This study examines the influence of group or individual housing placement and consumer characteristics on the number of days subsequently homeless among formerly homeless mentally ill persons.

OBJECTIVE: The study examined the influence of group or individual housing placement and consumer characteristics on the number of days subsequently homeless among formerly homeless mentally ill persons.

METHODS: A total of 303 homeless shelter residents with severe mental illness were screened for dangerousness, 118 were randomly assigned to either independent apartment or staffed group living sites, and 110 were followed for 18 months. Study participants' sociodemographic characteristics, diagnosis, and residential preferences and the residential recommendations made by clinicians were measured at baseline.

RESULTS: Overall, 76 percent of the study participants were housed at the end of the 18-month follow-up period, although 27 percent had experienced at least one episode of homelessness during the period. The number of days homeless was greater for individuals assigned to independent apartments than for those placed in staffed group homes, but only for members of minority groups. Substance abuse was the strongest individual-level predictor of days homeless. Individuals whom clinicians identified as needing group living experienced more days homeless, irrespective of the type of housing they received. Consumers who stated a strong preference for independent living had more days homeless than those who were amenable to staffed group homes.

CONCLUSIONS: Although consumers more frequently prefer independent living, placement in staffed group housing resulted in somewhat fewer days homeless for some groups of consumers. Further experience of homelessness by formerly homeless mentally ill individuals may be reduced by providing effective substance abuse treatment and by paying special attention to consumers identified by clinicians to be at particular risk for housing loss. (Authors)