Abstract

KIE: Marquis analyzes the design of randomized clinical trials of anticancer drugs and concludes that randomization is incompatible with the ethical obligations of the physician patient relationship. Statistical requirements and administrative complexities make it difficult to halt a trial before it has run its course, yet it is unethical to continue once there is evidence that one therapy is superior to another. Strategies for resolving the dilemma are proposed, but are shown to be ethically flawed upon closer examination. There appears to be no solution to the problem of reconciling the demonstrable benefits of randomized trials with inevitable violations of the therapeutic obligation.