Spatial Analysis of Racial & Economic Disparities in COVID-19 Infection in Louisiana

Background. After the SARS-CoV-2 virus entered the United States it became rapidly apparent that there were stark racial disparities in COVID-19 infection. Minorities, especially Black or African Americans, bore the greatest burden of disease and death. Researchers quickly began examining these disparities and social determinants of COVID-19. Early cohort studies reported that minorities and individuals of low socioeconomic status were more likely to test positive. Studies assessing the role of area-based social measures (ABSMs) found a range of factors associated with infection, including racial composition, poverty, and population density, as well as environmental factors like housing, transportation and air pollution. Often, ABSM studies assess risk using small-area average measures and there has been little focus on how infection rate relates to relative concentrations of risk within an area, i.e. segregation of the most privileged from the most disadvantaged. We aim to examine this with in relation to the two factors that have shown greater likelihood of positive testing, race and income, in the state of Louisiana, a state known for high racial and economic inequality.

Method. We linked administrative data on COVID-19 from the Louisiana Department of Health from March to December 2020 and population data from the United States Census Bureau’s American Community Survey (ACS) 5-year estimates at the census tract level using ArcMap. Residential racial and economic segregation was measured using the Index of Concentration of the Extremes (ICE). We first conducted ordinary least squares (OLS) regression to assess a global association between ICE and COVID-19 case rate. Spatial analysis via geographic weighted regression (GWR) was used to model spatially varying relationships.

Results. OLS models indicated a significant global association between COVID-19 case rates and racial ICE (p=0.015), where areas with greater concentrations of Black residents had increased case rates. However, there was significant spatial variation in the relationship between COVID-19 cases and all three ICE indicators (racial, economic and racial-economic). A GWR model with racial ICE exhibited the best model fit and explained approximately 24% of the variation in COVID-19 case rates (adjusted R²=0.24).

Implications. We have found that racial and economic disparities in COVID-19 cases vary geographically, which supports the notion that social determinants, not biological, drive racial disparities. Future research should investigate the role of public health policy, testing resources or other environmental factors that may explain why some areas exhibited greater disparities than others. In general, policy makers and public health officials would do well to target public health and vaccination campaigns not only to members of racial and ethnic minorities or those with lower income, but also to people who live in areas of extreme inequality.