Community Reentry of Persons with Severe Mental Illness Released from State Prison

Joseph P Morrissey, PhD, Co-Principal Investigator
Gary Cuddeback, PhD, MSW, MPH, Co-Investigator
Marisa Domino, PhD, Co-Investigator

Linda Frisman, PhD, Co-Principal Investigator; Director of Research, Connecticut Department of Mental Health & Addiction Services
David Mancuso, PhD, Responsible Investigator, Senior Research Supervisor, Washington State Department of Social and Health Services, Research and Data Analysis Division

David Lovell, PhD, Dept of Psychosocial & Community Health, School of Nursing, University of Washington
Thomas McGuire, PhD, Dept of Health Care Policy, Harvard Medical School
Jeffrey Swanson, PhD, Dept of Psychiatry and Behavioral Sciences, Duke University
Marvin Swartz, PhD, Dept of Psychiatry and Behavioral Sciences, Duke University

Mentally ill persons involved in the criminal justice system represent a new frontier for community mental health and disability research. The scope of this problem is truly staggering with upwards of 86,000 persons with severe mental illness (SMI) released from state prisons each year with high rates of recidivism. The loss of Medicaid benefits by incarcerated offenders is thought to be a major obstacle to successful community reentry. Several states have adopted expedited restoration programs to reconnect eligible prisoners to their Medicaid benefits prior to release. Early benefit restoration can avoid disruptions in medication regimens and treatments for offenders with SMI when they are released to the community. However, few rigorous research studies have been conducted to establish the impact or cost-effectiveness of these policies.

This research project responded to RFA-MH-09-050 from the NIMH with a three-year study that addressed these concerns by undertaking an assessment of expedited Medicaid benefits restoration policies in Connecticut and Washington State for released prisoners who have SMI. Two specific aims were addressed:  1.The impact of EMBR on three outcomes—subsequent hospitalizations, use of outpatient mental health and substance abuse services, and criminal recidivism,. Two complementary designs were employed: (a) a difference-in-differences design to compare individuals with SMI and those without SMI centered on a policy adoption date of April 1, 2005 in Connecticut and January 1, 2006 in Washington with three years pre and two years post data in each state and (b) a quasi-experimental study design focused on the post-period only comparing individuals with SMI who received expedited Medicaid restoration (treatment group) vs. individuals with SMI who did not (comparison group) over a two-year follow-up period using a hierarchical linear modeling framework. Propensity score matching was used with both designs to achieve balance between groups on all observed baseline covariates.  2. The cost of EMBR in maintaining prison releasees with severe mental illness in the community from both individual state agency (Medicaid, mental health, corrections) and total government perspectives. These analyses used community days as the composite measure of clinical effectiveness. Regression analyses estimated the marginal effect of EMBR on costs and community days for the restored population as compared with the comparison group, controlling for differences in target populations between states and for trends in costs and community days over time. Incremental cost-effectiveness ratios were formed from the model coefficients.

This study was well-aligned with RFA-MH-09-050 in that it addressed an important mental health policy issue for state authorities, took advantage of a wealth of administrative data from two states, applied an array of analytical strategies from the fields of economics and statistics to address an important state mental health policy issue, and responded to the call for research about justice-involved persons with severe mental illness as put forth in the President’s New Freedom Commission on Mental Health (2002).