Examining the interaction between area-level immunity coverage and individual-level immunity to help quantify indirect herd benefits: a population-based retrospective cohort study of COVID-19 diagnoses and deaths.

Empirical evidence on the indirect herd benefits of COVID-19 vaccination and/or prior infection is limited. We examined how area-level immunity interacts with individual-level immunity to affect COVID-19 diagnoses and deaths. Ontario residents aged ≥18 years were followed from August 1, 2021, to January 30, 2022. Individual-level immunity was defined as receipt of a primary series of COVID-19 vaccines or a positive SARS-CoV-2 test in the past 165 days. Area-level immunity was based on the proportion of immune individuals in an individual's residing area. Logistic regression and cause-specific hazard models were used to examine the relationship between immunity and COVID-19 diagnosis, and between immunity and COVID-19 death, with an interaction term between individual-level and area-level immunity. Of 11 122 816 adults, 7 518 015 (67.6%) were classified as having individual-level immunity at baseline. After accounting for potential confounders, area-level immunity (highest vs. lowest quintiles) was associated with lower odds of COVID-19 diagnosis. Higher area-level immunity (highest vs. lowest quintiles) was also associated with lower hazard of COVID-19 death among non-immune individuals (hazard ratio, 0.77 [0.60, 1.00]). Findings provide evidence supporting the herd benefits of vaccination or prior infection on COVID-19 diagnosis and deaths, and provide insights for interpreting vaccine effectiveness estimates in the context of herd immunity.