Vaccine efficacy against naturally asymptomatic infections: A novel estimand for quantifying vaccine effects

The naive approach to estimating the effects of a vaccine on asymptomatic infections, which compares the risk of asymptomatic infection among vaccinated and unvaccinated individuals, can be misleading because it is comprised of two effects: the vaccine preventing asymptomatic infections and the vaccine converting symptomatic to asymptomatic infections. When the latter effect is strong, vaccines can appear harmful with respect to asymptomatic infections. Using a causal principal stratification framework, we formalize an estimand, vaccine efficacy against naturally asymptomatic infection (VENAI), that describes the effectiveness of a vaccine in preventing asymptomatic infections among individuals who would naturally (i.e., in the absence of vaccine) be expected to be asymptomatic. This estimand excludes vaccine effects that convert symptomatic cases to asymptomatic infections, and we demonstrate how this makes it a more natural analogue of the usual vaccine efficacy estimands against infection and symptomatic disease. We describe the assumptions under which this estimand can be identified and estimated from randomized and observational studies. We further identify and estimate bounds that do not require cross-world independence assumptions and characterize sensitivity analyses around the main assumption needed for identification. Finally, we apply these methods to a randomized trial of the COVID-19 mRNA-1273 vaccine. In this trial, VENAI was higher than standard estimates of efficacy against asymptomatic infections and was similar in magnitude to efficacy against any infection. Reporting VENAI in vaccine trials in addition to other vaccine effects would improve interpretability, could broaden understanding of vaccine impact on transmission, and provide insights into immunological mechanisms.