Abstract
This study investigates the changes in physical church closings years 2013 to 2019 in New York City (NYC), Philadelphia, and Baltimore and the association with COVID-19 infection rates. We applied Bayesian spatial binomial models to analyze confirmed cases of COVID-19 as of February 28, 2022, in each city at the zip code-level. A one unit increase in the number of churches closed corresponded to a 5% higher COVID-19 infection rate, in NYC (rate ratio = 1.05, 95% credible interval = 1.02-1.08%), where the association was significant. Church closings appears to be an important indicator of neighborhood social vulnerability. Church closings should be routinely monitored as a structural determinant of community health and to advance health equity.