• 中国核心期刊(遴选)数据库收录期刊
  • 中文科技期刊数据库收录期刊
  • 中国期刊全文数据库收录期刊
  • 中国学术期刊综合评价数据库统计源期刊等

中国药物评价 ›› 2022, Vol. 39 ›› Issue (6): 483-788.

• 评价技术与方法 • 上一篇    下一篇

分析数据缺失对生物等效性评价的影响

姚青, 刘莹, 王晓芳, 高宁, 王苗*   

  1. 新领先医药科技发展有限公司, 北京 100094
  • 收稿日期:2022-09-20 修回日期:2022-10-22 出版日期:2022-12-28 发布日期:2022-12-28

Analyzing the Impact of Missing Data on Bioequivalence Evaluation

  1. Leadingpharm Medical Technology Development Co., Ltd., Beijing 100094, China
  • Received:2022-09-20 Revised:2022-10-22 Online:2022-12-28 Published:2022-12-28

摘要: 目的:全方面分析数据缺失对生物等效性研究和临床试验结论的影响。方法:对存在缺失的结局数据进行完整意向性分析(Full intention-to-treat analysis,FITT分析)、按接受治疗分析(As-treated analysis,AT分析)、方差分析和受试制剂(Test preparation,T)与参比制剂(Reference preparation,R)比值分析,并比较Phoenix WinNolin和Excel统计软件计算的生物等效性(Bioequivalence,BE)结果,考察数据缺失对生物等效性评价的影响。结果:数据缺失会使生物等效性结论发生偏倚,序列错误和退出增大药动学参数周期间和个体间差异,对其他受试者T/R值影响不大。对于不均衡BE数据,Phoenix WinNolin和Excel软件计算结果不完全一致,临床结论可信度降低。结论:数据缺失对生物等效性评价有显著性影响,应尽可能减少数据缺失的发生,对于数据缺失的结局数据要进行全面分析以尽可能提高临床结论的可信度。
 

关键词: font-size:medium, ">数据缺失;生物等效性;FITT分析;AT分析;方差分析;T/R分析

Abstract: Objective:To comprehensively analyze the impact of missing data on the conclusions of bioequivalence studies and clinical trials. Methods:Full intention-to-treat(FITT) analysis, As-treated(AT) analysis, variance analysis and T/R analysis were performed on the outcome data, and the results of BE by Phoenix WinNolin and Excel statistical software were compared, and the impact of missing data on the bioequivalence evaluation was investigated. Results: Missing data would bias bioequivalence conclusions, sequence errors and dropouts would increase inter-week and inter-individual differences of pharmacokinetic parameters, and had little effect on the T/R values of other subjects. For unbalanced BE data, the calculation results of Phoenix WinNolin and Excel software are not completely consistent, and the credibility of clinical conclusions is reduced. Conclusion:Data missing has a significant impact on bioequivalence evaluation, the occurrence of data missing should be minimized, and the outcome data with missing data should be comprehensively analyzed to improve the reliability of clinical conclusions as much as possible.

Key words: font-size:medium, ">Missing data; Bioequivalence; FITT analysis; AT analysis; Variance analysis; T/R analysis

中图分类号: