New designs for clinical trials.

Abstract

The pressures of ethics and equity on clinical trials have always been severe. The highest level of evidence about the usefulness of therapy is that obtained from well-planned, carefully conducted randomized, and where possible, double-blind clinical trials. Traditional frequentist statistics have been used predominantly, but a Bayesian approach may offer more flexibility by leveraging more information from the results of a clinical trial. Innovative designs can minimize the number of patients included in a trial and decrease the chance of maintaining a patient in an unfavorable treatment arm.