We propose a new sample size calculation method for trial-based cost-effectiveness analyses. Our strategy is based on the value of perfect information that would remain after the completion of the study.
