| EFFICIENCY GAINS USING GROUP SEQUENTIAL METHODS FOR CLINICAL TRIALS
| Michelle Sonnenberg, Master's Candidate, Department of Statistics, Colorado State University
Interim analyses are often made during clinical trials to declare, prior the pre-designated end of the trial, the efficacy of a drug or the futility of the trial to produce favorable results. Since multiple analyses are being done in the same trial, unless adjustments are made to the interim analyses, this can lead to an inflation of the overall type I error rate. Group sequential methods provide a framework for completing these interim analyses while maintaining the desired overall level of significance. This seminar will summarize and discuss the Pocock, O’Brien-Fleming, and Wang-Tsiatis group sequential methods, and subsequent efficiency gains due to their implementation in the clinical trial setting. Specifically considered will be the case of a two group clinical trial, with normally distributed data and equally spaced interim analyses. The case of unequally spaced analyses will also be addressed in the context of alpha spending functions. A comparison of the methods will be given considering efficiency gains of the various methods with respect to sample size and power.