Relationship between neighbourhood-level killed oral cholera vaccine coverage and protective efficacy: evidence for herd immunity

Abstract

The effectiveness of vaccines in populations must consider both direct and indirect protection. This study reanalyses data from a large individually randomized oral cholera vaccine trial that was conducted in rural Bangladesh from 1985 to 1990. A recent analysis of the results of that trial showed that the proportion of people in household clusters who received the vaccine was inversely related to placebo incidence during the first year of surveillance, which was attributed to herd immunity. In this study we measure the relationship between neighbourhood-level oral cholera vaccine coverage and protective efficacy (PE) during a 2 year follow-up period, controlling for known effect modifiers. We link trial data to a household geographic information system to facilitate the neighbourhood-level analysis. Findings Neighbourhood-level PE can be partially explained by vaccine coverage after adjusting for ecological variables. The inverse relationship between vaccine coverage and efficacy illustrates that people living in high-coverage areas may be indirectly protected from cholera because people living around them are vaccinated.