Geographical Variation in COVID-19 Transmission Potential and Case Burden, West Virginia, 2020-2023

Objective Substate-level analysis reveals geographical variation in COVID-19 epidemiology and facilitates improvement of prevention efforts with greater granularity. Methods We analyzed daily confirmed COVID-19 case count in West Virginia and its 9 regions (March 19, 2020-March 9, 2023). Nonparametric bootstrapping and a Poisson-distributed multiplier of 4 were applied to account for irregular and under-reporting. We used the R package EpiEstim to estimate the time-varying reproduction number R_t with 7-day-sliding-windows (2020-2023) and non-overlapping-time-windows between 5 policy changes (2020 only). Poisson regression was used to estimate the incidence rate ratio (IRR) between each region and West Virginia (2020, 2021, and 2022). Results Statewide R_t fluctuated over the study period, with the highest in March 2020 (close to 2) and the lowest R_t (<1) seen in June 2020. The Stay-at-home Order, Face Mask Mandate, and Virtual Learning Resumes saw 38.7% (95% credible interval [CrI]: 21.9%-57.5%), 10.6% (95% CrI, 3.2%-18.9%), and 9.4% (95% CrI, 3.2%-15.4%) corresponding decreases in R_t statewide. All regions experienced incidence rates different from the state. The IRRs ranged from 0.32 (95% CI, 0.32-0.33) (Northern region) to 1.90 (95% CI, 1.87-1.94) (Wood-Jackson region) in 2020. Conclusions Policies reducing human contacts, e.g., Stay-at-home Order and Virtual Learning Resumes, effectively reduced transmission statewide.