Kandula S, Shaman J, et al. - Researchers extracted county-level estimates of 14 indicators associated with COVID-19 mortality from public data sources and identified the existence of significant spatial autocorrelation in COVID-19 mortality in the US; 43% of the variability in county-level mortality can be explained by population health/socioeconomic indicators.

• Spatial simultaneous autoregressive (SAR) models that account for spatial autocorrelation in response and predictors were used.

• The largest individual effect estimates were of the prevalence of chronic kidney disease and the proportion of population resident in nursing homes.

• Researchers noted strong spatial autocorrelation in COVID-19 mortality as well as considerable collinearity in COVID-19–linked health conditions.