ABSTRACT
According to the U.S. Food and Drug Administration (FDA) Draft Guidance on Non-inferiority Trials,1 “The rst and most critical task in designing an NI study is obtaining the best estimate of the effect of the active control in the NI study (i.e., M1).” The FDA draft guidance on non-inferiority trials1 provides instances for which a non-inferiority margin can be dened in the absence of controlled clinical trials evaluating the active control effect. The circumstances are similar to those for which historically controlled trials can provide persuasive evidence.2 For example, there should be a good understanding or estimate of the outcome (e.g., spontaneous cure rate) without treatment, and the outcomes or cure rate for the active control from multiple historical experiences should be substantially different from those seen without treatment (e.g., substantially different spontaneous cure rates). The assumed effect of the active control in the setting of the non-inferiority would be conservatively chosen.
