Standardized incidence ratio of the COVID-19 pandemic: a case study in a Midwestern state

MotivationThe Coronavirus disease 2019 (COVID-19) has made a dramatic impact around the world, with some communities facing harsher outcomes than others. We sought to understand how counties in the state of South Dakota (SD) fared compared to expected based on a reference population and what factors contributed to negative outcomes from the pandemic in SD. MethodsThe Standardized Incidence Ratios (SIR) of all counties, using age-adjusted and crude adjusted hospitalization and death rates were computed using the SD age-adjusted rate as a reference population. In addition, a penalized generalized linear regression model was used to identify factors that are associated with COVID-19 hospitalization and death rates. This model was then used to compute a new SIR after controlling for other socio-demographic and -economic factors. ResultsWe identified counties that had more or less severe outcomes than what would be expected based on the rate of SD after age adjustment. Additionally, race, education, and testing rate were some of the significant factors associated with the outcome. The SIR values after controlling for these additional factors showed change in magnitude from the range of 4 times more severe to 1.5 times more severe out-come than what is expected. Interestingly the lower end of this interval did not have a major change. ConclusionThe age adjusted SIR model used in this study allowed for the identification of counties with more or less severe than what is expected based on the state rate. These counties tended to be those with high nonwhite percentage, which mostly included counties with American Indian reservations. Although several predictors are associated with hospitalization and deaths, the penalized model confirmed what is already reported in literature that race and education level have a very high association with the outcome variables. As can be expected the further adjusted SIR mostly changed in those counties with higher than expected outcomes. We believe that these results may provide useful information to improve the implementation of mitigation strategies to curb the damage of this or future pandemics by providing a way for data-driven resource allocation.