Achieving Drug and Alcohol Abstinence Among Recently Incarcerated Homeless Women: A Randomized Controlled Trial Comparing Dialectical Behavioral Therapy-Case Management With a Health Promotion Program

Abstract

Background: Homeless female ex-offenders (homeless female offenders) exiting jail and prison are at a critical juncture during reentry and transitioning into the community setting.

Objective: The purpose of the study was to compare the effect of a dialectical behavioral therapy-case management (DBT-CM) program with a health promotion (HP) program on achieving drug and alcohol abstinence among female parolees/probationers residing in the community.

Methods: We conducted a multicenter parallel randomized controlled trial with 130 female parolees/probationers (aged 19-64 years) residing in the community randomly assigned to either DBT-CM (n = 65) or HP (n = 65). The trial was conducted in four community-based partner sites in Los Angeles and Pomona, California, from February 2015 to November 2016. Treatment assignment was carried out using a computer-based urn randomization program. The primary outcome was drug and alcohol use abstinence at 6-month follow up.

Results: Analysis was based on data from 116 participants with complete outcome data. Multivariable logistic regression revealed that the DBT-CM program remained an independent positive predictor of decrease in drug use among the DBT-CM participants at 6 months (p = .01) as compared with the HP program participants. Being non-White (p < .05) and having higher depressive symptom scores (p < .05) were associated with lower odds of drug use abstinence (i.e., increased the odds of drug use) at 6 months.

Discussion: DBT-CM increased drug and alcohol abstinence at 6-month follow-up, compared to an HP program.

Trial registration: ClinicalTrials.gov NCT02258425.

Conflict of interest statement

The authors have no conflicts of interest to report.

Figures

FIGURE 1

Changes in outcomes for Health Promotion vs. DBT-CM groups, showing the primary analysis using complete case data and sensitivity analyses. The p-values for the coefficient of the intervention x time point interaction term estimated using logistic regression models with generalized estimating equation (GEE) are shown.