Predicting Retention of Drug Court Participants Using Event History Analysis

Abstract

This paper presents the results of a discrete-time event-history analysis of the relationships between client and program characteristics and the length and outcome of participation in a drug court program. We identify factors associated with both successful completion and premature termination. Having an African-American case manager, being older, having little criminal history, and not being a user of crack cocaine are strongly predictive of successfully completing the program. Predicted probabilities of successful completion ranged from 0.16, in the most pessimistic scenario, to 0.88 for the most optimistic scenario.

Key Words:

• Drug courts
• substance abuse treatment
• treatment retention
• case management
• event history analysis
• logistic regression