ABSTRACT
Multiple hypothesis testing is becoming an ever more prominent issue in clinical development. Much of this stems from the idea that sponsors want to learn as much as possible from every clinical trial. This leads to clinical trials with multiple objectives to be studied. These objectives could include comparing multiple doses, multiple endpoints, and multiple analyses such as interim analyses or futility analyses. The complexities of designs with multiple objectives have caused regulatory concerns as well (ICH 1998; CHMP 2002). One other trial design that will often lead to multiplicity issues is an adaptive trial design. An adaptive design will, by design, have an interim analysis in which decisions can be made and, depending on the type of adaptation, there could be a selection between which doses to continue, what is the primary hypothesis, or primary population of interest among
others. In order to ensure the validity of the trial, these issues have to be addressed in the statistical design.
