Closing service system gaps for homeless clients with a dual diagnosis: integrated teams and interagency cooperation.

BACKGROUND: There is great concern about fragmentation of mental health service delivery, especially for dually diagnosed homeless people, and apprehension that such fragmentation adversely affects service access and outcomes.

AIMS OF THE STUDY: This study first seeks to articulate two alternative approaches to the integration of psychiatric and substance abuse services, one involving an integrated team model and the other a collaborative relationship between agencies. It then applies this conceptualization to a sample of dually diagnosed homeless people who participated in the ACCESS demonstration.

METHODS: Longitudinal outcome data were obtained through interviews at baseline, 3 months, and 12 months with homeless clients with a dual diagnosis (N = 1074) who received ACT-like case management services through the ACCESS demonstration. A survey of ACCESS case managers was conducted to obtain information on: (i) the proportion of clients who received substance abuse services directly from ACCESS case management teams, and the proportion who received services from other agencies; and (ii) the perceived quality of the relationship (i.e. communication, cooperation and trust) between providers--both within the same teams and between agencies. Hierarchical linear modeling was then used to examine the relationship of these two factors to service use and outcome with mixed-model regression analysis.

RESULTS: Significant (p<.05) and positive relationships were observed in 4 of the 20 analyses of the association of service use and measures of communication, cooperation, and trust (either intrateam or inter-agency) while none were significant and negative. At 12 months, receipt of a higher proportion of services from agencies other than the ACCESS team was associated with fewer days homeless, and greater reduction of psychiatric symptoms, contradicting the hypothesis that integrated team care is more effective than interagency collaborations.

DISCUSSION AND LIMITATIONS: This study broadens the conceptual framework for addressing service system fragmentation by considering both single team integration and interagency coordination, and by considering both program structure and the quality of relationships between providers. Data from a multi-site outcome study demonstrated suggestive associations between perceptions of communication, cooperation and measures of clinical service use. However, the proportion of clients treated entirely within a single team was associated with poorer housing and psychiatric outcomes. These empirical results must be regarded as illustrative rather than conclusive because of the use of a non-experimental study design, imperfections in the available measures, and the incomplete sampling of case managers.

IMPLICATIONS FOR HEALTH POLICY: This study suggests that fragmentation of services for dually diagnosed clients may be reduced by improving the interactions within and between agencies providing these services. While primary emphasis has been placed on developing integrated teams, interagency approaches should not be prematurely excluded.

IMPLICATIONS FOR FUTURE RESEARCH: Research on approaches to reducing system fragmentation have focused on either global efforts to integrate numerous agencies in a community or highly focused efforts to develop specialized teams. Future research should also focus on the possibility of fostering constructive relationships between selected pairs or subsets of agencies. Research in this area will also benefit from the further development measures of team integration and of both intra-team and inter-agency communication, collaboration, and trust.