Causal Inference of Vaccine Effects on Infectiousness

• Activity Number: 6
• Type: Invited
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): M. Elizabeth Halloran and Michael G. Hudgens*+
• Companies: Fred Hutchinson Cancer Research Center/University of Washington and The University of North Carolina at Chapel Hill
• Keywords: causal inference ; principal stratification ; interference ; vaccine ; infectiousness

Abstract

Evaluating the effect of vaccination on infectiousness has important public health consequences. If a vaccine does not protect well against infection, it could still substantially reduce the number of cases if vaccination reduces infectiousness of infected individuals. Assessing causal vaccine effects on infectiousness incorporates both principal stratification based on potential infection outcomes under vaccine and control and interference between individuals within transmission units. Here we define causal estimands for vaccine effects on infectiousness in transmission units of size two. Four different scenarios are considered. In the general case, both individuals can be exposed to infection outside the transmission unit and both can be assigned either vaccine or control. In the simplest case, only one individual can be exposed outside the transmission unit and the same individual can be assigned either vaccine or control, while the other is not vaccinated. We demonstrate the identifiability of the effects under different assumptions about postinfection selection bias. We derive large sample bounds for causal vaccine effects on infectiousness.