Long-Term Outcomes of a RCT Intervention Study for Women Convicted of Violent Offenses

Background and Purpose: Violence prevention interventions have focused mainly on men, and as a result, there is little effective intervention to address the often coexisting trauma-related anger, violent behaviors and substance misuse among women. Beyond Violence (BV), a 20 session program for women who engage in, or are convicted of, violent behaviors, has shown efficacy among incarcerated women in terms of feasibility and short-term outcomes through a multi-phase study that included a randomized control trial (RCT). Specifically, BV was shown to significantly decrease symptoms of depression, anxiety, and PTSD and influence anger expression when compared to a treatment-as-usual (TAU) – a 44 session intervention designed primarily for incarcerated men. Building upon these prior studies, we assess long term outcomes with an overarching research question: Is BV more effective than the TAU in reducing recidivism and relapse after prison release?

Methods: The study utilized 37 women involved in the RCT during their incarceration and on parole for one year. Among them, two women were out of state and one was deceased. As a result 34 women were included in the analysis (TAU n=16; BV n=18). Initial inclusion criteria: violent offense; within 2 years of release; no serious mental health disorder; and presence of a substance use disorder (SUD). Data consisted of extracted parole officer case notes and department records from a state-level administrative database, for the study period (12 months post prison release).  Recidivism was defined in two ways: a) return to prison for a parole violation, and b) new arrest/jailed.  Relapse was defined as a positive drug screen, proportion of positive to number of screens, and time to first positive.

Results: There were no differences between groups in age at treatment (M=34, SD=9.0), race (47% white, 53% African American), or the proportion of positive drug tests (1.6 over 12 months). Although none of the women returned to prison, 11 women (32%) were arrested or jailed. Results of logistic regression showed that women who received BV were less likely to recidivate to arrest or jail than those who received TAU (OR =.20 for BV compared to TAU, p=.05, CI=.04, .97). Cox regression on recidivism showed that the hazard for recidivism was lower for BV compared to TAU (hazard ratio = .28; LR chi square = 4.12, p< .05). Although women in BV were less likely to relapse (25% v 47%) there were no significant differences between groups on the proportion or timing of relapse.

Conclusion/Implications: This study found that compared to TAU, BV had a significantly positive impact on recidivism among incarcerated women. The lack of differences on relapse may be a result of sample size, but is concerning for long term prevention outcomes given the relationship between use and violence. Nonetheless, results indicate that the gender-specific and trauma informed BV intervention can provide better results for women than the higher dosage intervention designed for men (20 vs 44 sessions). Implications for research and practice include insights about importance of developing gender specific prevention program for those involved in criminal/legal systems.