Keywords: Ranking approach, parametric and non-parametric statistical tests, bootstrap method, HIV.
Clinical trial data have a variety of clinical and laboratory measures to assess a new treatment benefit over the standard care on the basis of outcomes such as death or heart failure hospitalization. Data from these studies are often analyzed with combining measures in one clinical outcome. We call the primary outcome as composite endpoint or combined endpoint. We perform an analysis on the composite endpoint with a ranking approach based on a principle that each subject in the trail is compared to rest of the subjects. In this research, we propose a parametric test and a non-parametric bootstrap procedure for comparing the effect of treatment groups. We conduct an extensive simulation study to assess the operating characteristics of the proposed methods and compare proposed methods with an existing method. We illustrate the use of the methods using public data from a study of HIV subjects.