Assessment of COVID-19 Hospitalizations by Race/Ethnicity in 12 States

Given the reported health disparities in coronavirus disease 2019 (COVID-19) infection and mortality by race/ethnicity,^1,2 there is an immediate need for increased assessment of the prevalence of COVID-19 across racial/ethnic subgroups of the population in the US. We examined the racial/ethnic prevalence of cumulative COVID-19 hospitalizations in the 12 states that report such data and compared how this prevalence differs from the racial/ethnic composition of each state’s population.

Using data extracted from the University of Minnesota COVID-19 Hospitalization Tracking Project,³ we identified the 12 states that reported the race/ethnicity of individuals hospitalized with COVID-19 between April 30 and June 24, 2020. We calculated the percentage of cumulative hospitalizations by racial/ethnic categories averaged over the study period and then calculated the difference between the percentage of cumulative hospitalizations for each subgroup and the corresponding percentage of the state’s population for each racial/ethnic subgroup as reported in the US Census.⁴ The race/ethnicity categories included were White, Black, American Indian and/or Alaskan Native, Asian, and Hispanic. Descriptive statistical analyses were conducted using Stata/MP, version 14 (Stata Corp). The University of Minnesota Institutional Review Board reviewed the study data and deemed it exempt from review and informed consent requirements because the study was not human subjects research.

This analysis of COVID-19 hospitalizations in 12 US states during nearly a 2-month period represented a total of 48 788 cumulative hospitalizations among a total population of 66 796 666 individuals in 12 US states by the end of the study period on June 24, 2020. The share of the hospitalizations of White patients was substantially smaller vs their share of state population in all 12 states (Table and Figure). For example, in Minnesota, the share of hospitalizations of White patients was 52.9%, whereas their share of the state population was 84.1%. Conversely, the percentage of hospitalizations among Black patients exceeded the percentage of their representative proportion of the state population in all 12 states. Differences between the cumulative percentage of hospitalizations and the state population of Black individuals were greatest in Ohio (31.8% vs 13.0%), Minnesota (24.9% vs 6.8%), Indiana (28.1% vs 9.8%), and Kansas (22.0% vs 6.1%).

Eleven states reported the number of COVID-19 hospitalizations for Hispanic individuals, and in 10 states, the percentage of hospitalizations for Hispanic individuals was higher than their representative proportion of the state population. The disparity among Hispanic individuals was most pronounced in Virginia (36.2% of hospitalizations vs 9.6% of population), Utah (35.3% of hospitalizations vs 14.2% of population), and Rhode Island (33.0% of hospitalizations vs 15.9% of population).

The pattern was largely reversed for the Asian population. In 6 of 10 states that reported data for this subgroup, the proportion of hospitalizations was lower compared with their population representation. In Massachusetts, for example, the Asian population comprised 7.0% of the population but only 4.0% of the COVID-19 hospitalizations.

Hospitalization data for American Indian and Alaskan Native populations were only reported by 8 states. However, the disparity was substantial in select states. In Arizona, the American Indian and Alaskan Native population accounted for 15.7% of the hospitalizations but only 4.0% of the state’s population. Similarly, in Utah, this subgroup accounted for 5.0% of the hospitalizations in contrast with 0.9% representation of the state population.

This analysis identified considerable disparities in the prevalence of COVID-19 across racial/ethnic subgroups of the population in 12 US states. These findings are consistent with an earlier Centers for Disease Control and Prevention analysis of 580 hospitalizations between March 1 and March 30, 2020, that found disproportionately high COVID-19 hospitalizations for the Black population.⁵ Similarly, a study of 1052 confirmed COVID-19 cases between January 1 and April 8, 2020, at a California health system reported higher odds of hospitalization in non-Hispanic Black individuals compared with non-Hispanic White individuals.⁶ In addition, we observed high hospitalization rates for Hispanic individuals in most of the states analyzed and high hospitalization rates for American Indian and Alaskan Native populations in select states.

These findings highlight the need for increased data reporting and consistency within and across all states. Only 12 of 50 US states have consistently reported hospitalizations by race/ethnicity during our study period. New Jersey and Florida recently started reporting data on COVID-19 hospitalizations by race/ethnicity. The present study is limited in that there was no adjustment for age, sex, comorbidities, and socioeconomic factors within each racial/ethnic group that are likely to be associated with COVID-19 hospitalizations.

A large body of research has identified racial/ethnic health disparities in the risk of infection associated with a higher prevalence of comorbidities, less access to health care, adverse economic conditions, and service-related occupations.² The unique clinical, financial, and social implications of COVID-19 for racial/ethnic populations that are often systematically marginalized in society must be well understood to design and establish effective and equitable infrastructure solutions.

Corresponding Author: Pinar Karaca-Mandic, PhD, Carlson School of Management, Department of Finance, University of Minnesota, 321 19th Ave S, Minneapolis, MN 55455 (pkmandic@umn.edu).

Published Online: August 17, 2020. doi:10.1001/jamainternmed.2020.3857

Author Contributions: Dr Karaca-Mandic had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Karaca-Mandic, Georgiou.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Karaca-Mandic, Georgiou.

Obtained funding: Karaca-Mandic.

Administrative, technical, or material support: All authors.

Supervision: Karaca-Mandic.

Conflict of Interest Disclosures: Dr Karaca-Mandic reported receiving funding from the University of Minnesota, Office of Academic Clinical Affairs and grants from United Health Foundation during the conduct of the study; personal fees from Tactile Medical, Precision Health Economics, and Sempre Health; and grants from the Agency for Healthcare Research and Quality, American Cancer Society, National Institute for Health Care Management, National Institute on Drug Abuse, and National Institutes of Health outside the submitted work. Dr Georgiou reported receiving personal fees from HealthGrades outside the submitted work. No other disclosures were reported.

Additional Information: This research uses publicly available data from the University of Minnesota COVID-19 Hospitalization Project, which is partially funded by the University of Minnesota Office of Academic Clinical Affairs and the United Health Foundation. Although the race and ethnicity breakdown of hospitalizations are not reported on the project website, data can be requested from the project team.

Additional Contributions: Yi Zhu, MA, Carlson School of Management, University of Minnesota, contributed to this project by helping launch the University of Minnesota COVID-19 Hospitalization Project website. Mr Zhu did not receive compensation. Zachary Levin, BA, School of Public Health, University of Minnesota, provided research assistance with data analysis.Khoa Vu, Applied Economics, University of Minnesota, provided research assistance with data collection. Messrs Levin and Vu were compensated as research assistants.