Petersen Elijah J, Elliott John T, Gordon John, Kleinstreuer Nicole C, Reinke Emily, Roesslein Mattias, Toman Blaza
National Institute of Standards and Technology (NIST), Gaithersburg, MD, USA.
US Consumer Product Safety Commission, Rockville, MD, USA.
ALTEX. 2023;40(1):174-186. doi: 10.14573/altex.2205081. Epub 2022 Jul 15.
New approach methodologies (NAMs) are in vitro, in chemico, and in silico or computational approaches that can potentially be used to reduce animal testing. For NAMs that require laboratory experiments, it is critical that they provide consistent and reliable results. While guidance has been provided on improving the reproducibility of NAMs that require laboratory experiments, there is not yet an overarching technical framework that details how to add measurement quality features into a protocol. In this manuscript, we discuss such a framework and provide a step-by-step process describing how to refine a protocol using basic quality tools. The steps in this framework include 1) conceptual analysis of sources of technical variability in the assay, 2) within-laboratory evaluation of assay performance, 3) statistical data analysis, and 4) determination of method transferability (if needed). While each of these steps has discrete components, they are all inter-related, and insights from any step can influence the others. Following the steps in this framework can help reveal the advantages and limitations of different choices during the design of an assay such as which in-process control measurements to include and how many replicates to use for each control measurement and for each test substance. Overall, the use of this technical framework can support optimizing NAM reproducibility, thereby supporting meeting research and regulatory needs.
新方法学（NAMs）是体外、化学和计算机模拟或计算方法，有可能用于减少动物试验。对于需要实验室实验的新方法学而言，确保其提供一致且可靠的结果至关重要。虽然已提供了关于提高需要实验室实验的新方法学可重复性的指南，但尚未有一个总体技术框架详细说明如何在方案中加入测量质量特征。在本手稿中，我们讨论了这样一个框架，并提供了一个逐步流程，描述如何使用基本质量工具完善方案。该框架中的步骤包括：1）对测定中技术变异性来源的概念分析；2）实验室内部对测定性能的评估；3）统计数据分析；4）确定方法的可转移性（如有需要）。虽然这些步骤中的每一个都有不同的组成部分，但它们都是相互关联的，任何一个步骤的见解都可能影响其他步骤。遵循该框架中的步骤有助于揭示在测定设计过程中不同选择的优缺点，例如应包括哪些过程控制测量以及每个控制测量和每种测试物质应使用多少重复。总体而言，使用此技术框架可支持优化新方法学的可重复性，从而满足研究和监管需求。
