Keywords: prioritized endpoints, Wilcoxon test, informative censoring, weighted method, bootstrap
We consider a two-group randomized clinical trial of prioritized endpoints, where mortality affects the assessment of a follow-up continuous outcome. Using the continuous outcome as the principal outcome, we incorporate mortality via the worst-rank paradigm to define a single composite endpoint. Then, we develop a weighted Wilcoxon test statistic to analyze such a composite endpoint. We determine the optimal weights for the Wilcoxon test statistic that maximize its power. In the process, we derive a formula for power and sample size calculations for the test statistic. Finally, we run simulation studies to assess the characteristics of the test and demonstrate its accuracy. As an illustration, we apply the proposed method to a dataset from an acute ischemic stroke clinical trial of normobaric oxygen therapy.