含量测定类方法验证中的统计学评价

目的：针对双环醇片主成分的测定方法，论证含量测定类方法验证的科学性和可靠性，为今后该类方法的验证研究提供更严谨的模板或范例。方法：通过科学的实验设计，制定合理的实验方案，获得专属性、线性与范围、准确度和精密度等性能参数。结果：双环醇片中主成分与有关物质可完全分离。在标示量70%~130%的范围内线性关系良好（决定系数R²>0.999 9，剩余标准偏差RSME=0.299 6）；在80%~120%的5个浓度点下的相对偏倚均不超过0.78%，其95%置信限均不超过±2%；各浓度下的中间精密度均不超过0.89%，中间精密度的置信上限均不超过1.43%；方法的总变异在120%浓度下最大，为0.87%，所对应的95%/90%容忍区间为98.11%~102.78%，95%预测区间为98.95%~101.95%；对可接受标准为90.0%~110.0%的双环醇片，其方法能力指数均超过2，即该方法的误判率低于1/100万。结论：该方法经验证可满足其对双环醇片含量定量测定的预期用途。同时通过实验设计可获得更多科学可靠的方法性能指标，这些指标使得方法验证的结论更加可靠和直观，且能够从整体把握方法的适用性。新提供的方法容忍区间、预测区间和方法能力指数等参数，为今后判断方法确认和转移是否成功提供了依据。

Objective: To demonstrate the scientificity and reliability of the method validation for major content determination in bicyclol tablets as an example in order to provide a more rigorous template or example for future validation of such methods. Methods: Through scientific experiment design and reasonable experimental scheme to obtain performance parameters, such as specificity, linearity and range, accuracy and precision. Results: The main components in bicyclol tablets could be completely separated from the related substances. The linearity was excellent in the range of 70% to 130% of labeling quantity(R²>0.999 9,RSME=0.299 6). All relative biases did not exceed 0.78% at the five concentration points of 80%-120%, and all 95% confidence limits did not exceed ±2%. All intermediate precisions were not more than 0.89% and all upper confidence limits of the intermediate precision were not more than 1.43%. The maximum total variation of the method was 0.87% at the concentration of 120% and the corresponding 90% tolerance interval with 95% confidence level was 98.11%-102.78%.The prediction interval with 95% confidence level was 98.95%-101.95%. For bicyclol tablets with the acceptable criteria of 90.0%-110.0%, the method capability index was more than 2, which means the misjudgment rate of this method was less than one millionth. Conclusion: The method has been validated to meet its intended use for quantitation of bicyclol tablets. More scientific and reliable method performance indicators can be obtained through the design of experiment,which make the conclusions of the method validation more reliable and intuitive,and can grasp the applicability of the method from the whole. The newly provided tolerance interval,prediction interval and the method capability index,etc.,provide a basis for judging the success of method verification and transfer in the future.

[1] USP 41-NF 36[S]. 2018:7665(<1225>Validation Of Compendial Procedures)
[2] 中华人民共和国药典2015年版.四部[S]. 2015:374 Ch P 2015. Vol Ⅳ[S]. 2015:374
[3] ICH Q2(R1):Harmonised Tripartite Guideline Validation of Analytical Procedure:Text and Methodology[S]. 2005
[4] 唐静.双环醇片防治抗癌药物所致肝损害临床研究[J].中国药业, 2015, 24(16):43 TANG J. Clinical research on bicyclol tablet for preventing and treating anti-cancer drug induced liver[J]. China Pharm, 2015, 24(16):43
[5] USP 41-NF 36[S]. 2018:7663(<1210>Statistical Tools For Procedures Validation)
[6]<1200>Requirements for Compendial Validation[J/OL]. USP Pharmacoperial Forum, 2015, 41(5)[2017-08-07]. http://www.usppf.com/pf/pub/index.html
[7] 中华人民共和国药典2015年版.二部[S]. 2015:93 ChP 2015. Vol Ⅱ[S]. 2015:93
[8] ERMER J, MILLER J. Method Validation in Pharmaceutical Analysis[M]. Weinheim:Wiley-Vch,2005:30
[9] USP 41-NF 36[S]. 2018:1803(<1033>Bilogical Assay Validation)
[10] FDA. Analytical Procedures and Methods Validation for Drugs and Biologics Guidance for Industry[S]. 2015
[11] 谭德讲,马双成.药品监督与检定中的统计学应用[M].北京:中国科学技术出版社,2011 TAN DJ, MA SC. Statistical Application in Drug Supervision and Verification[M]. Beijing:China Science and Technology Press, 2011
[12] 胡良平.检验医学科研设计与统计分析[M].北京:人民军医出版社, 2004 HU LP. Scientific Research Design and Statistical Analysis in Laboratory medicine[M]. Beijing:People' Military Medical Press, 2004
[13] USP 41-NF 36[S]. 2018:6706(<1010>Analytical Data-Interpretation and Treatment)
[14] DESIMONI E, BRUNETTI B. Uncertainty of measurement and conformity assessment:a review[J]. Anal Bioanal Chem, 2011, 400(6):1729
[15] 张横,谭德讲.统计区间的作用及在药品质量控制中的应用[J].药物分析杂志, 2013, 33(11):1840 ZHANG H, TAN DJ. The effects of statistical intervals and their application in pharmaceutical quality control[J]. Chin J Pharm Anal, 2013, 33(11):1840