Emergency Department Utilization by Race, Ethnicity, Language, and Medicaid Status.

introductionEmergency department (ED) use varies by age, sex, race, ethnicity, language preference, and payor type. Most studies comparing ED use by patients with English vs non-English preference (ELP/NELP) have used racially aggregated data, potentially masking differences across population subgroups. In this study we aimed to disaggregate the associations of race, ethnicity, language preference, and Medicaid coverage with ED utilization.

methodsWe used cross-sectional study electronic health record data for 2,047,105 Kaiser Permanente Northern California members who were 25 - 85 years of age in January 2019 and had been continuous health plan members during 2018 - 2019. We tabulated the percentages of adults in seven racial and ethnic groups (White, Black, Hispanic, Chinese, Filipino, Vietnamese, South Asian) within three age groups (25 - 44, 45 - 64, 65 - 85) who had ≥1 ED visit in 2019. Modified log-Poisson regression was used to examine racial, ethnic, and language preference differences after adjusting for demographic and Medicaid status covariates.

resultsThe study population was 51.8% White, 53.2% female, 9.6% NELP, and 6.2% Medicaid-insured. Overall, 18% had ≥ 1 ED visit. Compared with White adults, Black and Hispanic adults were more likely and Chinese, Vietnamese, and South Asian adults were less likely to have ≥ 1 ED visit. After adjusting for all covariates, NELP adults 25 - 64 years of age were 10% less likely to have had an ED visit. However, while NELP was associated with a 10-20% lower ED visit prevalence among Hispanic, Filipino, Chinese, and Vietnamese adults 25 - 64, the prevalence was 10% higher among White and South Asian adults 45 - 64 and Filipino and South Asian adults aged 65 - 85. Adults with Medicaid coverage aged 25 - 64 were twice as likely and adults aged 65 - 85 were 50% more likely to have had ≥ 1 ED visit.

conclusionThis study of a US adult health-plan membership found several significant differences in ED use across racial, ethnic, and language subgroups and a higher prevalence of ED use by Medicaid-covered adults ≤ 65 years of age in most racial and ethnic groups. Our findings highlight the importance of using disaggregated data, particularly for Asian ethnic groups, when comparing ED use in different populations. Further research is needed to identify similarities and differences in social, personal, and policy factors driving ED use in diverse adult populations to better inform population-specific health interventions.