Exclusion of clinically atypical or microbiologically mixed diarrhoeal episodes from outcome events in a field trial of oral cholera vaccines

Abstract

We investigated whether alternative clinical and microbiological criteria for outcome events affected estimates of vaccine efficacy in a randomized, double-blind field trial of B subunit-killed whole cell (BS-WC) and killed whole cell-only (WC) oral cholera vaccines among 62,285 rural Bangladeshi participants. At one year of follow-up estimates of vaccine protective efficacy (PE = 60%, P less than 0.0001 for BS-WC; PE = 54%, P less than 0.0001 for WC) against all treated diarrhoeal episodes associated with V. cholerae 01 were similar to estimates of efficacy against only those episodes which were clinically typical and unassociated with additional enteric pathogens (PE = 62%, P less than 0.0001 for BS-WC; PE = 52%, P less than 0.0001 for WC). In contrast, estimates of vaccine cross-protection against episodes associated with each of several agents antigenically related to V. cholerae 01 (LT-ETEC, non-cholera Vibrio sp, Aeromonas sp) were substantially reduced when mixed infections with V. cholerae 01 were excluded. We conclude that restrictive criteria intended to improve the specificity of the definition of cholera did not increase the detectability of vaccine efficacy against V. cholerae 01, but that exclusion of mixed infections with V. cholerae 01 was necessary to avoid false-positive conclusions about vaccine cross-protection against other potential target pathogens