COVID-19 Prevalence Estimates for the State of Ohio
- Problem
- Early in the pandemic, case counts could not show how many Ohio adults had current or past infection.
- Methods & tools
- Bayesian multilevel regression and poststratification, with adjustment for imperfect diagnostic tests, low positivity, and nonresponse.
- Outputs
- Peer-reviewed estimates in PNAS (2021) and Annals of Epidemiology (2022).
This statewide research estimated the prevalence of current and past SARS-CoV-2 infection among Ohio adults when routine case counts could not provide a population-level measure of epidemic burden. My work integrated probability sampling, multiple diagnostic tests, Bayesian multilevel poststratification, and adjustment for nonresponse, sparse positives, and imperfect tests. The resulting studies were published in PNAS and Annals of Epidemiology and demonstrate substantive experience in infectious-disease surveillance as well as population-health methodology.
