ABSTRACT
Multiple-endpoint problems can arise in many different situations. For in-
stance, when there is a single primary efficacy endpoint with one or more
secondary endpoints, we try to claim the drug effect on primary and sec-
ondary endpoints; when there are two endpoints, at least one needs to be
statistically significant to claim efficacy of the test drug; in a clinical trial
we can use a surrogate endpoint for an accelerated approval and a clinically
important endpoint for a full approval. In this chapter, we will first review
the single-stage and stepwise methods in classical designs, then discuss
multiple-endpoint issues in adaptive designs. We will use trial examples to
give step by step instructions.