The impact of indirect benefits of vaccination on postlicensure vaccine effectiveness estimates: A scenario analysis

Abstract

Vaccine efficacy is measured in randomized, prelicensure clinical trials where vaccination typically affords only direct protection to the vaccinated individual. Vaccine effectiveness is measured in postlicensure observational studies where vaccination might provide indirect benefits to a population as a whole in addition to directly protecting the vaccinated individual. The potential discrepancy in effectiveness and efficacy estimates would depend on the postlicensure study design. We developed a mathematical model to assess the impact of indirect benefits on vaccine effectiveness as measured by the common cohort study design under scenarios of homogenous and heterogenous vaccine allocation. We found that under the cohort design, effectiveness estimates equaled efficacy if either the indirect effects were assumed to be negligible or vaccine allocation in the community was homogenous. However, in presence of indirect benefits, effectiveness estimates would be biased upward compared with vaccine efficacy if one of the two sub-populations in the same study had a higher rate of vaccination. Because of indirect effects of vaccination, even in studies where other biases can be eliminated, the presence of distinct sub-populations with varying rates of vaccination can lead to discrepancies between effectiveness and efficacy estimates.