Performance Improvement 1995. Making a Difference: Interim Status Report of the McKinney Research Demonstration Program for Homeless Mentally Ill Adults



This report synthesizes the findings from five separate demonstration projects that evaluated the effect of offering case management and housing to the homeless mentally ill. Early findings indicate general success at reducing homelessness and at improving mental health of study participants: homeless adults with severe mental illness were willing to use accessible services targeted to their needs. The factors contributing to the success of the project include mental health treatment, receipt of entitlement income, and reliance on an interdisciplinary team of staff, including some with mental illness who were formerly homeless. These findings offer strategies for the design and management of programs for the homeless mentally ill.


The goal of the 1987 Stuart McKinney Act was to create programs to reduce homelessness. Among numerous provisions, the Act authorized demonstration grants to examine new approaches to relieving homelessness. This project was designed to explore new ways of reducing homelessness and improving the mental health of homeless people by offering distinct combinations of case management and housing to address one of the most challenging problems facing our cities.


An estimated 600,000 people are homeless in the United States, based on 1987 estimates. One-third, or 200,000, are homeless adults with severe mental illness. Besides suffering from schizophrenia, depression, or other mental disorders, they commonly are afflicted by substance abuse, tuberculosis, and HIV. This constellation of mental and physical conditions is believed to place the homeless mentally ill beyond the reach of the conventional social service and health care system.

Launched by the National Institute of Mental Health in 1990, the McKinney Research Demonstration Program for Homeless Mentally Ill Adults was funded for 3 years with a total of $16.8 million. The program was transferred to the Center for Mental Health Services in the newly created Substance Abuse and Mental Health Services Administration as a result of the ADAMHA Reorganization Act of 1992.


A total of 896 homeless adults with severe mental illnesses were studied at five sites--Boston, Baltimore, San Diego, and two sites in New York City. Fifty-seven percent of the participants had psychotic disorders (e.g., schizophrenia) and 32 percent had affective disorders (e.g., depression). More than one-third were living in shelters, slightly less than one-third were living on the street, and the remainder were living in community settings, hospitals, or jails. Across all sites, the underlying goals were to offer a combination of housing and social services that would train participants in daily living, link them to needed social and medical services, help them receive entitlements for which they are eligible, and assist them in securing housing. Study sites varied in intensity, duration, or approach. The demographic portrait that emerged was sufficiently similar to the general homeless population that the final results are expected to have widespread applicability to thousands of communities throughout the Nation.

Interventions tested varied across sites according to the level of case management, the type of housing, and the degree of client engagement with the social service or health care system at the outset of the study. The comparison group of homeless mentally ill at most sites received traditional services. Study participants were randomly assigned to the new services or to traditional services.

Five broad outcome measures unified the analysis across study sites--psychiatric symptomatology, substance abuse, quality of life, physical health, and residential stability. Participants were interviewed at several stages (at baseline and at followups at 6 months, 9 or 12 months, and 18 or 24 months) to determine their progress on these measures. They were given fairly uniform questions along with a standard battery of widely accepted measures of mental health status. Some of the interview data were corroborated by case manager reports. While investigators are performing separate analyses at each site, additional analyses are being undertaken across sites. The effectiveness of the interventions in reducing costly hospitalizations and other health expenditures was examined at two study sites.


Results are still preliminary, covering the 6- to 12-month followup to the intervention. Early findings reveal that each site's intervention was successful at reducing homelessness and improving the mental health of study participants. These preliminary findings, expected to be sustained when the study is concluded, should also hold true when the findings from all sites are combined.

These general findings were amplified by many discrete, yet interrelated, findings. Most important, it was found that the homeless who are mentally ill were willing to use accessible services targeted to their needs, repudiating the belief that these people are beyond help. When they did take advantage of mental health treatment, there was a decrease in psychiatric symptoms and in the use of costly inpatient hospitalization. In one of the projects, inpatient days were halved.

Another finding was that targeted services decreased homelessness, making permanent housing an attainable goal--as long as there were appropriate levels of support. A critical period for intervention was during the transition from the shelter to community-based housing. The receipt of entitlement income was another vital element to realizing study goals. At one of the sites, the intervention yielded a two- to threefold increase in the number of participants being supported by Supplemental Security Income and Social Security Disability Income.

Formerly homeless people with mental illnesses also were found to be a valuable staff resource: they staffed four of the five sites, helping to identify and engage participants, sensitizing professional staff to their needs, and acting as role models.

Substance abuse, which was widespread among participants when the study began (47 to 78 percent), exacerbated homelessness. Participants' substance abuse was felt by clinical staff to be more significant than their mental illness in preventing them from finding or keeping housing.

Use of Results

Emerging policy recommendations thus far integrate service systems at all levels for the homeless mentally ill, bringing together mental health, substance abuse treatment, social services, and the criminal justice system; emphasize substance abuse treatment as an integral part of mental health services; offer clients a greater range of housing options; and provide preventive health care and education to reduce morbidity and mortality from severe illnesses experienced by this population.


More than 30 publications had emanated from the study at the time of this report, in such journals as Community Mental Health Journal, Psychiatric Services, Hospital and Community Psychiatry, Psychosocial Rehabilitation, American Journal of Psychiatry, Evaluation Bulletin, and Schizophrenia Bulletin.

Agency sponsor:

Substance Abuse and Mental Health Services Administration

Federal contact:

Roger Straw, Ph.D.
Center for Mental Health Services
Substance Abuse and Mental Health
Services Administration
Parklawn Building, Room 11C26
5600 Fishers Lane
Rockville, MD 20857
(301) 443-3606 Fax: (301) 443-0541

Principal investigators:

Anthony Lehman, M.D., The Baltimore Project, Baltimore, MD

Stephen Goldfinger, M.D., The Boston Project, Boston, MA

David Shern, Ph.D., The New York City Street Outreach Project, New York, NY

Elie Valencia, J.D., The New York City Critical Time Project, New York, NY

Richard Hough, Ph.D., The San Diego Project, San Diego, CA

View full report


"PerformanceImprovement1995.pdf" (pdf, 941.92Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®