ABSTRACT
BACKGROUND: To promote equitable allocation of scarce COVID-19 resources, most US states added place-based social disadvantage indices in allocation plans. Here we compare how 4 common indices of social disadvantage (differing on social variables including race, and geographic levels)-the Social Vulnerability Index (SVI), Area Deprivation Index (ADI), COVID-19 Community Vulnerability Index (CCVI), and Minority Health-Social Vulnerability Index (MH-SVI)-are associated with COVID-19 incidence/mortality. METHODS: This cross-sectional study uses aggregated COVID-19 cases/ deaths in 3,135 US counties/county-equivalents reported by health departments (New York Times repository), merged with social disadvantage indices from CDC (SVI and MH-SVI), Surgo Ventures® (CCVI), and University of Wisconsin Neighborhood Atlas® (ADI). The SVI, MH-SVI, and CCVI are available at the US county level. ADI was transformed from census block group level to county-level using a population-weighted average. All indices/subindices are national percentile rankings. SVI, MH-SVI, and CCVI range from 0-1 and ADI range from 0-100;higher scores indicating greater disadvantage. For analysis, we converted all indices/subindices into deciles. Mixed effects negative binomial regression models adjusted for population density, urbanicity, and including an offset for county population, were used to estimate associations of each index/subindex with COVID-19 incidence/ mortality, as of July 31, 2021. RESULTS: All 4 disadvantage indices had similar positive associations with COVID-19 incidence (incidence rate ratios [IRR] ranging from 1.03-1.04). Each index was also significantly associated with COVID-19 mortality, but ADI had a stronger association (IRR 1.17, 95%CI 1.16-1.18) than CCVI (IRR 1.07, 95%CI 1.06-1.08), SVI (IRR 1.06, 95%CI 1.05-1.07), and MH-SVI (IRR 1.04, 95%CI 1.03-1.04). Each SVI, MH-SVI, and CCVI subindex was significantly associated with COVID-19 incidence, and most were significantly associated with mortality. CONCLUSIONS: With Omicron and other emerging COVID-19 variants, the need may again arise for allocation of scarce resources like testing, vaccines, and treatments. Despite differences in component measures and weighting, all 4 indices demonstrated an association between greater disadvantage and increased COVID-19 incidence/mortality, suggesting that any index can be used to assist public health leaders in targeting COVID-19 resources to regions most vulnerable to negative COVID-19 outcomes. Of note, unlike SVI, MH-SVI, and CCVI, the ADI does not include race, which can matter for legal/political issues associated with prioritization. Targeting underserved populations with indices that include race as a variable has been challenged by some state policymakers with allegations of reverse discrimination. Policymakers may weigh potential tradeoffs in the political/ practical acceptability when considering use of these indices to target equitable allocation of COVID-19 resources.