Can a correctional group treatment program prevent re-offending among justice-involved people with mental illness?
We are testing the impact of adding a group-based Cognitive Behavioral Intervention (CBI) program to usual services, in reducing justice system involvement for people with mental illness.
Need to reach beyond psychiatric services
Over two million annual jail admissions in the U.S. involve people with serious mental illnesses like schizophrenia, bipolar disorder, and major depression. Traditional responses focus on linking these people with psychiatric care, including medication and therapy under court supervision. However, research shows that symptom reduction alone rarely prevents reoffending.
According to an innovative policy framework, mental health treatment is necessary to address public health goals (e.g., symptom reduction), but often must be paired with correctional interventions to achieve public safety goals (e.g., recidivism reduction). Many justice-involved individuals with mental illness also exhibit risk factors for recidivism—such as criminal attitudes, impulsivity, and antisocial peer associations—that must be addressed to support successful reintegration.
Promise of cognitive behavioral interventions (CBI)
One way to implement this framework is by adding cognitive-behavioral interventions (CBI) to usual psychiatric services. CBI targets risk factors for recidivism while building participants’ skills for self-management, problem-solving, and communication. CBI is structured, applicable in groups, and can be monitored for fidelity.
To advance this approach, the Council of State Governments partnered with the University of Cincinnati Corrections Institute (UCCI) to develop Interventions—a 56-session, seven-month, manualized CBI program tailored for justice-involved people with mental illness. Rooted in UCCI’s evidence-informed programs for general offenders, Interventions incorporates simplified materials and relevant examples adapted to this population’s unique needs. The program’s impact, however, has not been evaluated.
Testing whether and how CBI improves outcomes
The goal of this seven-year project was to rigorously test whether and how Interventions adds value to existing services, to reduce recidivism for this group. To achieve this goal, we conducted two randomized controlled field trials—one in San Francisco and one in Sonoma—recruiting, assessing, and randomizing 722 participants with mental illness to either “services as usual” or “Interventions-added” conditions.
To measure outcomes, we conducted follow-up interviews, coded official records (i.e., probation and services), and accessed state arrest records. Treatment fidelity was assessed by observing and rating CBI facilitators. Fidelity scores were high at both sites, but higher in San Francisco (d= .38). Other key differences between the two sites are summarized below.
Results & implications
Intent to treat analyses indicate that participants randomized to Interventions in San Francisco had significantly lower two-year re-arrest rates compared to those in the usual services group. In Sonoma, no significant differences were observed. Implementation challenges were evident, as many participants assigned to Interventions either did not start or failed to complete the program. Dose-response analyses that rigorously controlled for selection effects indicated that completing each Intervention module significantly reduced the likelihood of re-offending, underscoring the importance of treatment dosage.
These findings demonstrate that Interventions adds value to usual services for justice-involved people with mental illness when delivered effectively. However, the results also highlight the challenges of implementing high-intensity treatment programs for this high-need population. Currently, we are publishing papers and briefs to provide empirical guidance on integrating CBI with usual services to improve outcomes for this group.
Partners and funding
Jennifer Skeem led this project, with Co-I Susan Turner at the Center for Evidence-Based Corrections at UC Irvine. Sharon Farrell directed data collection. This project was made possible through our close collaboration with Citywide Forensics, San Francisco Probation, San Francisco Behavioral Health Court, and Sonoma County’s Day Reporting Center. The University of Cincinnati Corrections Institute, which developed Interventions, provided training and materials. This project was funded by Arnold Ventures.
