Epidemiological factors in clinical trial design.

Randomised, double-blind, placebo-controlled clinical trials are human experiments in which the process of randomisation and blinded observation greatly simplify the interpretation of the results. However, epidemiological issues when properly addressed allow the optimal preparation of the study with anticipation of factors to be included in the analysis which are independent of the randomisation process but can affect the results, for example, the effect of age, dose, time since vaccination and exposure to other diseases. Alternative methods of estimating vaccine efficacy may be necessary in some settings, and such methods may have limitations which must be acknowledged in interpretation. This paper reviews some recently published studies of vaccine efficacy for the application of epidemiological principles in design and analysis.