Estimating the Association Between COVID-19 Vaccine Coverage and Deaths in Connecticut

• Activity Number: 90 - Novel Statistical Methods for COVID Pandemic and Other Current Health Policy Issues
• Type: Contributed
• Date/Time: Monday, August 9, 2021 : 10:00 AM to 11:50 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Samantha G Dean* and Olga Morozova and Matthew Cartter and Amanda Durante and Forrest W. Crawford
• Companies: Department of Biostatistics, Yale School of Public Health and Program in Public Health, Stony Brook University and Connecticut Department of Health and Connecticut Department of Health and Yale University
• Keywords: immunizations; mortality; Sars-cov2; coronavirus

Abstract

Connecticut’s COVID-19 vaccination program has been one of the more successful in the U.S, with 61% of residents fully vaccinated as of June 30, 2021. COVID-19 case counts have declined statewide. To study the causal impact of vaccine uptake in Connecticut, we fit a log-link Poisson marginal structural model for case and death counts as a function of vaccine coverage in census tracts from December 14 to May 30, 2021. We adjusted the coverage model for demographics, prior case counts, and prior deaths within each tract. We lagged cases by 2 weeks and deaths by 4 weeks to account for the time between vaccination and immunity, and between infection and diagnosis or death. Our analysis measured the effect of coverage on deaths and cases that occurred in non-congregate (community) settings. In the first six months of the vaccine rollout, each additional 1 percent of vaccine coverage was associated with a 5.7 percent (95% CI 5.3, 6.4) reduction in the weekly tract death rate. Further, each additional 1 percent in vaccine coverage was associated with a 6.3 percent (95% CI 6.3, 6.4) reduction in the weekly tract case rate.