Title: Mathematical Statistician
Street: 9201 Corporate Boulevard, HFD-725
Authors: Qian Li and Mohammad Huque
Title: A general multiplicity adjustment approach and its application to evaluating several indenpendently conducted studies
Abstract: Usually, in applying for market approval of a new drug, more than one similarly designed clinical trial is conducted to support efficacy or safety claims. How to evaluate these studies collectively and assess the decision error for a decision rule used for such evaluation can be a challenging statistical issue. In this paper, we propose a general multiplicity adjustment approach to evaluating several similarly designed and independently conducted studies collectively, and to control the decision error at a desired level. The Bonferroni correction is a special case of this general approach. A concept of an overall hypotheses, which is essentially union or intersection combinations of individual study's hypotheses, is used to define the decision error. Using this new approach, we can derive decision rules using p-value cut points that control the decision errors under certain given overall null hypotheses. Properties of the general multiplicity approach are discuss! ed as well, including the choice of overall hypothesis and its corresponding decision error, and the power characteristics. A couple of examples are presented to illustrate the application of the approach in evaluating several independently conducted studies.
KEY WORDS: Multiplicity; Clinical trials; Overall hypotheses; p-value cut points; Decision rules; Decision errors; Rejection regions; Optimal power.