Comparison Of Design Efficiency For Two-phase Bioequivalence Studies

Bioequivalence evaluation is one of the important topics in pharmacokinetics,and it is also a hot field of statistical application in recent years.Two-stage bioequivalence studies have been approved by many national bioequivalence guidelines,but how to ensure target power while controlling Type I Error when implementing two-stage bioequivalence studies is a great challenge.This paper reviews the current two-stage design bioequivalence research methods published in domestic and foreign literature,and introduces in detail the research strategies of literature methods,calibration methods for testing standards,sample size reestimation and other methods in order to provide reference for domestic drug sponsors to carry out two-phase bioequivalence studies.Based on the background of staged evaluation and sample size adjustment of two-stage bioequivalence studies,this paper combines group sequential design with adaptive adjustment for bioequivalence evaluation.The design efficiency optimization scheme is established for the oriented group sequential design,and the specific contents are as follows:Firstly,the basic principles of bioequivalence test and two-stage design,their status and importance in clinical medical trials,the proposal and design scheme of two-stage bioequivalence study,and specific measures to control the cumulative Type I Error rate are briefly introduced from the theoretical aspect.Secondly,using SAS and Win Nonlin statistical software,citing the project data in the technical report,the pharmacokinetic parameter model was established,the statistical data was analyzed,and good bioequivalence evaluation results were obtained through variance analysis and equivalence evaluation.Finally,combining group sequential design and adaptive adjustment,starting from the direction of conditional power,a promising zone was constructed to re-estimate the sample size,so as to make an early termination or sample size re-estimation decision,by adjusting the final test statistics to control the probability of Type I Error.The effectiveness of the method was verified by examples,and unnecessary waste of interim analysis can be reduced in the sample size re-estimation stage.