Duration and Continuity of Medicaid Enrollment Before the COVID-19 Pandemic.

Importance: COVID-19 relief legislation created a temporary moratorium on Medicaid disenrollment, but when the public health emergency ends, states will begin to "unwind" Medicaid enrollment. Prepandemic data shed light on factors that can affect Medicaid coverage stability. Objective: To assess factors associated with the duration and continuity of Medicaid enrollment. Design, Setting, and Participants: In this cross-sectional analyses of a Medicaid data set for 2016 that was released by the Agency for Healthcare Research and Quality in June of 2022, we analyze a nationally representative data set of 5.7 million persons, weighted to represent 70 million Medicaid beneficiaries in 2016. We focus on 22 million nondisabled, nonelderly adults for this analysis. The data were analyzed between July and September of 2022. Main Outcomes and Measures: The main outcomes were the average months of Medicaid enrollment in 2016 and the probability of churning, defined as a break in coverage between 2 periods of enrollment during the calendar year. We compared these outcomes by eligibility category, state, demographic characteristics, and key Medicaid policies, including whether the state expanded Medicaid and whether it used ex parte reviews (automated reviews of other administrative data to reduce renewal paperwork burdens). Results: In this cross-sectional analysis, we analyze a nationally representative Medicaid data set of 5.7 million persons, weighted to represent 70 million Medicaid beneficiaries in 2016, released by the Agency for Healthcare Research and Quality in June of 2022. The analysis focused on nonelderly, nondisabled adults (aged 18-64 years) with a weighted population size of 22.7 million, of which 18.4% were Black, 19.2% were Latino, 39.5% were White, 7.3% were other/Asian/Native American, and 15.5% had unknown race. Multivariable regression analysis indicated that those living in states that expanded Medicaid but did not use ex parte reviews had longer average duration (0.31 months longer; 95% CI, 0.03-0.59) and lower risk of churning(odds ratio [OR], .40; 95% CI, 0.39-0.40), whereas those living in nonexpansion states that used ex parte reviews had lower odds of churning (OR, .68; 95% CI, 0.66-0.70) but also had shorter average duration (3.1 months shorter; 95% CI, -3.4 to -2.8). Those living in expansion states that used ex parte reviews also had reduced churning (OR, .83; 95% CI, 0.82-0.85). The average duration varied widely by state, even after adjustments for demographic and state policy factors. Conclusions and Relevance: If state Medicaid programs revert to prepandemic policies after the temporary moratorium ends, Medicaid coverage, particularly for nondisabled, nonelderly adults, is likely to become less stable again. Medicaid expansions are associated with improved continuity, but ex parte review may have a more complex role.