| ||Literature Coverage Dates||Number of Studies||Number of Study Participants|
|Meta-Analysis 1||1994 - 2013||6||1588|
|Meta-Analysis 2||2000 - 2013||37||21795|
Tripodi and colleagues (2011) conducted a systematic review to examine the effectiveness of substance abuse interventions for incarcerated women. The authors reexamined the studies included in the meta-analysis conducted on interventions for female offenders conducted by Dowden and Andrews (1999) and performed a comprehensive search of the literature to locate studies, including unpublished dissertations, which had been conducted in the United States between 1988 and 2008. To be included, studies had to have evaluated the effectiveness of correctional-based interventions for women inmates in jails or prisons and used experimental, quasi-experimental, or one-group pretest–posttest research designs that provided sufficient information to calculate effect sizes. The researchers excluded studies with attrition rates of more than 30 percent that did not include study dropouts in the analyses.
A total of six studies were identified and fit the criteria for inclusion in the risk-reduction model group of studies. The final sample comprised 1,588 offenders. The treatment group participated in substance-abuse prevention programs that used a therapeutic community approach, with the exception of one program that used a cognitive–behavioral approach.
The outcome of interest was recidivism, which was measured in a variety of ways across the studies, including convictions since release, incarceration since release, return to custody, arrests, criminal activity, drug-related crime, and sex crime. A fixed-effects model and random-effects model were calculated.Meta-Analysis 2
Gobeil and colleagues (2016) conducted a meta-analysis on gender-informed and gender-neutral interventions in which they examined the effectiveness of correctional interventions for incarcerated women on reducing recidivism. The authors performed a comprehensive search of published and unpublished literature between January 2000 and December 2013. To be included, studies had to be focused on correctional programs or interventions for women offenders, have a minimum of one comparison group, and a measure of recidivism after intervention completion. The researchers excluded studies reporting on populations of immigration detention and secure mental health settings. Studies had to have at least 10 participants per group and use research designs that provided sufficient information to calculate effect sizes.
A total of 37 studies providing 38 unique effect sizes were identified and fit the criteria for inclusion in the meta-analysis. The final sample comprised 8,831 incarcerated women in the treatment group and 12,964 in the control group. The treatment group participated in cognitive–behavioral and therapeutic community programs that focused on substance use, multi-targets, trauma, cognitive skills, education, life skills, peer support, case management, and parenting.
The outcome of interest was recidivism, which was measured in a variety of ways across the studies, including conviction, charge/arrest, any return to custody, and not defined. Both fixed-effects and random-effects models were calculated.