Data verification and sensitivity analysis of statistical analysis in bioequivalence studies / 药学实践杂志

ABSTRACT

Objective To sort out the key points, problems and countermeasures of data verification through data verification of statistical analysis in 18 bioequivalence studies, and provide reference for improving the statistical analysis reports of related research. Methods SAS and WinNonlin software was used to illustrate that whether the random number tables, main pharmacokinetic parameters and bioequivalence data could reproduce the corresponding results in the original statistical analysis reports. Results Among the 18 studies, sensitivity analysis was supplemented or re-performed for 5 studies due to sampling time deviation of individual subjects or adjustment of sensitivity data sets, resulting in differences in pharmacokinetic parameters from the original statistical analysis report, but same for the bioequivalence evaluation. Other verified data was consistent with the original statistical analysis reports. Conclusions The verification of statistical analysis data of bioequivalence studies is extremely important, and the problems found in this paper should be fully considered when writing such statistical analysis reports.