After adjusting for baseline covariates, and the baseline value for each client service use measure, significant associations at p<.05 were found in 10 of 64 network-client service use relationships examined. The network measure associated with the largest number of measures of client service use was the implementation of best practices. While previous analyses of CICH service use data showed increasing use of preventive primary health care services, discussion of healthcare behaviors with one's primary care provider, and receipt of case management visits in the community (Table 2), use of all three of these types of services were found, paradoxically, to be lower among clients served by networks that reported implementing more evidence-based practices (Table 5, Best Practices data column). Clients served by networks implementing best practices to a greater degree were more likely to receive money management services, but to report being treated by fewer different service providers. Networks more involved in co-planning and integration of services showed greater increases in the proportion of CICH clients who reported having a usual provider for both physical health and mental health problems over time (1.5, p<.05; and, 2.5, p<.01, respectively) (Table 5, Services Co-Planning data column).
Clients served by networks which exchanged resources to a greater extent were also more likely to have a primary treater for mental health over time (0.59, p<.05), but were less likely to receive case management visits in the community over time (-0.60, p<.05) (Table 5, Exchange Resources data column).
Finally, clients served by networks with higher levels of interagency trust were more likely to show increasing proportions of clients with health insurance over time clients (-0.31 uninsured, p<.05)
Significant associations at p<.05 were found on only 2 of the 72 network-client outcome relationships, and among only 1 of the 4 network measures. Clients served by networks implementing more evidence-based practices expressed more trust in their primary healthcare provider (0.24, p<.05), and had lower total treatment costs (-2260, p<.05), than clients treated in networks reporting lower levels of implementation of best practices (Table 5, Best Practices data column).
At the Bonferroni-adjusted p<.001 level of significance the only significant relationship between network measures and either client service use or outcomes was the positive relationship between increased implementation of best practices and increased likelihood of receiving money management services (Table 5).