Background: The protective effect of a vaccine following an outbreak is often measured by the vaccine efficacy statistic, namely one minus the ratio of attack rates in vaccinees and non-vaccinees. This quantity is not an adequate measure of the population-level benefits of the vaccine.
Methods: We discuss two measures of the effectiveness of a vaccination programme. The first is the commonly used vaccine efficacy statistic. This is called here the individual vaccination effectiveness. The second measure, called the population effectiveness, is defined as one minus the ratio of the overall (or average) attack rate in the population when the vaccination programme is implemented to the expected attack rate in the same population without vaccination. We outline a method for computing the population effectiveness following an outbreak of a directly transmitted acute infectious disease in a closed heterogeneous population. We then explore and compare the behaviour of the two measures of vaccination effectiveness under various conditions.
Results: The population vaccination effectiveness is more robust than individual effectiveness to factors that may interfere with the evaluation of the performance of vaccination. Such factors are non-uniform vaccination, changes in contact patterns by vaccinees, and the ability of the vaccine to reduce infectiousness.
Conclusion: The population vaccination effectiveness is a more adequate measure of the population-level benefits of a vaccination programme. The main disadvantage of this measure is that it cannot be readily calculated from observed attack rates.