Modeling Vaccine Trials in Epidemics with Mild and Asymptomatic Infection

Abstract

Background: Vaccine efficacy against susceptibility to infection (VES), regardless of symptoms, is an important endpoint of vaccine trials for pathogens with a high proportion of asymptomatic infection, as such infections may contribute to onward transmission and outcomes such as Congenital Zika Syndrome. However, estimating VES is resource-intensive. We aim to identify methods to accurately estimate VES when only a limited amount of information is available and resources are constrained. Methods: We model an individually randomized vaccine trial by generating a network of individuals and simulating an epidemic. The disease natural history follows a Susceptible, Exposed, Infectious and Symptomatic or Infectious and Asymptomatic, Recovered model. The vaccine is leaky, meaning it reduces the probability of infection upon exposure. We then use seven approaches to estimate VES, and we also estimate vaccine efficacy against progression to symptoms (VEP). Results: A corrected relative risk and an interval censored Cox model accurately estimate VES and only require serologic testing of participants once at the end of the trial, while a Cox model using only symptomatic infections returns biased estimates. Only acquiring serological endpoints in a 10% sample and imputing the remaining infection statuses yields unbiased VES estimates across values of R0 and accurate estimates of VEP for higher values of R0. Conclusion: Identifying resource-preserving methods for accurately estimating VES is important in designing trials for diseases with a high proportion of asymptomatic infection. Understanding potential sources of bias can allow for more accurate VE estimates in epidemic settings.