Disparities in Coronavirus Disease 2019 Outcomes for African Americans: More Studies Are Warranted.

To the Editor—We read with interest the article by Wiley et al [1], in which they describe coronavirus disease 2019 (COVID-19) outcomes among 94 683 patients presenting to emergency departments in 87 US health systems. Compared with white patients, the authors found no difference in hospitalization risk for African Americans, but higher risk of in-hospital death. These findings contradict several other studies, some with similar or longer study periods, reporting higher risk of hospitalization and similar or lower risk of in-hospital death for African Americans [2-8]. While their use of a large data set might support more generalizable results, the disparities that may exist in smaller individual settings could have been overlooked.

Wiley et al [1] attributed the racial disparities in mortality rate observed in their study to differences in prescribing dexamethasone. Because the magnitude and attributes of COVID-19 racial disparities might be changing during the course of the pandemic [6-8], the authors could assess temporal variations in disparities by stratifying the data into different time periods (eg, before vs after dexamethasone). It would also be useful to assess whether the prevalences of certain underlying conditions and risk factors were higher among black hospitalized patients, which might explain the higher risk of in-hospital death. Moreover, multivariate analyses should be performed separately in each racial group to identify potential differences in independent predictors (eg, receipt of dexamethasone) of hospitalization for COVID-19 or in-hospital death, even in the absence of outcome disparities between racial groups.

Systematic reviews and meta-analyses are useful in addressing discrepancies in study findings. To date, 2 relevant reviews have been published. One of them did not conduct meta-analysis owing to the heterogeneity caused by combining data from ecological studies and public databases [9]. In the other review, not only did the meta-analysis combine UK and US studies, which in itself is cause for concern, but African Americans were also underrepresented (<3%), and the major roles for social determinants were overlooked [10]. Moreover, the outcome measures were inadequately defined, and they were grouped as one without carefully examining the denominator (eg, mortality rate among infected, hospitalized, or intubated patients) or performing subgroup analyses. Precise definition of the outcome is important, because the disparities in one outcome are likely to influence those in the next event. We reviewed the descriptions of the outcomes in the original studies included in these 2 articles and found that some were misrepresented, resulting in only a limited number of studies eligible for assessing the relative risk of each specific COVID-19 outcome for African Americans (Supplementary Table 1). Notably, there has been no meta-analysis regarding racial disparities in the risk of hospitalization for COVID-19.

Thus, more studies, including systematic reviews and meta-analyses, are needed to further understand the trends in and dynamics of COVID-19–related racial disparities and possible reasons for inconsistent findings. These should include studies ranging from smaller community hospitals to larger integrated healthcare databases. Racial disparities in comorbid conditions and other risk factors should be assessed for all well-defined outcomes along the COVID-19 health continuum, to determine whether they are correlated with or predict disparities in subsequent events. This will provide mechanistic insights into possible causes for disparities and aid in tailored public health strategies for eliminating disparities. To this end, future studies addressing COVID-19 health disparities should present data disaggregated by race, perform multivariate analyses stratified by race, and assess temporal and geographic variations and trends in disparities.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Click here for additional data file.