Covid-19 Mortality in US Counties

This work follows a healthcare-systems question across multiple waves of COVID-19 in the United States that "as clinical care, variants, vaccination, and hospital strain changed, did the factors linked to mortality-based severity change as well?" Using county-level data from August 2020 to March 2023, we apply rolling multivariable regression to repeatedly re-estimate associations between two severity measures (case fatality rate, and a second measure relating deaths to hospitalizations) and a set of operational and population factors such as hospital bed availability, vaccination coverage, age structure, rural–urban context, and broader community risk indicators. Rather than assuming one stable "answer" for the whole period, the analysis is designed to show which relationships persist over time and which ones shift direction or strength, producing an interpretable timeline of how system capacity and population context relate to mortality outcomes as conditions evolve.