Bias reduction for risk ratio and vaccine effect estimators

We examine the structural bias for established estimators of vaccine effect s on susceptibility and for newer estimates of vaccine effects on infectiou sness. We then propose and analyse new bias corrections for vaccine effect estimators of both susceptibility and infectiousness, as well as their comb ined effect on infection transmission. Each estimator is evaluated empirica lly with computer simulations. Of the estimators examined in this paper, th ose with the least bias and root mean squared error are computed by adding one to the positive count in the placebo population. We also identify a sou rce of bias for a standard Bayesian estimator of risk ratios. Copyright (C) 2001 John Wiley & Sons, Ltd.