Design and Analysis of Clinical Trials with Multiple Co-primary Endpoints

点击翻译按钮获取中文摘要

Statistical methodologies for design and analysis of clinical trials with multiple co-primary endpoints were developed, where the study objective of the trials is to demonstrate statistical significance for all of the endpoints.(1) Power and sample size determinations for each scale of endpoint were provided. The derived convenient sample size formula for continuous normally distributed endpoints provided a clear relationship between the required sample size and the correlations among the endpoints.(2) A further investigation was necessary to improve the testing method combining superiority and non-inferiority of a new treatment compared to a control treatment.(3) The required sample sizes of two methods to alleviate conservative approach of testing co-primary endpoints were practically comparable in typical clinical trial settings.