Expedited Medicaid Enrollment and Substance Use Service Utilization among Released Prisoners with Severe Mental Illness

People released from criminal incarceration are at especially high risk of death from drug-related causes, pointing to the importance of connecting individuals to treatment after release from prison. An important risk factor for substance use disorders (SUD) among incarcerated individuals is serious mental illness (SMI). Though Medicaid plays an important role in financing behavioral health treatments for vulnerable groups, many states terminate individuals’ Medicaid coverage during incarceration. In January 2006, Washington State began a program of expedited Medicaid enrollment for individuals with mental illness being discharged from state prisons, jails, and psychiatric hospitals. The current paper examines the effect of referral to expedited Medicaid on use of SUD treatment for prison releasees with SMI. Our sample consisted of 3,086 individuals with a diagnosis of SMI who were released from prison from January 1, 2006 to December 31, 2007. Of the sample we identified, 869 individuals received referrals for expedited Medicaid and 2,217 did not. To control for potential endogeneity in the selection of prisoners for referral, we used inverse probability of treatment weights (IPTW) to balance the referred and not-referred groups on more than 50 baseline covariates that may predict service utilization. We used doubly robust logistic regression models using IPTW to estimate the effect of referral to expedited Medicaid on use of any SUD treatment in the 12 months following release. In the unweighted data, 33% of individuals in the referred group and 27% of individuals in the non-referred group used any SUD treatment services following release. When controlling for baseline difference using IPTWs and covariates in a doubly robust logit model, referral to expedited Medicaid enrollment was associated on average with a 4.8 (SE 1.9, p<.01) percentage point increase in the predicted probability of using any SUD treatment in the 12 months following release. Referral to expedited Medicaid led to a moderate increase in the predicted probability of using any SUD treatment in the 12 months after index release. Our results support the use of expedited Medicaid enrollment programs for people with SMI released from prison, though caution should be taken concerning the generalizability of findings from this state-specific study.