Qian Li, Laura Lu, Mo Huque (FDA, USA)
A Flexible Multiplicity Adjustment Approach and Its Application
We propose a general multiplicity adjustment approach that is flexible enough to allow one to specify decision criteria more freely compared to some of the conventional approaches, such as Bonferroni and Simes approaches. In fact, Bonferroni and Simes approaches are special cases of this flexible approach. Due to the flexibility of this approach, decision criteria can be chosen based on the direction of alternatives for the purpose of enhancing power. An example is illustrated for the application of this approach. In this example, several decision criteria are discussed to control the overall type I error when more than one studies were conducted for the same efficacy claim in clinical trial setting.