Multiple Testing in Group Sequential Design

In recent clinical trials, it has become increasingly common to see the analysis of multiple hypotheses. This includes scenarios such as comparing multiple treatments with a common control arm, assessing the effectiveness of a new drug across multiple endpoints, and evaluating treatment responses in different dose levels for both overall and subpopulations. This book chapter provides a comprehensive overview of multiple test procedures within the context of group sequential designs, with a focus on graphical approaches. Moreover, we introduce two options for inference: comparing either adjusted significance levels with unadjusted p-values or adjusted p-values with the unadjusted significance level \(\alpha \) . This book chapter provides not only technical details of the discussed methods, but also offers comprehensive examples with detailed step-by-step implementations, along with comparisons of the methods. Additionally, we incorporate practical considerations from the viewpoint of pharmaceutical industry practitioners. Finally, the chapter provides a summary and offers high-level suggestions for multiplicity designs.