Smith W C, Sittampalam G S
Statistical and Mathematical Sciences, Eli Lilly and Company, Greenfield Laboratories, Indiana 46140, USA.
J Biopharm Stat. 1998 Nov;8(4):509-32. doi: 10.1080/10543409808835257.
The discovery, research, and development of a pharmaceutical product relies on the availability of validated assays for assessing product characteristics and drug effects in vivo and in vitro. Development of a validated assay is a multifaceted activity that provides many interesting challenges for bioanalytical chemists and statisticians. In this paper, the similarity condition for fundamental validity of an analytic dilution assay is reviewed as a basic concept underlying the validation of assays for pharmaceutical applications. The distinction between the validity and the acceptability of an assay is considered in terms of the characteristics evaluated during four stages of validation. Recent guidelines on the validation of analytical procedures published by the U.S. Food and Drug Administration are appraised from a statistical perspective, and statistical issues in the validation process are discussed.
药物产品的发现、研究和开发依赖于用于评估体内和体外产品特性及药物效应的经过验证的分析方法。开发经过验证的分析方法是一项多方面的活动，给生物分析化学家及统计学家带来了诸多有趣的挑战。本文回顾了分析稀释法基本有效性的相似性条件，将其作为药物应用分析方法验证的一个基本概念。从验证四个阶段所评估的特性方面考虑了分析方法有效性与可接受性之间的区别。从统计学角度对美国食品药品监督管理局发布的分析程序验证最新指南进行了评估，并讨论了验证过程中的统计学问题。
