Sequential trials adaptive designs

An alternative (or addition) to repeating the fixed-sample size trial is to use a sequential design in which the trial is run until a useful result is reached. These adaptive designs, in which decisions are taken on the basis of results to date, can assess results on a continuous basis as data for each subject that becomes available or, more commonly, on groups of subjects... 

Group sequential designs, particularly in noninferiority safety studies where many events may occur after accrual during the follow-up phase, are unlikely to stop early. The design described here, instead, uses accruing information to predict how many events the current cohort is likely to provide with addition follow-up. Therefore, rather than having many patients with short exposure times, it relies on a smaller number of patients with greater exposure times—with that optimal number chosen adaptively and based upon internal trial data. 

---