Keywords: hospital closures, affordable care act, nonprofit, Medicaid, multilevel modeling
Since 2010, there has been a wave of hospital closures and mergers across the United States. Given that decisions to expand Medicaid were state-specific, there may be differences in number and type of hospital closures between states that expanded and did not expand Medicaid. The present study aimed to investigate the association between state level Medicaid expansion and short-term hospital closures from 2010-2016 using multilevel modeling of hospitals, counties, and states. Hospital data was obtained from the Center for Medicaid and Medicare Services. For contextual county effects, small area health insurance, income, and poverty estimates were included. Smaller hospitals, nonprofit hospitals, and hospitals with a history of ownership change were more likely to close. Critical access hospitals were less likely to close. Medicaid non-expansion was not associated with a greater risk of hospital closure once included in the multilevel model. Further, rural vs. urban status was not predictive of hospital closure, though population density was a significant predictive factor. Policies supportive of small and nonprofit hospitals may be beneficial in preventing more hospital closures.