Introduction
In 2003 the Department of Housing and Urban Development (HUD), the Department of Health and Human Services, and the Department of Veterans Affairs initiated a major service demonstration, the Collaborative Initiative to Help End Chronic Homelessness (CICH). This jointly funded $35 million demonstration represents an extension of efforts to integrate services for homeless people fostered for many years by HUD's Continuum of Care initiative. CICH focused on improving outcomes for chronically homeless people by making funding available to provide five core services at each site: (1) permanent supportive housing, (2) mental health treatment, (3) substance abuse treatment, (4) primary health care, and (5) veteran health services. In this study we utilize data from the 11 communities in which CICH was implemented to examine four questions reflecting central objectives of this initiative at the service system level. First, to what degree is CICH associated with implementation of practices that encourage system integration; with improvements in coordination of service delivery and planning among participating agencies over time; and with increased trust and respect between providers? Second, was the initiative associated with changes in the type of housing provided at CICH sites, with the implementation of homeless information management systems, or with the availability of evidence-based mental health practices? Third, did some sites and some types of agencies show greater change in measures of system-wide performance than others? Lastly, to complement the focus on organizational integration, we examine whether relationships specifically characterized by exchanges of funds are associated with greater levels of inter-agency integration, collaboration and trust, both cross-sectionally and over time. A second report will address client outcomes.
Methods
A "network definition" survey of key informants at core agencies (the lead agency and partnering agencies that provided housing assistance, mental health care, substance abuse services, primary care, and veteran services) was used to identify participating agencies and key informants at these agencies. A more extensive "network participation" survey was then administered in three waves before CICH was implemented (from November 2003 to March 2004) and at the end of the first and second years of operation (from November 2004 to February 200, and from January to March 2006 respectively) . The data collected allowed for the creation of eleven measures, eight that relied on each participating agency at each of the eleven CICH sites as the unit of analysis, and three others that used dyadic relationships between each pair of agencies as the unit of analysis. An average of 6.7 agencies (standard deviation=1.66) were surveyed at each site in each wave along with 44.4 dyadic relationships (standard deviation=19.8). The eleven measures were used to assess five broad dimensions of CICH service systems: 1) system connectedness and integration; 2) emphasis on providing permanent supported housing services ; 3) development of homeless services management information systems; 4) use of evidence-based mental health practices; and 5) the existence of interagency fiscal relationships. Hierarchical linear modeling, general linear modeling and correlation analyses were used to examine change over time and the interrelationships between measures.
Results
The most notable trend was the significant increase over the study period in the implementation of practices that encourage system integration, as well as in levels of system integration themselves, particularly the measure of joint service planning and coordination. Implementation of practices intended to encourage system integration was significantly and positively correlated with multiple indicators of actual levels of integration.
We also observed a significant increase in the availability of information on client and service delivery and in the implementation of homeless management information systems as well as in the use of evidence-based mental health practices. There were no significant changes in ratings of the extent to which various types of housing were provided at CICH sites or in the prevalence of fiscal relationships, primarily due to ceiling effects on these measures. Significant variation between sites or agency types in the amount of change experienced in these system characteristics over the study period was limited.
While the major emphasis in CICH was put on encouraging organizational integration, it was also of interest that agencies with ongoing fiscal relationships had significantly higher levels of joint planning and coordination as well as trust and respect.
Conclusions
This report highlights several positive trends in the characteristics and activities of CICH networks over the course of this initiative. The most notable trend was the significant increase over the two year study period in the implementation of practices that encourage system integration, as well as in levels of system integration, particularly on the measure of joint service planning and coordination. We also found that the implementation of practices intended to encourage system integration was significantly and positively correlated with measured levels of integration. These findings provide evidence of the success of the participating sites in meeting CICH program goals.