Group Sequential Methods with Applications to Clinical Trials

Group sequential methods answer the needs of clinical trial monitoring committees who must assess the data available at an interim analysis. These interim results may provide grounds for terminating the study-effectively reducing costs-or may benefit the general patient population by allowing early dissemination of its findings. Group sequential methods provide a means to balance the ethical and financial advantages of stopping a study early against the risk of an incorrect conclusion.
Group Sequential Methods with Applications to Clinical Trials describes group sequential stopping rules designed to reduce average study length and control Type I and II error probabilities. The authors present one-sided and two-sided tests, introduce several families of group sequential tests, and explain how to choose the most appropriate test and interim analysis schedule. Their topics include placebo-controlled randomized trials, bio-equivalence testing, crossover and longitudinal studies, and linear and generalized linear models.
Research in group sequential analysis has progressed rapidly over the past 20 years. Group Sequential Methods with Applications to Clinical Trials surveys and extends current methods for planning and conducting interim analyses. It provides straightforward descriptions of group sequential hypothesis tests in a form suited for direct application to a wide variety of clinical trials. Medical statisticians engaged in any investigations planned with interim analyses will find this book a useful and important tool.

Booknews

In clinical trials of new medical treatments, it is standard practice to monitor accumulating data with specialized statistical methods called group sequential methods. This work surveys and extends current methods for planning and conducting interim analyses, and describes group sequential stopping rules which can reduce average study length while controlling error probabilities. Procedures are presented in a manner which allows their easy implementation to a variety of data types arising from clinical trials. Of interest to medical statisticians in the pharmaceutical industry and medical research, as well as academic statisticians. The authors are professors of statistics at the University of Bath, UK, and at Cornell University. Annotation c. Book News, Inc., Portland, OR (booknews.com)