Asian Journal of Probability and Statistics

Clinical trials demand a comprehensive evaluation of multiple endpoints to provide a thorough understanding of intervention efficacy and safety. In response to a discerned gap in existing literature, this study introduces an innovative global nonparametric testing procedure, grounded in multivariate ranks, for the holistic analysis of these diverse endpoints. Unlike conventional approaches that heavily rely on pairwise comparisons for individual endpoints, our method takes a novel approach by directly incorporating multivariate ranks. This methodology builds on the strengths of previous models while addressing their limitations, ensuring a more nuanced and robust assessment. By considering the joint ranking of all endpoints, our approach exhibits heightened resilience against diverse data distributions and common censoring mechanisms commonly encountered in clinical trials. The proposed method emerges from a thoughtful integration of existing models, contributing to the methodological evolution in the field. To illustrate its superiority, extensive simulations have been conducted. The results unequivocally demonstrate the superior performance of our multivariate rank-based approach, showcasing its ability to effectively control type I error and achieve higher power compared to conventional rank-based methods. This empirical validation not only underscores the method's efficacy but also highlights its versatility and robustness across a spectrum of clinical trial settings.