Deaths averted by vaccination due to reduced transmission can exceed those from direct protection of vaccinated individuals for SARS-CoV-2

Vaccination programs often focus on direct protection of vaccinated individuals against disease and discount reductions in transmission, which can result in preventable disease and death. Initial clinical trials for most COVID-19 vaccines only measured direct protection, and dosing and vaccine selection decisions have, so far, ignored effects on transmission. Here we provide a novel empirical framework for incorporating effects of COVID-19 vaccination on transmission in a continuous dynamic immune landscape. We quantified relationships between neutralizing antibody titers and vaccine effectiveness for both susceptibility (VE _S ) and infectiousness (VE _I ) and quantified changes in VE with waning and boosting of immunity. We used these relationships to quantify the impact that additional doses of mRNA vaccines (BNT162b2 and mRNA-1273) could have had in reducing transmission and deaths caused by the deadliest SARS-CoV-2 variant, Delta, in Autumn 2021. Neutralizing antibodies waned 8-fold in 2021 over the six months following initial vaccination with mRNA vaccines, which reduced VE _S 33-38% (from 75-81% to 47-54%) and VE _I 62-65% (from 47-57% to 16-22%) against the Delta variant. Third doses increased neutralizing antibody titers 13-26-fold, which more than restored VE and reduced the relative risk of transmission 7-10-fold. Administering third doses by September 1, 2021, could have reduced the effective reproductive number R _t by 19%, stopped surges in transmission in many populations, and averted an estimated 113,000 deaths in the United States, which is 2.6-12.4 fold higher than direct effects in vaccinated individuals. Vaccination programs that incorporate effects on transmission in trial design, vaccination frequency, and vaccine choice are needed to address current and future public health challenges.