Application of Group Sequential Methods to Clinical Trial

History & Objective: One of the most important issues in scientific research is determination of sample size for caculation of statistical indices such as mean and ratio. Researchers, usually use a fix confidence interval of 95%, a test power of 90% with fixed error and one or two sample formula to estimate a fix sample size. Due to some ethical, administrative and economical reasons using fix sample size in clinical trials is not practical. Group sequential methods are part of design and analysis of experiments. In group sequential methods, we try to solve problems by reducing the sample size and choosing a suitable test such as Pocock, O’Brien-Fleming, Lan-DEMets and the like. The purpose of this study is to evaluate the application of group sequential methods in determination of sample size and to analyze the results of the effect of single and multiple courses of corticosteroid therapy on Pretem Premuture Rupture of Memberance (PPROM) with a gestational age of 28-36 weeks.

Results: Analysis of data showed application of group sequential method reduced both the sample size and the length of the study.