Randomized, clinical trials are the strongest method we have to define the efficacy of new drugs. Recently, however, novel agents that show remarkable responses with minimal toxicity in patients with advanced cancer have been discovered [1], and these present a challenge for the ethical conduct of randomized trials. Consequently, there is a debate on the randomized trial as the gold standard for bringing these agents to approval.
For several decades, the standard for determining that a randomized trial is ethical has been a matter of discussion. Some doctors espouse the uncertainty principle, whereby randomization to treatment is acceptable when an individual doctor is genuinely unsure which treatment is best for a patient. Others believe that ‘clinical equipoise,’ reflecting collective professional uncertainty about treatment, is the soundest ethical criterion [2, 3]. Both sides agree that uncertainty as to the more effective therapy is necessary for a randomized trial to be ethical, the point of contention being only whether it is individual physician uncertainty or collective uncertainty of the medical community. Until the recent proposal to reject the equipoise (balance of uncertainty) standard [4], it has been widely accepted that doubt as to the better arm must exist at the start of the trial, and safeguards are also required so that there is intermittent monitoring of the trial by an independent committee, who can stop the trial should one of the arms prove to be superior.
The problem arises when substantial evidence suggests that a new treatment has strikingly superior efficacy, and the equipoise standard may therefore not be met. To address this concern, it has recently been proposed that, to inform coverage decisions for populations, payors may require rigorous assessment of a new agent beyond the point at which physicians and informed patients would …