Using Randomized Controlled Trials to Evaluate Interventions for Releasing Prisoners

Abstract

Randomized controlled trials (RCTs) are rarely used to evaluate social and behavioral interventions designed for releasing prisoners. Objective: We use a pilot RCT of a social support intervention (Support Matters) as a case example to discuss obstacles and strategies for conducting RCT intervention evaluations that span prison and community contexts. Method: First, we summarize existing studies with current and former prisoner samples that highlight RCT methodological techniques. Then, we introduce the trial of Support Matters and discuss obstacles and efforts to overcome those obstacles. Results and Conclusions: We focus on factors related to sampling frame, participant eligibility, sample size, engagement, and attrition. We conclude with general recommendations related to project delays, compromised research sites, and retention. This article provides guidance based on lessons learned from an RCT of a social support intervention with releasing prisoners and their loved ones. Barriers discussed and recommendations presented here will help interventionists develop and implement RCTs in prison to community contexts.