439,451
All: Outpatient, number of visits $79,998 $45,969 –$34,029
All: Total $877,145 $340,782 –$536,363
All: Average per client $8,198 $3,185 –$5,013
Housed: ER, number of visits $91,242 $32,058 –$59,184
Housed: Inpatient, number of nights $443,289 $80,598 –$362,691
Housed: Outpatient, number of visits $53,133 $28,855 –$24,278
Housed: Total $587,664 $141,511 –$446,153
Housed: Average per client $8,162 $1,965 –$6,197
Un-housed: ER, number of visits $24,660 $20,961 –$3,699
Un-housed: Inpatient, number of nights $237,956 $161,196 –$76,760
Un-housed: Outpatient, number of visits $26,865 $17,114 –$9,751
Un-housed: Total $289,481 $199,271 –$90,210
Un-housed: Average per client $8,271 $5,693 –$2,577
Note: Data included 107 clients who completed the follow-up interview and answered this set of questions; 72 of them
were housed and 35 were un-housed at follow-up. Nine clients didn't respond and three of them were housed clients.
Estimated costs were calculated by multiplying the total usage (Table 26) with the average cost for each type of service
based on recent research literature.
13
Moulton, S. (2013). Does increased funding for homeless programs reduce chronic homelessness? Southern Economic Journal, 79(3),
600–620.
System Outcomes 43
Based on clients’ self-reported health care utilization Results of this analysis should be interpreted with
for the 30 days prior to their interviews, the estimated caution due to data sources and certain assumptions
health care cost was $8,198 per client at baseline and involved in the calculations. For one, accuracy of the
$3,185 per client at follow-up, representing a 61% data relies on clients’ abilities to recall health care
decrease and an estimated cost savings of $5,013 utilization in the past 30 days. Secondly, individuals’
per client. Cost reduction was more drastic among health care utilization levels vary across time.
clients who were housed at follow-up (76% decrease, Moreover, this analysis assumed that the 30-day
estimated cost savings of $6,197 per client) compared service utilization is the average monthly utilization,
to those who were not housed (31% decrease, which may not be true. An average calculated from
estimated cost saving of $2,577 per client). The 12 months of data could be very different from the
decrease in health care cost was statistically significant 30-day data. Thirdly, cost estimation is based on the
(p < 0.1) among all clients and the housed clients, average cost for specific types of health services from
but not among the non-housed group. This analysis the latest published studies and may not represent
showed that, while providing treatment and recovery typical health care cost for the chronically homeless
services helped lowering health care cost in general, population. Nevertheless, these are the best data and
stable housing was the key contributor to significant methods available for this study.
cost savings among those who experienced chronic
homelessness and behavioral health disorders.
The estimated costs were calculated by multiplying Now I am housed. Over the last seven years,
the total usage with the average cost for each type of I went from occasional drug user to being
service: a drug addict, to being depressed, [having]
• ER visit: Median ER charge was $1,233 for the anxiety, mental problems, physical problems
ten most frequent outpatient diagnoses based all because [I was] living in a vicious cycle, you
on a national study utilizing the 2006‒2008 know? I didn’t have no mental problems until
Medical Expenditure Panel Survey (MEPS) I became homeless.... I used to run buildings;
data.14 I’m a building engineer. Because I was
• Inpatient care: Hospital adjusted expenses per homeless [with] no address, it is really tough
patient day for Hawai‘i was $1,919 based on to get a place. You need an address. The only
2013 Annual Survey of the American Hospital way to break the cycle is you got to get out
Association.15 of the cycle. You got to go back to reality and
• Outpatient visit: A physician office visit get a job... my next move is get my resume
averaged $199 in the U.S. in 2008 based on together. Get my certificates together. Because
the Medical Expenditure Panel Survey (MEPS) now I am thinking clearly because I don’t have
data.16 to deal with drugs. I don’t have to deal with
alcohol. I deal with stress and anxieties and all
of that stuff. But on a clearer level.
—Harvey, Client
14
Caldwell, N., Srebotnjak, T., Wang, T., & Hsia, R. (2013). “How much will I get charged for this?’’ Patient charges for top ten diagnoses in
the emergency department. PLoS ONE, 8(2), e55491.
15
The Henry J. Kaiser Family Foundation. (n.d.). Hospital adjusted expenses per inpatient day, Hawaii, 2015. State Health Facts.
Retrieved from https://www.kff.org/health-costs/state-indicator/expenses-per-inpatient-day/?currentTimeframe=0&select-
edRows=%7B%22states%22:%7B%22hawaii%22:%7B%7D%7D%7D&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22as-
c%22%7D
16
Agency for Healthcare Research and Quality. (2011, April 20). Hospital outpatient care represents more than 20 percent of all ambula-
tory care costs in U.S.. AHRQ News and Numbers. Retrieved from http://archive.ahrq.gov/news/newsroom/news-and-numbers/042011.
html

ng,
bering ER, est. @
$ 91,242 $ 32,058 -$ 59,184
$1,233 per visit
Inpatient, est. @
$ 443,289 $ 80,598 -$ 362,691
76%
$1,919 per night
ly
ms Estimated Cost
Outpatient, Savings
est. @ for Housed Clients
$ 53,133 $ 28,855 -$ 24,278
$199 per visit
Among the housed
Average costsclients, the health care costs per client per month decreased by $6,197 at follow-up, which
$ 8,162 $ 1,965 -$ 6,197
was 76%
perless
client
than the costs at baseline. The average monthly rent was $1,100 per client housed by the Hawai‘i
Pathways Project, and the average cost for providing supportive services by Pathways was $850 per month per
Note:After
client. Of the 75 clients who
considering were
these interviewed
costs, at program
the net savings intake to
equated and$4,247
after housing placement,
per month 72 were
per client.
included in the health care analysis and 3 were excluded due to missing data. Estimated costs were
calculated by multiplying the total usage with the average cost for each type of service based on recent
research literature. References are provided in the "Hawai'i Pathways Project: Final Report" available
Figure 11:Center
from the Estimated Cost Savings
on the Family website. for Housed Clients
Average Cost per Client
ESTIMATED HOUSING FIRST ESTIMATED

REDUCTION IN $1,950 NET SAVINGS
HEALTH CARE $4,247
COSTS Housing Pathways
$6,197 voucher supportive
$1,100 Services
$850
I just want to get out and really try

to work on my physical health. Get
back in shape [by] getting out jogging,
going around maybe the clubhouse,
and then coming back around.
Maybe going swimming [too]. Just
try to stay really busy and active.
—Steve, Client
System Outcomes 45
13. Impacts on Programs and Policies
The Hawai‘i Pathways Project was the first treatment

and supportive services project of its kind in the state
and was an agent of change that had system level
impacts. According to the Governor’s Coordinator on
Homelessness, Scott Morishige, there was a reciprocal
relationship between the Hawai‘i Pathways Project The CABHI grant was very positive in
and the Hawai‘i Interagency Council on Homelessness triggering a lot of things at the system level.
(HICH) that he chairs: the framework of the HICH It triggered better city-state collaboration,
assisted the implementation of the Pathways Project, and within the state, [helped us] to look at
such as opening doors to collaboration between the resources beyond just the Department of
Pathways Project and other agencies and securing Human Services and Homeless Programs
housing vouchers for Pathways clients. In turn, Office to address homelessness.
through the barriers and challenges experienced, —Governor’s Coordinator of Homelessness
the Pathways Project helped the council to be more
effective and purposeful in addressing systematic
issues. A significant step the council took was to seek
consultation and training from national experts on
the subject matter. As a relatively small part of the
Pathways Project grant was dedicated for system- System’s Capacity Building
level work, some funding was instead used to
secure a contract with the Corporation for Support The Project has allowed for additional training to
of Housing (CSH) to contribute technical support better support the chronically homeless population.
and a broad skill set. The CSH provided consultation Examples include trainings on providing housing-
on matters that included financial modeling for focused case management for health plans, trainings
supportive housing, case management for chronically for health plans to examine service gaps in the system,
homeless individuals through Medicaid waiver, and and trainings for Community Care Services (CCS)
strategies for Interagency Council to coordinate workers to engage more directly with the homeless
services at a system level and to leverage resources service sector. These types of training helped health
from various departments. In addition to CSH, the care providers to navigate the homeless service system
training and technical assistance the Project received and be more effective in helping their clients who are
from the Pathways to Housing Institute also made homeless. Besides workshops and boot camps, the
a lasting impact on the state’s response to chronic partnerships that were created between the Pathways
homelessness. The impact the Project had at the Project and the state and city’s Housing First programs
system level influenced training, collaboration, created a venue through which providers could share
Medicaid waiver, and programs. their practices.

Collaboration among State Agencies
Another major impact of the Project was the
facilitation of collaboration among state agencies and
Hawaii’s request recognizes that despite the
community providers through the HICH. The HICH
expansive nature of its Medicaid program, it remains
received technical assistance from the CSH to help
with evaluating the coordination of services, with difficult to coordinate care for some individuals.
the goal of creating a more streamlined approach The high prevalence of people with complex needs
to engage various departments and agencies. The among those who are homeless is evidence that
technical assistance built stronger relationships with the service system is currently unable to provide
departments such as the Department of Health effective services to this group…. While Medicaid
and the MedQuest Division of the Department of cannot reimburse providers for the cost of rent or
Human Services in creating ways to get homeless other direct housing costs, Medicaid can provide
people housed at faster rates. The collaboration reimbursement for the health and social services
between the state and Honolulu city and county necessary to keep the person housed. Inability to
strengthened as they worked together to see finance services is often the reason new supportive
which resources complemented each other. As an housing units cannot be created. Building owners
example, the state had a list of chronically homeless and landlords are unwilling to lease units to people
individuals in Pathways who needed housing, and who are chronically homeless and have behavioral
the city and county had new housing projects that health disorders without the assurance that service
had just become available with the help of the providers will handle any problems that arise.
Governor's Emergency Proclamation. As a result of Providing Medicaid payment for services increases
working together, the Project helped house a group the capacity of service providers and allows these
of individuals, which accelerated the overall housing individuals to be housed.
placement rate. Beyond the state and city and county —Center on Budget and Policy Priorities, public
agencies, collaboration within the state among the comments submitted on Hawai‘i’s waiver
Departments of Human Services, Health, Public Safety, application.
Transportation, and Land and Natural Resources have
unfolded, and some departments have established
homeless coordinator positions as a result. As different
departments, community providers, and agencies
encounter homeless individuals, they are better able Medicaid 1115 Waiver Amendment
to link these individuals to the appropriate office to Application
get services.
The technical assistance provided through the CSH was
critical to the state’s submission of the Medicaid 1115
Demonstration Project’s amendment on September
19, 2017. The amendments, if approved, would allow
The relationships between different agencies for Medicaid to provide supportive housing services
involved, such as Med-Quest, the State for chronically homeless individuals with a behavioral
Homeless Coordinator, and ADAD, have gotten or physical illness, or a substance abuse diagnosis.
better. There’s a lot more familiarity between These services will include pre-tenancy and tenancy
each other and more trust. The relationship is support with the goal of assisting the target population
stronger so that we can support each other in to obtaining and maintaining permanent housing. At
multiple efforts and different ways—that’s one the time of writing, the Medicaid’s decision on the
good thing that has come out of this project. application remains pending. Comments submitted
—ADAD’s Project Coordinator by the Center on Budget and Policy Priorities, a
nonpartisan Washington DC-based research and
policy organization, on Hawai‘i’s application during-
System Outcomes 47
the federal public comment period provided strong
support for the application, while highlighting the Housing Plans and Financing Model
importance of this proposed amendment for Hawai‘i
to better serve those who experience chronic The state have been making progress in providing
homelessness.17 housing assistance for homeless families, making rapid
re-housing programs more available, and expanding
The process of preparing the application also provoked the Housing First program to Neighbor Islands.
important discussion among Hawai‘i MedQuest However, the biggest challenge continues to be to find
Division and contracted health plans on the training, housing for people experiencing chronic homelessness
piloting and staffing that would need to occur in order due to Hawai‘i’s tight housing market and scarce
to make the transition possible, once the amendment supply of supportive housing vouchers. The technical
to provide supportive housing services is approved. assistance from the CSH through the Pathways
The proposed amendment would better align mental Project’s funding provided Hawai‘i with a snapshot
health and homeless services that function separately, of current housing options, as well as a proposal for
but often have shared, overlapping interests. With an how to finance the development of more affordable
increased focus on housing and permanent support housing to meet the needs of the homeless population
housing programs, providers in these two service areas based on the Point-in-Time Count estimation. While
would be more likely to hone in on the intersection it would take time and resources to develop the
between the two and what is needed to better assist housing units needed, the housing planning and
the chronically homeless. financial modeling provides the necessary data for
policymakers, government agencies, and community
development organizations to continue the discussion
and determine the feasibility of a broad system change
that would house the homeless.
This is the different mix [of] white

and Asian tourists. [Here’s] what’s
happening in Hawaii.... [Tourists] think,
“Oh gee, [the homeless] could get a job
or they could do this [work].” They have
no idea how expensive it is to live here....
When they get off the flight or the boat,
[they were told,] “Don’t be giving money
to the homeless people, because all they
do is drugs.”
—Mary, client
17
Center on Budget and Policy Priorities. (2017, October 17). Public comments on Hawaii Quest’s amendment 17—supportive housing
services (ID: #311765). Retrieved from https://public.medicaid.gov/connect.ti/public.comments/ showUserAnswers?qid=1892579&vo-
teid=311765&nextURL=%2Fconnect%2Eti%2Fpublic%2Ecomments%2FquestionnaireVotes%3Fqid%3D1892579%26sort%3Drespon-
dent%5F%5FcommonName%26dir%3Dasc%26startrow%3D1%26search%3D

DISCUSSION AND
RECOMMENDATIONS
The Hawai‘i Pathways Project demonstrated a successful model in housing the hardest-to-serve population
among the homeless—chronically homeless adults with mental illnesses, addictions, or co-occurring disorders.
Many of them also suffer from other long-term physical disabilities, HIV or other chronic health conditions.
In Hawai‘i, the chronically homeless population is relatively small, about 1,600 on any given day.18 But they
represent people who have the most severe service needs, being over-represented in the criminal justice system,
and over-utilizing emergency and acute services in the health care system. Most have gone through shelters and
attempted to stay off the streets with the help of programs but none of these services seemed to work for them;
many have become reluctant to engage in services and lost their sense of empowerment. The Hawai‘i Pathways
Project service team members showed that, with time and patience, it is possible to engage and build trusting
relationships with this highly vulnerable population and to support clients in pursuing their own journey to
recovery and re-integrating into the community. The Pathways Housing First model works in Hawai‘i!
In this section, we discuss the accomplishments of the Project, critically examine the barriers and challenges to
the Housing First implementation, and offer some recommendations for the future considerations.
Accomplishments
Implementing Housing First with High Fidelity

The Hawai‘i Pathways Project was the first and only to maximize housing stability and prevent
program to implement an evidence-based Housing eviction;
First model in the state. The fidelity assessment c. Service philosophy: Delivered client-driven
conducted by the Pathways to Housing Institute services using a harm reduction approach;
concluded that Hawai‘i Pathways was implemented d. Service array: Provided supportive services and
with high fidelity. Certain aspects of housing and limited treatment services through a multi-
the Assertive Community Treatment (ACT) approach disciplinary team and supplemented treatment
received lower rating in fidelity and are discussed services by connecting clients with other
in the barriers section. Key features of the program community-based providers;
implementation were: e. Program structure: Delivered services primarily
through home visits on a weekly basis,
a. Housing structure: Offered scattered-site conducted team meetings 2‒3 times a week
housing, with clients paying 30% of income for to discuss client needs and service plans, and
rent; adopted an individual (versus shared) caseload
b. Separation of housing and services: Offered approach to case management.
housing without preconditions for treatment
or sobriety, and with supportive services
18
Bridging the Gap and Partners in Care. (2017). State of Hawaii homeless point-in-time count. Retrieved from http://www.partnersincar-
eoahu.org/sites/default/files/2017%20Statewide%20PIT%20Report%20-%20Full%20Report%20-%20FINAL.pdf
Discussion And Recommendations 49

Achieving High Housing Retention Rate Reducing Costs
The Pathways Project served the most vulnerable Housing takes a fundamental role in our physical and
group of homeless people with complex behavioral psychological well-being. From baseline to follow-
health issues. About half of clients had been homeless up, Pathways clients reported a 53% reduction in the
for a continuous period of six years or more, and 30-day health care utilization. However, the decrease
almost three-quarters were diagnosed with SMI was found to be statistically significant (p<0.05) only
(serious mental illness) or co-occurring SMI and among housed clients and not among un-housed
addiction disorders. The Project successfully placed 99 clients. The reduction in health care utilization among
clients into permanent housing units using vouchers the housed clients validated the association between
obtained from permanent supportive housing housing and health care. After living in stable and
programs on O‘ahu. At the end of the grant period, safe housing with supportive services, the estimated
the Project achieved a 90% housing retention rate. 30-day health care costs of Pathways clients dropped
The average length of housing was 13.9 months, with 76%, from an average of $8,162 per client at baseline
a range of 0.1‒34.2. Shorter lengths of housing were to $1,965 per client at follow-up, representing a
due to clients being placed in the third year of the reduction of $6,197 per client. The largest drop was
Project. The service team was committed to relocating found in the costs of hospital stays, followed by ER
clients. Of those who remained housed, 13 were visits and outpatient visits. Considering the costs of
relocated one to three times due to lease violations, providing permanent supportive housing for Pathways
illegal or drug related activities, client’s choice, or poor clients were $1,950 per month per client, which
physical housing conditions. included $1,100 for rental subsidies and $850 for
Pathways services, the estimated net cost savings was
$4,247.
Transforming Clients’ Lives

Stable housing transformed the lives of those who
Filling the Service Gap
experienced chronic homelessness and complex
In the current system, intensive case management is
behavioral health issues. Personal stories shared
available to people with serious mental illness (SMI)
through the PhotoVoice interviews and the focus
through Medicaid’s Community Care Services (CCS)
group interviews conducted by the Pathways to
program. The majority of Pathways clients had similar
Housing Institute validated the impacts of the Hawai‘i
diagnosis as those served by the CCS program. In
Pathways Project on participants’ lives. There were
fact, many of them were enrolled in the CCS prior to
many accounts of rebuilding relationships with adult
becoming Pathways clients, indicating that there were
children, slowly regaining self-dignity, feeling safe and
still some service gaps that needed to be met. People
peaceful, taking small steps to get well, thriving in
suffer behavioral health issues other than SMI are not
keeping their apartment clean and neat, and searching
eligible for the CCS program. While they may have
for volunteer and work opportunities. From baseline
access to a lower level of care-service coordination,
to follow-up, there were statistically significant
it is not enough to meet their needs due to the
increases (p<0.1) in housed clients who reported,
additional challenges they have from being homeless
in the past 30 days, of not being physically hurt;
for a very long time, such as having difficulty in
abstaining from alcohol or illegal drugs; and not having
keeping track of appointments and lacking motivation
experienced drug-use related health, behavioral, or
to get well. By providing intensive case management
social consequences. They also reported significant
that included tenancy support, the Hawai‘i Pathways
decreases (p<0.1) in non-drug-related psychological
Project showed that even the most difficult clients of
or emotional problems; alcohol and drug use; and
the hard-to-serve population were able to transition
decreases in drug-related health, behavioral and social
successfully from being chronically homeless to stably
consequences. The number of clients reporting good,
housed.
very good or excellent health increased almost half at
follow-up.

Accelerating System Change
The Hawai‘i Pathways Project was piloted during a chronic homelessness. This Project also impacted the
critical time when the state’s homeless service system system by sharing the evidence-based practices with
was undergoing a paradigm shift in adopting a Housing other organizations providing permanent supportive
First approach to address chronic homelessness. housing services. Through client advocacy, the Project
The involvement of the Hawai‘i Interagency Council worked with the Interagency Council to facilitate
on Homelessness, through the State Homeless collaboration across government agencies in housing,
Coordinator who chaired the council, ensured that this health, social service and criminal justice with the goal
pilot program was fully integrated into the larger effort of stopping the revolving doors in the current service
of the state’s response to homelessness. The State systems. At the program and policy level, the Project
Coordinator’s participation contributed to shaping consulted with the Corporation for Supportive Housing
the grant application that aimed to address service to identify unmet service needs to be addressed via
gaps in the system and to bring the Housing First Medicaid Demonstration Project and to complete a
approach to the center stage of policy discourse on supportive housing and financial plan for the state.
Barriers and Challenges
Grant Administration
The first challenge faced by the Project was the request was not approved by SAMHSA. When the
lengthy start-up period involved with implementing one-year extension approval finally came through in
federal grants in Hawai‘i. From the start of the grant the last month of the original grant period, September
period in October 2013 to the first referral received 2016, the uncertainty of future employment had
by the Project in August 2014, 10 months were already led to several staff members leaving the
spent on establishing the Project at the Alcohol and Project prematurely. Furthermore, the service contract
Drug Abuse Division (ADAD) of the Department of was initially extended for six months only, which added
Health, executing the service contracts, and hiring to the challenges of the service team to recruit and
key personnel for Project coordination at ADAD and retain staff during the last year of implementation. In
the service team at the Helping Hands Hawai‘i and short, grant administration delays from various sources
Catholic Charities Hawai‘i. The delays had detrimental affected the lower-than expected program enrollment
effects on the Project. To meet the grant’s enrollment and the Project’s ability to recruit staff.
expectation, the Project was given five months to
catch up with the enrollment goal set for the first
15 months. While the goal was achieved, the shift Workforce Availability
of the service team’s effort to focus on enrollment
caused delays in providing housing and supportive Workforce availability was a major challenge
services to clients. In the subsequent months, the experienced by the Hawai‘i Pathways Project. The
team was able to refocus on housing placement, but service team had a total of 16 full-time equivalent
the opportunity to build rapport with some clients (FTE) positions, however only eight positions
right after enrollment was lost, which contributed to were filled on average throughout the three-year
a number of clients (21) who lost touch, were unable implementation period. Certain positions were more
to be located, or declined services later on. Due to difficult to fill than others due to required work
the compressed Project period, after less than two experience or unique qualifications. There were two
years of program implementation, the service team major negative impacts on the Project due to this.
had to stop enrolling clients and to shift their effort First, some functions of the Assertive Community
to transition/discharge planning in case the extension Treatment (ACT) were restricted because some

positions were filled partially; the plan for establishing vouchers were out of the Project’s control, coming into
a Consumer Advisory Board had to be put aside due the Project in small installments throughout the three-
to the lack of peer navigators to coordinate the effort; year period. This caused the Project to halt housing
and the total enrollment reached only 86% of the placement periodically when vouchers were out.
goal. Secondly, the staff members had to take on an The process was further complicated by the special
increased number of clients and responsibilities, due conditions attached to certain vouchers, such as those
to the looming pressure of the enrollment target. In designated for veterans, people with HIV/AIDS, and
order to improve the team’s efficiency when serving specific geographic placements, which the pool of
clients, the team adopted an individual caseload clients waiting for housing did not meet. While it took
approach, rather than a shared caseload approach time to look for apartments and engage landlords,
that was required by the ACT model. To address the service team did not think these processes caused
staff shortage, the Project sub-contracted with other delays in housing placement. A third challenge was
providers to offer additional services in peer coaching housing choices. O‘ahu’s tight housing markets and
and housing navigation. In addition to the recruitment the inadequate supply of low-income apartment units
challenges, the Project also faced with setbacks due limited the housing choices clients had in location and
to staff turnover that occurred both within the service other features of their housing.19 Nevertheless, most
team and at the grant administration level at ADAD. clients were satisfied with their housing and were
At the system level, the change of state administration appreciative of the assistance received in furnishing
after the 2014 gubernatorial election resulted in their apartments.
a change of leadership and membership for the
Interagency Council. Coupled with the staff change at
ADAD, the relationship between the Project and the
council went through a short period of uncertainty
Treatment and Recovery
before the partnership regained its strength in the last
year of the Project. Treatment and recovery is a long journey. During
the relatively short time that the Project’s service
team had in working with the clients, significant
Housing Placement improvements were observed in the physical and
psychological well-being of clients after they were
The Hawai‘i Pathways Project placed 99 clients in housed. The use of a client-centered harm reduction
permanent housing with about half of these clients approach by the service team contributed to a
housed within six months upon program entry, significant reduction in the use of illicit drugs among
including 38% placed within four months and met clients from baseline to follow-up interviews; however,
the benchmark of the Pathways Housing First model. the same impact is yet to be seen in alcohol abuse
While these could be considered great achievements, as many clients continued to self-medicating with
the Project was not able to house the remaining 35 alcohol. Besides, fewer clients reported participation
clients due to the limited availability of permanent in self-help and recovery groups or meetings after
supportive housing vouchers on O‘ahu. During the enrolling in Pathways, which was an unexpected
Project’s implementation period, the scarcity of outcome due to many clients turning to the Project for
housing vouchers was evident in the fact that about recovery support. Hawai‘i’s current system is not set
900 people experienced chronic homelessness on any up for people to get addiction treatment on demand.
given day on O‘ahu but only about 220 permanent While the service team utilized a variety of techniques
supportive housing vouchers were available each year to engage clients for treatment and recovery, only
through turnover in existing programs funded by HUD a few clients were admitted to treatment programs
and VA, and through new Housing First programs during the Project period.
funded by the state and the city. A second challenge
was in quickly placing clients into housing when the
19
Tsemberis, S., & Walker, J. (2017).

Recommendations
Several recommendations as a result of the Hawai‘i Pathways Project pilot are provided and include expanding
the Housing First program; prioritizing the needs of chronically homeless individuals and allocating appropriate
resources for services; developing a Housing First learning community; and addressing the needs for positive
social inclusion.
Expand the Housing First Program Develop a Housing First Learning

Community
Expanding the Housing First program would assist
more of Hawai‘i’s chronically homeless individuals A learning community composed of Housing First
in getting them off the streets and reintegrated team members and experienced Housing First support
into the community. Housing subsidies and support professionals could assist in addressing staff shortages,
services are the two main components of Housing adhering to program fidelity, improving outcomes, and
First programs and would need to be funded at the providing professional support. A learning community
levels appropriate to the needs of the participants. can take the form of monthly meetings or conference
As Housing First has been proven to be effective, calls that create venues for all Housing First teams
expanding the program could go a long way in assisting to meet and share their practices, challenges and
this population, ultimately saving public costs, solutions, which will benefit their work and prevent
especially on healthcare. In addition, the Medicaid burnout. With the Housing First program being
Demonstration’s amendment, if approved, will extend new to the state, many staff may rely on their past
housing support services to the chronically homeless knowledge and experiences based on past traditional
individuals with a behavioral or physical illness, or a housing processes. The learning community will help
substance abuse diagnosis. With a growing program, to educate staff and help them to deprogram their
it is essential that state and counties are equally thinking about the traditional housing system in order
committed to its effective implementation and that to maximize the benefits of the Housing First program.
care coordination among programs, providers and Professional supports, from technical-assistance
other related parties be strong. consultants for example, could provide needed
training to strengthen the effectiveness of the team
and program.
Prioritize Needs
Address the Needs for Positive Social
The Housing First program is a costly intervention, Inclusion
but when it is implemented effectively to target the
hardest-to-serve people among those experiencing Reintegrating those who were chronically homeless
chronic homelessness, it is proven to save significant back into the community requires positive social
public costs compared to providing services in past inclusion. Part of this inclusion is to address the
traditional ways. The use of the common screening common resulting loneliness of homelessness and
tool, VI-SPDAT, is the first step but should not be the to assist these individuals in finding meaningful
only method for prioritization. It is important that volunteer and other social engagement activities
providers and clients alike provide input on how to and opportunities. Furthermore, opportunities for
prioritize needs and how to determine best allocation clients to be included in decision-making about the
of appropriate resources for services. Tracking changes future directions of the Housing First program should
in needs, resources expansion, and program outcomes be provided, perhaps through the establishment of
will inform policy and programs to ensure public a Consumer Advisory Board. They are the ones who
resources are used responsibly. know best about being chronically homeless and can
inform services that work best in lifting individuals out
of homelessness.

Bb9f9 Hawaii Pathways Final Report

Uploaded by

Copyright:

Available Formats

Bb9f9 Hawaii Pathways Final Report

Uploaded by

Document Information

Original Description:

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

Bb9f9 Hawaii Pathways Final Report

Uploaded by

Copyright:

Available Formats

HAWAI‘I

2 Hawai‘i Pathways Project: Final Report

About the Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2. Referral, Admission and Discharge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

5. Treatment and Support Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

6. Permanent Housing Placement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

9. Involvement in the Criminal Justice System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

10. Progress in Personal Recovery Goals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

11. Health Status and Health Care Service Utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

13. Impacts on Programs and Policies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Discussion and Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Barriers and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Hawai‘i Pathways Project: Final Report 3

4 Hawai‘i Pathways Project: Final Report

Hawai‘i Pathways Project: Final Report 5

Overview mental health, or co-occurring disorders. The Project

6 Hawai‘i Pathways Project: Final Report

8 Hawai‘i Pathways Project: Final Report

While the adoption of a Housing First approach to

About the Project

10 Hawai‘i Pathways Project: Final Report

OBJECTIVES INPUTS ACTIVITIES OUTPUTS OUTCOMES

• Place chronically • Administration: • Implement • Offered a Client Level System Level

12 Hawai‘i Pathways Project: Final Report

Table 1: Evaluation Areas

1. Program fidelity and training 7. Housing stability 12. Cost reduction

6. Permanent housing placement

Table 2: Data Collection Methods

These are my slippers. They symbolize for me for

14 Hawai‘i Pathways Project: Final Report

1. Program Fidelity and Training

Referral and Admission

Figure 2: Numbers of Referrals, Admissions, and Target Enrollment

Table 3: Referral Sources

Number of referrals 191 23 30 12 6 262

% of total referrals 73% 9% 11% 5% 2% 100%

% of referrals admitted 57% 43% 27% 50% 33% 51%

16 Hawai‘i Pathways Project: Final Report

Figure 3: Length of Admission Process

60 interest in being housed, but

7.0 Other cont Face‐to‐face contacts

2.0 4.2 4.0

Figure 5: Reasons for Denial

The remaining 108 clients (81%) were all discharged

18 Hawai‘i Pathways Project: Final Report

Table 5: VI-SPDAT Score and Homeless Service Utilization History

Total 34 100 134 100%

Table 6: Length of Homelessness

1‒2 years 23%

Behavioral Health Status

20 Hawai‘i Pathways Project: Final Report

Note: Diagnosis was reported in ICD-9 codes.

Table 8: Substance Use and Mental Health Disorders among Veterans

Serious mental illness (SMI) only 5 20%

Co-occurring substance use and SMI 12 48%

Here’s the bus stop on my next station where every

ICM – CSAC, other 0 1 3 2

22 Hawai‘i Pathways Project: Final Report

Figure 6: Service Team Positions Available and Filled