Segregated Patterns of Racial and Socioeconomic Inclusivity of Access to Hospital Care Among the Medicare Population

ImportanceMeasures of health care disparities across racial and socioeconomic groups are needed to guide research and policy, particularly for hospitals, because they are the major locus of health care. ObjectiveTo quantify socioeconomic differences between a hospitals patients and the surrounding area, and to assess associations with hospital characteristics. DesignA cross-sectional observational study using Medicare fee-for-service claims and the 2019 American Community Survey (ACS) 5-year zip code data. Quantile regressions identified hospital characteristics associated with lowest and highest inclusivity. SettingInpatient admissions to non-specialty, non-federal hospitals (N = 3,426) in calendar years 2018 and 2019. ParticipantsMedicare beneficiaries enrolled in Parts A & B, and over the age of 65 years Main Outcome and MeasuresInclusivity was defined as the differences between education level, income, and proportion of racial groups, between a hospitals community area and its patient zip codes (obtained from US Census data). The community area was calculated based on a hospitals patient counts from contributing zip codes and by finding this travel time radius from each hospital campus. Percentiles of the inclusivity scores and their composite were compared across hospital characteristics using quantile regressions. ResultsWe included 10,221,387 patients with a mean ({+/-} standard deviation) age of 77.4 (8.0) years and 55.3% were female. The median travel time radius was 48.2 minutes (IQR: 34.7 to 75.0). Weighted median incomes for hospitalized patients zip codes ranged from $27,060 less than their community area to $31,505 more. Black people had the widest weighted percentage differences across hospitals: one hospital had a Black patient proportion 41.6% greater than its community area, while another had a proportion 37.2% less than its community area. Safety net status was the characteristic consistently associated with a higher median inclusivity score (0.4 points greater than non-safety net hospitals, CI: 0.31 to 0.45; p < 0.001). Metropolitan area hospitals had both higher 90th and lower 10th percentile scores compared to non-metropolitan, as did teaching hospitals compared to non-teaching. Conclusions and RelevanceUS hospitals measured inclusivity varies widely, with patterns of greater segregation in metropolitan areas. The results for safety net hospitals reflect their essential role in the US healthcare system. Key pointsO_ST_ABSQuestionC_ST_ABSWhat are the variation patterns of socioeconomic and racial characteristics of hospital inpatient populations in relation to the demographics of their surrounding communities? FindingsUsing Medicare claims data, we calculated the differences in race, income and educational attainment between hospitalized patients and a defined surrounding community area. Quantile regressions showed that the most and the least inclusivity existed in larger, teaching, and metropolitan area hospitals. Safety net hospitals consistently maintained higher inclusivity across both urban and rural settings. MeaningSystematic patterns of racial and class segregation exist among hospitalized Medicare patients.