Use of cost-effectiveness analysis in planning cancer chemoprophylaxis trials

We present a conceptual framework for the use of cost-effectiveness analysis in planning cancer chemoprophylaxis trials and evaluating such choices as sample size, study design, diagnostic evaluation, and duration of follow-up. The approach is illustrated by cost-effectiveness calculations for a trial of synthetic retinoids as cancer prophylaxis agents. Our work emphasizes the need for studies large enough to have a reasonable power of detecting 10%-20% reductions in mortality. Conversely, we note that rare but serious idiosyncratic side effects such as those seen the the "swine-flu" vaccination program are likely to escape detection even in trials involving tens of thousands of subjects.