ABSTRACT
I. INTRODUCTION Recent developments in group sequential methods have had a great impact on the design and analysis of randomized clinical trials. This is especially true for those trials that are of "pivotal" or "confirmatory" nature in support of a new drug application (NDA). The pros and cons of performing interim analyses on accumulating data in these trials have been discussed in numerous papers in the literature (e.g., Rodda et aI., 1988; Pocock and Hughes, 1989; Davis and Hwang, 1992), and its use for the determination of efficacy and safety is a frequent and necessary practice.
