Vaccination with BNT162b2 reduces transmission of SARS-CoV-2 to household contacts in Israel

Abstract

The effectiveness of vaccines against COVID-19 on the individual level is well established. However, few studies have examined vaccine effectiveness against transmission. We used a chain binomial model to estimate the effectiveness of vaccination with BNT162b2 [Pfizer-BioNTech messenger RNA (mRNA)-based vaccine] against household transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Israel before and after emergence of the B.1.617.2 (Delta) variant. Vaccination reduced susceptibility to infection by 89.4% [95% confidence interval (CI): 88.7 to 90.0%], whereas vaccine effectiveness against infectiousness given infection was 23.0% (95% CI: -11.3 to 46.7%) during days 10 to 90 after the second dose, before 1 June 2021. Total vaccine effectiveness was 91.8% (95% CI: 88.1 to 94.3%). However, vaccine effectiveness is reduced over time as a result of the combined effect of waning of immunity and emergence of the Delta variant.

Fig. 1.. Schematic representation of the multiple imputation process for an example household.

Each infected household member is associated with: (A and D) a distribution for time from onset of infectiousness to testing; (B and E) a distribution for the infectious period; and (C and F) a distribution for the latent period to infer the time of infection. The filled ovals represent observed events, and the circles and stars represent unobserved events in the infection timeline. Panels (A to C) and (D to F) represent two possible sample sets from the delay distributions, each with a different index case, who is not necessarily the first person to test positive in the household. We generated 100 samples of the latent data for each infected individual.