treatment duration. These findings suggest the importance of further investigating relationships among management capacity (financial and human resources) and the ability of SDUs to engage in treatment and service delivery practices known in the clinical research to be positively related to program outcomes.
The average revenue per client varied significantly across CSAT program (grant) type (p=0.0014) and treatment modality (p < 0.0001, see Table 4). The relatively large standard deviations of revenue per client in this table also indicate that there is considerable variation across service delivery units in revenues per client even within CSAT program types and treatment modalities. The Target Cities programs (with the highest average number of clients) had the lowest average revenue per client. Not surprisingly, average revenues per client for the outpatient programs were considerably lower than those for the two residential program types. Gerstein et al. (1997) showed that the typical client in a residential SDU received a greater variety and intensity of services (including ancillary and medical services), likely because clients are in treatment 24 hours per day and require on-site service provision. They also observed that service availability was restricted in Target Cities programs (with the lowest average revenue per client), which suggests there is likely a link between revenues and service provision that is not fully accounted for by the modality of treatment.
Probably because the government provides, on average, more than three-fourths of the funding for substance abuse treatment programs, SDU revenues per client were not significantly correlated with the percent of patients privately insured, publicly insured, or unable to pay. However, there were substantial differences in the sources of revenue among CSAT program types and treatment modalities (see Table 5). Among program types, the Target Cities programs