A group sequential design provides detailed specifications for a group sequential trial. In addition to the usual specifications, it provides the total number of stages (the number of interim stages plus a final stage) and a stopping criterion to reject, accept, or either reject or accept the null hypothesis at each interim stage. It also provides critical values and the sample size at each stage for the trial.
At each interim stage, all the data collected up to that point are analyzed, and statistics such as a maximum likelihood test statistic and its associated standard error are computed. The test statistic is then compared with critical values generated from the sequential design, and the trial is stopped or continued. If a trial continues to the final stage, the null hypothesis is either rejected or accepted.
Below are highlights of the capabilities of the SAS/STAT procedures that perform group sequential design and analysis: