Housing Stability and Recovery Among Chronically Homeless Persons With Co-Occuring Disorders in Washington, DC

Abstract

Individuals who remain chronically homeless frequently suffer debilitating effects of serious mental illness and addiction. These frequently co-occurring disorders represent an extremely difficult hurdle for individuals to overcome. Typical housing programs demand sobriety and compliance with psychiatric and behavioral treatment as a condition of admission and continued enrollment. A more realistic, compassionate, and effective approach provides housing without such prerequisites and instead provides immediate access to permanent housing with supports as a foundation for recovery.

PATHWAYS HOUSING FIRST

The Pathways Housing First (PHF) program has demonstrated its effectiveness for individuals with extensive service needs^1,2 and now provides housing and intensive community-based treatment to chronically homeless individuals in Washington, DC. Fundamental needs of housing and support services for this project were provided through funding by the federal Department of Housing and Urban Development (HUD). The program uses a client-centered psychiatric rehabilitation and harm reduction approach to engage, permanently house, and provide Assertive Community Treatment (ACT) rehabilitative services to clients presenting with co-occurring disorders who have been homeless at least 5 years—in short, the most vulnerable and challenging individuals to engage and serve.

A description of the ACT team approach is available elsewhere.³ It calls for a 10-to-1 ratio of clients to providers, who include psychiatrists, nurses, addiction and employment counselors, and peer support specialists. PHF provides community-based services, and a service coordinator is always on call to help clients address emergency needs. Individuals are enrolled in the program on a first-found, first-served basis and begin with intake, psychiatric assessments, and initiation of public benefit applications (e.g., Medicaid, Social Security Disability Insurance, food stamps). Individuals are immediately housed in affordable, scattered-site apartments of their choice that meet federal housing quality standards. Client choice determines the type, intensity, and frequency of treatment and support services provided. Each client receives at least 1 weekly visit by the team and must fulfill the responsibilities of a standard lease.

The HUD program is 1 of 10 sites across the United States testing various rapid housing strategies with individuals meeting its need severity criteria of alcohol dependence and homelessness history. At this site, HUD provided a $1035 subsidy per client toward fair market monthly rent and $55 per client toward the cost of monthly housing support services. The remaining support and treatment services delivered were paid for by Medicaid and other local funding. Clients pay 30% of their total disability income toward rent. The remainder of housing and treatment costs are thought to be offset through savings of high-cost public services that would otherwise be consumed by this population.^4,5

Advocates argue that the PHF approach is a clinically successful and cost-effective strategy to end homelessness. A recent review of ACT programs concluded that evidence of the model's efficacy justifies its broader use,⁶ the federal Department of Health and Human Services, Substance Abuse and Mental Health Administration added PHF's approach to its national registry of evidence-based programs and practices,⁷ and its merits have been argued before policymakers.⁸

DISCUSSION AND EVALUATION

During the demonstration period, PHF operated at a full capacity of 36 individuals, all of whom were followed for 2 years. Of the clients, 83% were male; 58% were Black and 31% were White; 72% were diagnosed with schizoaffective disorder or schizophrenia and 25% with bipolar disorder; 100% were alcohol dependent; and 36% were aged 25 to 45 years and 64% were aged 46 to 63 years at intake. At the end of the first year, 97% of clients remained housed, decreasing to 84% after the second year (Table 1), with 14% requiring much less intensive and costly community support services. A thorough understanding of reasons for early client exit and differences among clients’ rates of improvement could be helpful in resource planning and in refining the selection criteria.

Predicted individual outcomes included reduced psychological distress, reduced consumption of alcohol, and improved recovery within 2 years. We conducted analyses to determine whether the average difference between client scores at 2 points in time was significantly different from zero. A comparison group was not feasible because of the stringent severity of the need enrollment criteria and immediate provision of services upon client engagement.

Open in a separate window

Formerly homeless tenant holding his apartment key. Photograph by Linda Kaufman

We collected evaluation data from client records and separate interviews with clients and ACT team members. We measured client self-report of psychological distress with Brief Symptom Inventory and Global Severity Index raw scores (Table 2).⁹ Clients who have had many contacts with service providers, sometimes over decades, are often wary of questions about both psychotic symptoms and alcohol consumption. Consistent with the PHF policy of no demands on addiction behaviors and assistance with disability income budgeting, we asked clients how much money, if any, they spent on alcohol in the past 2 weeks. As a check on self-report of a question that might elicit a response colored by previous experience with service organizations, we asked a staff member who worked closely with an individual client to estimate that client's use of alcohol and its impact. We computed an interaction score for use and impact responses on a Likert-type scale, which yielded an alcohol impact score.

We did not ask staff members for a baseline estimation of alcohol use and impact on the client's individual functioning at intake because they had too little client experience to formulate such estimates. We administered a Ridgeway Recovery subscale comprising 24 averaged items, which yielded a score with higher values presenting thoughts, feelings, and activities consistent with mental illness and addiction recovery.¹⁰ As expected, after 2 years we observed reductions in psychological distress and alcohol impact, as well as higher recovery scores. Regression to the mean was a potential threat to internal validity, especially in light of the severity of need enrollment requirement, but the strong impact of the sense of safety and belonging that clients may have gained from living in their own homes should not be underestimated. We attribute our results to the combined effects of housing and client-centered services.

The program could benefit from integrating physical and mental health services as recommended by the National Coalition for the Homeless¹¹ and incorporated into ACT teams, as advocated the National Association of State Mental Health Program Directors.¹² Clients served are likely to be affected by acute and chronic conditions requiring treatment to reduce interference with psychiatric treatment, and vice versa. Such services would not threaten the sustainability of the program if they incorporated a nurse practitioner from a federally qualified health center staff, reimbursed through each client's public benefits to treat known conditions, screen for unknown conditions, and provide disease prevention and health promotion information.

Outcome results from this and other demonstration sites will provide additional data about the impact of the PHF strategy. This approach could be widely disseminated because the model has been documented with measures of fidelity to assess its use and effectiveness in other client types and locations.¹³ Continued reporting of Housing First results should increase understanding and acceptance of the model to provide a solution to ending homelessness for people with complex needs.

Acknowledgments

Human Participant Protection

References