Excess deaths associated with COVID‐19 by rurality and demographic factors in the United States

Abstract

AbstractPurposeTo estimate percent excess deaths during the COVID‐19 pandemic by rural‐urban residence in the United States and to describe rural‐urban disparities by age, sex, and race/ethnicity.MethodsUsing US mortality data, we used overdispersed Poisson regression models to estimate monthly expected death counts by rurality of residence, age group, sex, and race/ethnicity, and compared expected death counts with observed deaths. We then summarized excess deaths over 6 6‐month time periods.FindingsThere were 16.9% (95% confidence interval [CI]: 16.8, 17.0) more deaths than expected between March 2020 and February 2023. The percent excess varied by rurality (large central metro: 18.2% [18.1, 18.4], large fringe metro: 15.6% [15.5, 15.8], medium metro: 18.1% [18.0, 18.3], small metro: 15.5% [15.3, 15.7], micropolitan rural: 16.3% [16.1, 16.5], and noncore rural: 15.8% [15.6, 16.1]). The percent excess deaths were 20.2% (20.1, 20.3) for males and 13.6% (13.5, 13.7) for females, and highest for Hispanic persons (49% [49.0, 49.6]), followed by non‐Hispanic Black persons (28% [27.5, 27.9]) and non‐Hispanic White persons (12% [11.6, 11.8]). The 6‐month time periods with the highest percent excess deaths for large central metro areas were March 2020‐August 2020 and September 2020‐February 2021; for all other areas, these time periods were September 2020‐February 2021 and September 2021‐February 2022.ConclusionPercent excess deaths varied by rurality, age group, sex, race/ethnicity, and time period. Monitoring excess deaths by rurality may be useful in assessing the impact of the pandemic over time, as rural‐urban patterns appear to differ.