PREDICTORS OF PROGRAM COMPLETION AND RECIDIVISM IN A JAIL-BASED TREATMENT PROGRAM

Purpose: Substance abuse is a major contributing factor to incarceration and recidivism. Jail substance abuse treatment programs can be effective in decreasing incarceration and recidivism. Understanding the barriers and predictors of program completion, relapse and recidivism in substance abuse treatment programs can facilitate tailored interventions directed at more adversely affected subgroups to improve treatment outcomes. This evaluation study examines the relationship between program completion and socio-demographic factors in predicting recidivism in a jail-based substance abuse treatment program for probationers as they re-integrate into the community. Methods: This study is an outcome evaluation study to determine predictors of program completion and recidivism in a substance abuse treatment program. The program is designed in three phases. Phase 1 is the in-jail component of the intervention, while in phases 2 and 3, the probationer resides in the community, but is required to attend treatment in the supervised(phase 2) and unsupervised(phase 3) phase. We analyzed Data from 197 offenders who participated in the treatment program between June 2009 and December 2012. Variables included age at first arrest, monthly income, gang involvement, prior arrest, employment status, level of education, offense level, mental health status, offence category, and risk level Results: Program completers reported a statistically significant higher income (t=2.95, p=0.02), higher age at first arrest (t=1.94, p=0.03). We found no statistically significant association between program completion and prior arrest (p>0.05) or level of offense (p>0.05). However, recidivism was positively associated with prior arrest (p=0.001), offense level (p=0.048), and offense type (p=0.007). There was no significant association between recidivism and marital status, employment status or education. Logistics regression analysis revealed risk level (OR=0.263) income level (OR=1.001) as significant predictors of program completion while marital status (OR=7.396), prior arrest(OR=0.112) and age at prior arrest (OR=0.906) showed significant association with recidivism. Conclusions: Age at first arrest, Prior arrest and marital status are associated with recidivism while drug risk level and income level are associated with recidivism