The COVID-19 pandemic continues to impact many countries including the United States (U.S.). In the U.S., much of this effect is due to the SARS-CoV-2 variants and the stagnation of vaccine uptake and coverage (Christie et al., 2021; GISAID, 2021). However, there are factors other than varying distributions of disease and vaccination rates that contribute to unequal impacts. Some of the disparate effects from COVID-19 among the U.S. population can be partially explained by measures of social vulnerability which highlight communities that are likely to be adversely impacted by an external stressor (Barry et al., 2021; Biggs et al., 2021; Garcia et al., 2021; Hughes et al., 2021; Johnson et al., 2021; Kim & Bostwick, 2020; Macias Gil et al., 2020; Oates et al., 2021). Of further concern, recent evidence suggests there is an increased likelihood of experiencing another pandemic from a disease other than COVID-19 in the coming decades (Marani et al., 2021). Therefore, research should continue examining social contributors to the relative risk of COVID-19 impacts, and those results considered when constructing current intervention strategies and build capacity to prepare for future events.

The purpose of this research is to examine the meaningful contribution to modeled relative risk of infection and death from COVID-19 for two of the more highly cited social vulnerability indices, 1.) the Centers for Disease Control and Prevention Social Vulnerability Index (CDC SVI) and 2.) Cutter et al.’s social vulnerability index (SoVI), using their respective composite scores, domains (components), and relevant input variables (Cutter et al., 2003; Flanagan et al., 2011). Social vulnerability indices have been developed to measure composite sociodemographic and environmental impacts and to attempt to quantify their combined effects on a population. These indices, especially the two utilized in this study, are often employed by health agencies in attempts to measure and spatially specify vulnerability. In this study the distribution of COVID-19 cases and deaths are modeled in Indiana at the census tract level using a Bayesian hierarchical spatiotemporal classification technique, which models the trend in relative risk through space and time (G. Li et al., 2014). This framework provides a robust methodology allowing straightforward examination of the meaningful contributions of CDC SVI and SoVI inputs to modeled relative risk of infection and death from COVID-19. There is an increasing need to understand differences between the two selected social vulnerability indices because they are widely used by public health officials for predicting locations of highly affected communities. Additionally, there are but a few examinations on the relationship between social vulnerability and COVID-19 at a statewide fine spatial scale conducted at the census tract-level (a geographic scale essential for examining conditions at the neighborhood or community (intra-county) level) (Biggs et al., 2021; Kim & Bostwick, 2020; Oates et al., 2021). Also, no study to date has examined relationships between CDC SVI, SoVI, and COVID-19 outcomes; existing studies focus on one index or the other. An examination into each index - their input variables, domains, and composite scores – and the respective contributions to the relative risk of SARS-CoV-2 infection and mortality is needed due to the increasing usage of these types of indices to predict effects. Not only will such an investigation aid in our understanding of variations in social vulnerability but will also contribute to the growing literature on COVID-19-specific vulnerability indices and which location-specific variables to include for maximum predictive effect.