ABSTRACT
Increasing use of statistical procedures that permit analyses of data collected in clinical trials prior to a trial reaching its preplanned conclusion has occurred over the last 30 years. Ideally, details of such procedures are incorporated into the protocol at the design stage. In doing so, the procedures from a statistical methodological point of view can be clearly presented. In addition, steps to be taken to ensure that by conducting such procedures the integrity of the trial will not be compromised, should also be clearly and unambiguously specified. Basically such procedures permit the analyses of specified efficacy or safety
endpoints as data accumulate over specified stages of a trial. Some of the more commonly used procedures are reviewed and discussed in this chapter. These include sequential procedures, group sequential procedures, stochastic curtailment procedures, and adaptive design methods.
