This two year project will focus on one of the least studied aspects of delivering care to persons with a dual diagnosis of serious substance abuse and mental health disorders- the process of creating and maintaining service linkages. Recent substance abuse services research findings are consistent in two regards: (1) this population is at an increased risk of negative life events and represent one of the, if not the, costliest populations to serve, and (2) successful intervention and prevention strategies must address a broad range of needs clustered in this population. When the situation of homelessness is present, meeting the needs of this population is exacerbated. To reduce the revolving door service pattern common among the homeless with substance abuse and mental health disorders, participation of providers is required. In the organizational literature, these across-service-sector linkages requisite to deliver needed services, are problematic. There is little literature informing service providers or policy makers on what factors encourage the creation or maintenance of service linkages for multi-need populations. This study will use interorganizational network methods and a dyadic analysis of relationships overtime between substance abuse and mental health programs in six cities across the US to investigate what factors predict the creation, and maintenance of, service linkages around dually-diagnosed homeless. Because the current service environment faces several socio-economic movements not conducive to the survivability or creation of service linkages among public service providers of chronic populations, these finding become even more critical. The project will use longitudinal, multi-site data collected as part of a national evaluation of the ACCESS demonstration project and a primary data collection to build the appropriate relational models. Findings will be related in terms of their implications for program managers and directors trying to build more responsive systems of care for this vulnerable population.
