Evaluation of the CMHS ACCESS Demonstration Project for Homeless Persons with Mental Illness
Joseph P. Morrissey, Ph.D., Principal Investigator
The Access to Community Care and Effective Support Services (ACCESS) initiative was a six year (1994-2000) Federal program of the Center for Mental Health Services designed to support demonstrations of services integration in 15 cities at 18 sites across the United States. It built on the efforts over the past two decades of service organization funding agencies and researchers to design and evaluate service models that effectively engage and treat homeless individuals with severe mental illness. The overall goal of the evaluation was to conduct an experimental evaluation of system integration strategies at 18 sites: 9 receiving funding to develop ACT services plus system integration activites (experimental sites) and 9 to develop ACT services only (comparison sites).
Interorganizational network analysis of mental health, substance abuse, primary care, housing, and entitlement service agencies at four time points (1994, 1996, 1998, 2000) provided an empirical basis for assessing the integration of services at each site. In addition to assisting with the design evaluation, researchers at UNC-CH were responsible for training and monitoring interviewers in the data collection process, with the processing and management of the data once collected, and with the analysis and reporting of the network data evaluation results. The ACCESS interorganizational evaluation was part of the system-level evaluation led by Howard Goldman, Principal Investigator, R.O.W. Sciences, Inc., Bethesda, MD. Several publications are available from this project.
Findings revealed that experimental sites did not achieve higher levels of systems integration than did the comparison sites. However, the grantee agency at each site (project integration) was much more integrated with the service system in experimental vs. comparison sites. Ultimately, neither system integration or project integration improved client outcomes. The policy implication is that investments in overall system integration do not yield cient-level improvements. Rather, investments in ACT-like services at the program integration level are needed to achieve the goal of improved client functioning.