Advanced Statistical Methods

With the development of medicine and data technologies, modern clinical trials are often situated to address sophisticated therapeutic questions that require advanced statistical techniques. This chapter introduces readers to several advanced statistical topics one may likely encounter in today’s clinical trials. These topics include multiple endpoints, subgroup analysis, site and operator heterogeneity, and time-to-event outcomes. Each topic will be covered in a separate subsection. The subsection on multiple endpoints, which are used to measure multiple aspects of a disease, includes a discussion of advanced multiplicity adjustment and composite endpoints construction. Subgroup analysis introduces methods to evaluate efficacy and conduct hypothesis testing in multiple subpopulations in addition to the overall population. Site and operator heterogeneity can be considered as a case of subgroup analysis of special importance for surgical procedures. Meta-analysis methods and mixed models are discussed. The time-to-event subsection introduces basic concepts of time-to-event derivation, censoring, and common analysis techniques such as Kaplan–Meier curve, proportional hazards regression, and restricted mean survival.