PHS80 - Contrasting Utilization of Inpatient Hospitalizations and Emergency Department Visits Between Medicaid and Commercially Insured Arkansans

Abstract

Arkansas expanded Medicaid coverage for individuals with incomes ≤138% federal poverty level to provide commercial insurance (qualified health plans, QHPs) through premium assistance to non-frail individuals. We compared utilization of inpatient hospitalizations (IH) and emergency department (ED) visits between traditional Medicaid and QHP enrollees. Non-institutionalized, ambulatory, newly enrolled Arkansans aged 18-64 years between 2014-16, with at least 180 days of continuous enrollment in any two consecutive years in either Medicaid or QHP were studied. Two approaches were used: a) propensity score matching (PS) used demographics, Charlson comorbidity index, insurance region, and first-year’s census block median income; b) regression discontinuity (RD) based on a continuous health needs score assessed prior to the plan enrollment. Using PS, the rate (per 10,000 person-years) of IH in follow-up year-1 was higher (mean[SE]: 757.7[23.0] vs. 643.0[22.2], p<0.001), average LOS (days) was lower (year-1:4.7[0.2] vs. 6.2[0.2], p<0.001; year-2:4.4[0.2] vs. 6.2[0.4], p<0.001) in Medicaid vs. QHPs. Similarly, rates (per 100 person-years) of total ED visits (year-1:82.4[1.2] vs. 73.5[1.0], p<0.001; year-2:90.3[1.9] vs. 87.7[1.7], p=0.041), rates of emergent-ED visits (year-1:14.5[0.4] vs. 12.7[0.3], p<0.001; year-2:15.8[0.6] vs. 14.6[0.4], p=0.002), and rate of non-emergent-ED visits were higher (41.6[0.7] vs. 37.1[0.6], p<0.001) in Medicaid. Using RD, the rates of IH (year-1:1663.3[55.7] vs. 528.0[12.8], p<0.001; year-2:1478.0[58.4] vs. 557.3[14.8], p<0.001) and rate of non-emergent-ED visit in year-2 (60.4[1.5] vs. 27.8[0.4], p=0.042) were higher in Medicaid vs. QHPs, but differences in the average LOS (p=0.997), rates of total ED visits (p=0.085) and emergent-ED visits (p=0.415) were not statistically significant. Although enrollment in QHPs was associated with lower rates of IH, these rates increased in QHPs over time and were not materially different from Medicaid in the final year of follow-up. The rates of ED visits, and particularly non-emergent-ED visits, were higher for Medicaid, suggesting better access to physician-based office care in QHPs.