Selection into mental health court

Abstract

How defendants are selected into mental health courts (MHC) is central to issues of fairness, efficacy, and successful program replication. Only recently has empirical research started to examine MHC selection, revealing a multi-stage process with multiple decision makers and multiple variables. In this study, we use classification and regression tree analysis (CART) to examine the variables suggested in recent research to predict selection into MHC. The analysis includes legal and diagnostic variables, treatment history, measures of treatability, motivation to change, violence risk, and symptom severity. We find that the MHC is more likely to accept defendants who did not have warrants issued for their arrest, who had diagnoses other than depression, and who did not report using illegal drugs around the time of their admission. Symptom severity and motivation to treatment also predict MHC admission, with their effects contingent on defendants' statuses on other variables.