Popis: |
Additional file 1: Figure 1. Covariate correlations. Heatmap of Spearman correlations between demographic, racial, socioeconomic, and health covariates. Figure 2. Demographic, racial, and socioeconomic covariate heatmaps. Demographic (yellow, aqua, blue), racial (pink, magenta, purple), and socioeconomic (red, orange) covariate heatmaps. Figure 3. Health covariate heatmaps. Health (white, blue, purple) covariate. Figure 4. Observed and estimated case, death, and case fatality rates. Observed cumulative case fatality rates through 12/21/2020 for all 3,142 US counties. Figure 5. Univariable and multivariable case fatality rate relative risks. Univariable and multivariable relative risks of demographic, socioeconomic, and health comorbidity factors on cumulative COVID-19 case fatality rates through 12/21/20 additionally adjust for state fixed effects and county random effects. Boxes are point estimates and error bars mark 95% confidence intervals. Relative risks are for a one standard deviation increase in a variable (see Additional Table 1), except for the metro/nonmetro categorical variable. Figure 6. Weekly case fatality rates. (A) Line plots of US national weekly case rates and death rates lagged by one week. Solid lines mark similar peaks between weekly case rates and lagged death rates. (B) Heatmaps of county case fatality rates by season. Table 1. List of county-level variables, transformations, and sources. Table 2. Multivariable weekly case fatality rates. Relative risks of county-level variables on weekly case fatality rates (39 repeated measurements per county) by season from 3/23/20-12/21/20 using a one-week and three-week lag for deaths. All results are from a single model that controls for state effects, US census region-specific time varying trends, and additional county overdispersion. Parentheses indicate 95% confidence intervals. Bold indicates confidence interval does not contain 1. Relative risks are for a one standard deviation increase in a variable, except for the metro/nonmetro categorical variable. |