Research in supported housing.

BACKGROUND: De-institutionalization has led to the provision of various forms of housing with or without support for people with mental illness in the community. In this paper, we review the conceptual issues related to the provision of supported housing schemes, the characteristics of residents, research methods and outcomes, and the factors influencing the quality of care provided.
METHODS: A Medline and hand search of published literature was complemented by information derived from contacting expert researchers in the field.
FINDINGS: There is considerable diversity of models of supported housing and inconsistent use of terminology to describe them. This makes it difficult to compare schemes, processes, and outcomes. Patients in supported housing are characterized by deficits in self-care and general functioning, whilst behavioral problems such as violence, drug abuse and extreme antisocial habits predict exclusion from supported housing. Most evaluative studies are merely descriptive. In terms of outcomes, it seems that functioning can improve, social integration can be facilitated, and residents are generally more satisfied in supported housing compared with conventional hospital care. Further evidence suggests that most patients prefer regimes with low restrictiveness and more independent living arrangements, although loneliness and isolation have occasionally been reported to be a problem. Little information is available on the factors that mediate outcomes and on skills required by staff.
CONCLUSION: Research in supported housing for psychiatric patients has so far been neglected. Large scale surveys on structure, process, and outcomes across a variety of housing schemes may be useful in the future to identify some of the key variables influencing outcomes. The use of direct observation methods in conjunction with other more conventional, standardized instruments may also highlight areas for improvement. In conducting research, structure and process, as well as outcomes, need to be considered. Thus, we need to know not just what to provide, but how to provide it in such a way that it will maximize beneficial outcomes. This represents a considerable research agenda.