Assessing gaps between policy and practice in Medicaid disenrollment of jail detainees with severe mental illness.

OBJECTIVE: This prospective cohort study in two large metropolitan jail systems examined whether Medicaid disenrollment policies for persons detained in jail were enforced. The extent to which persons with severe mental illness lost their Medicaid benefits while detained was determined.
METHODS: Mailed questionnaires to state Medicaid directors in 2000 yielded a 95 percent response rate. Directors responded to questions about procedures that are followed when enrollees become inmates in public institutions. In addition, community mental health service records, jail detention records, and Medicaid enrollment records were linked in King County (Seattle) and in Pinellas County (Clearwater and St. Petersburg), Florida, to identify persons with severe mental illness who were incarcerated at any time during a two-year period (1996-1998 in King County and 1998-2000 in Pinellas County). The samples consisted of 1,816 persons representing 4,482 detentions in King County and 1,210 persons representing 2,878 detentions in Pinellas County. Detentions were used as the unit of analysis to determine how often Medicaid disenrollment occurred during jail incarceration.
RESULTS: The stated policy in many states, including Florida and Washington, is to terminate Medicaid benefits upon incarceration, but termination occurred for only 3 percent of the detainees enrolled in Medicaid in each county. In both counties, in 97 percent of the detentions, persons who had Medicaid at entry also had it upon release. In both counties, the 3 percent who lost Medicaid while jailed had longer jail stays (three to five months compared with 16 to 30 days).
CONCLUSIONS: Stated policies do not align with actual Medicaid disenrollment of persons with severe mental illness who become incarcerated. In most instances, short jail stays allowed detainees with severe mental illness to retain their Medicaid benefits.