Interruptions in Medicaid Coverage and Risk for Hospitalization for Ambulatory Care-Sensitive Conditions

Background: Many low-income U.S. citizens experience interruptions in health insurance coverage. Objective: To determine the rate of hospitalization for ambulatory care-sensitive conditions among Medicaid beneficiaries with interruptions in coverage. Design: Retrospective cohort study. Setting: California Medicaid population. Patients: 4 735 797 adults in California age 18 to 64 years who received a minimum of 1 month of Medicaid coverage between 1998 to 2002. Measurement: Time to hospitalization for an ambulatory care sensitive condition. Results: Sixty-two percent of Medicaid beneficiaries experienced at least 1 interruption in coverage during the study period. The 3 most common ambulatory care-sensitive conditions resulting in a hospitalization were heart failure, diabetes, and chronic obstructive pulmonary disease. Interruptions in coverage were associated with a higher risk for hospitalization for an ambulatory care-sensitive condition (adjusted hazard ratio, 3.66 [95% CI, 3.59 to 3.72]; P < 0.001). In subgroup analyses, the association between interrupted coverage and hospitalization was stronger for beneficiaries eligible through the Temporary Aid to Needy Families program (adjusted hazard ratio, 8.56 [CI, 8.06 to 9.08]) than for beneficiaries eligible through the Supplemental Security Income program (adjusted hazard ratio, 1.72 [CI, 1.67 to 1.76]), who typically retain Medicare coverage even when their Medicaid coverage is interrupted. Limitation: The study lacked information on why interruptions occurred and whether beneficiaries with interruptions transitioned to other insurance coverage. Conclusion: Interruptions in Medicaid coverage are associated with a higher rate of hospitalization for ambulatory care-sensitive conditions. Policies that reduce the frequency of interruptions in Medicaid coverage might prevent some of the health events that trigger hospitalization and high-cost health care spending.