Statistical Analysis of Adverse Events in Randomized Clinical Trials

Adverse events are generally tabulated and listed in statistical reports in randomized clinical trials. This paper introduces how to apply statistical methods to assess adverse events using SAS. In addition to the crude percentage (rate), adverse events are also adjusted by exposure time or multiple repeat. The paper also introduces the commonly used chi-square test and Fisher‟s exact test, and other statistical methods such as Cochran-Armitage trend test, nonparametric approach and generalized linear mixed model. The application of statistical methods depends on study designs, data type and investigation purpose. Several clinical trial examples such as dose-response studies, cross-over trials, are analyzed using SAS for illustration purposes.