A modified TC for homeless MICAs in a shelter: An outcome study

Abstract

This study sought to compare outcomes of a modified therapeutic community (TC) in a shelter for homeless mentally ill chemical abusers (MICA) clients, with a general shelter for veterans, with a subgroup of MICA clients. The study attempted to answer the following question: Are specialized shelter programs for homeless MICA persons more effective than general shelter programs in the area of length of stay, stabilization (i.e., medication compliance and sobriety), and housing placement? Seven hypotheses were formulated based on the research question and sub-questions. The hypotheses are summarized as follows: greater effectiveness in the modified TC would be demonstrated in the area of longer periods of sobriety, fewer days of psychiatric hospitalizations, shorter lengths of stay in a shelter; positive discharge from the shelter, medication compliance, housing placement within the first year, and housing placement.;The study utilized a quasi-experimental design involving a retrospective review of closed case records that compared outcome data on length of stay, stability, and housing placement. The sample size was 140, with 70 subjects in each group. The modified TC shelter was the control group, and the general shelter served as the comparison group. Subjects were selected based on housing placement and discharge criteria from September 1, 1998--June 1, 2000 for the modified TC shelter, and from June 1, 1999--June 1, 2000 for the general shelter. Data were collected on independent (demographic) and dependent (outcome) variables.;Parametric and non-parametric tests (t-test and chi-square) were the primary statistical analyses conducted. The two groups tested different by age, length of homelessness, years of education, years of work experience, veteran status, marital status, and psychiatric diagnosis. Veteran status had to be controlled because of the disparity in numbers of veterans in the general shelter (100%) and veterans in the modified TC shelter (25%). Of the seven hypotheses, three were supported statistically significant at p ≤ .05 with a one-tailed analysis of significance. The supported hypotheses included discharge status, overall medication compliance, and housing placement. Although four hypotheses were not supported, the fact that the modified TC shelter subjects had more deficits, and the results were parallel to the general shelter despite controlling for veteran status, the results indicated that overall, the modified TC milieu was more effective in discharge status, housing placement, and medication compliance.