LAUSR

Study Objectives: To study the impact of COVID-19 hospitalizations on patients from differing ethnic and racial groups. Study Design and Methods: This is an IRB-exempt observational cohort study of de-identified patient data obtained from our central billing system comprising greater than 8.6 million emergency visits annually in over 175 United States hospitals. Inclusion criteria were adults aged 18 years and older, a presentation to one of our hospital emergency departments, and an admission for COVID-19 infection. Outcome variables were length of stay, in-hospital mortality, disease severity, and discharge disposition. Discharge disposition was further categorized into home, skilled nursing facility, hospice. Outcomes were stratified by racial groups: White, Black, Hispanic and Other. Statistical analysis consisted of summary statistics (distributions) with medians and interquartile ranges reported for non-normally distributed variables. Linear regression analyses determined factors predictive of outcome. A p-value of <0.05 was considered statistically significant. All statistical analyses were performed using JMP Pro 14.1 for Windows (R). Results: The dataset consisted of 45,360 patients. The cohort was 22% Black, 35% Hispanic, 37% White, and 6% Other. The mortality was 15% for all groups. White patients had the highest mortality rate at 17% compared to 10% in Blacks, 14% in Hispanics, and 15% in Other (ANOVA, p<0.0001). Whites were significantly older (Wilcoxon rank sum, <0.0001) with a median age of 71 years (IQR 59-80), compared to Blacks with a median age of 60 (IQR 46-71), Hispanics with a median age of 57 (IQR 44-70), and other races with a median age of 61 (IQR 48-73). Race was statistically significant in a multivariable model including age, sex, and race, with women having an odds ratio of 1.35 for survival. 6484 patients required ICU admission and intubation with hemodynamic support. This burden was disproportionate across racial groups, with 15.6% of Blacks and 13.9% of non-Blacks having such critical disease (P<0.0001, z-test for proportions). The overall median hospital length of stay (HLOS) for all races was 5 days (IQR 3-11). The median HLOS for all non-Whites was 5 days, whereas for Whites it was 6 days (p<0.0001). Whites were significantly less likely to be discharged home (P<0.001). A significantly higher proportion of Blacks and Hispanics were on Medicaid compared to Whites (p<0.0001). Conclusion: White patients had a higher mortality rate than non-White patients hospitalized for COVID-19;however, Black and other non-White patients were hospitalized for COVID-19 at a younger age than White patients. Black patients were significantly more likely to require admission to the ICU. These data suggest there is a multifactorial etiology behind the varying impact of COVID-19 on patients. Further examination of other social determinants of health are warranted to fully understand COVID-19 health disparities.