|
Abstract:
|
When the use of a placebo is unethical, the efficacy of an experimental therapy must be inferred from noninferiority trials. The usual approach involves defining a fixed noninferiority margin as a function of the lower confidence bound of the estimate of the effect of the active control relative to placebo based on a meta analysis of historical trials. An alternative approach, sometimes called a synthesis approach, involves combining the estimate of the effect of the treatment relative to the control based on the noninferiority trial with the estimate of the effect of the control relative to placebo. This presentation will consider sample size considerations for these two approaches, focusing on issues such as the relative sizes of the historical and noninferiority trials, preservation of the control's effect, and the assumed benefit of the treatment relative to the active control.
|
