Patterns of utilization and costs for persons with serious mental illness involved in the criminal justice system

Objective: We examined characteristics of persons with serious mental illnesses involved with the criminal justice system in relation to patterns of utilization and costs of this involvement within respective service sectors.

Methods: We used multiple administrative data sets to identify adults with a SMI arrested in a large urban county in FY 03-04 and followed them for one year before and two years after the 7/1/03-6/30/04 period. Cluster analysis was used with utilization data to identify groups with similar patterns within and across behavioral health and criminal justice systems. Per-capita and aggregate service costs for each group for the four-year period were determined, and multinomial logistic regression was used to determine the relationship between individual characteristics and group membership.

Results: Aggregated costs for the overall sample within criminal justice and health care sectors respectively were approximately equivalent, but individual utilization patterns differed, and psychiatric and substance use diagnoses, homelessness, race, age, and gender were all significantly associated with these differences. Three groups were identified: 1) 11% of the sample incurred 54% of the behavioral and other health care costs and was associated with psychotic or bipolar disorder, substance abuse diagnosis, older, and female gender; 2) 28% of the sample incurred 70% of the criminal justice costs and was associated with homelessness, substance abuse diagnosis, non-white, male, and age 21-39; 3) 61% of the sample incurred 25% of behavioral, other healthcare and criminal justice costs.

Conclusions: Distinguishing subpopulations with substantially different involvement profiles may help in identifying differential strategies for intervention.