Am J Epidemiol. 2021 Mar 12:kwab058. doi: 10.1093/aje/kwab058. Online ahead of print.
In their commentary, Zalla et al. argue that the approach taken by Centers for Disease Control (CDC) comparing the proportion of COVID-19 deaths by race/ethnicity to a weighted population distribution ignores how systemic racism structures the composition of places. While the CDC has abandoned their measure, they do so because of the changing geographic distribution of COVID-19, not because the measure underestimates racial disparities. We further Zalla et al.’s argument, advocating for a relational approach to estimating COVID-19 racial inequities that integrates the reciprocal relationship between context and composition through the interaction of places and people over time. To support our argument, we present a series of figures exploring the heterogeneous relationships between places, people, and time, using US county-level publicly available COVID-19 mortality data from February to December 2020 from Johns Hopkins University. Longitudinal and more geographically granular data that allows for disaggregation by person, place, and time will improve our estimation and understanding of inequities in COVID-19.