Abstract

For more than one half-century, variability observed in clinical test result measurements has been ascribed to three major independent factors: (i) pre-analytical variation, occurring at sample collection and processing steps; (ii) analytical variation of the test method for which measurements are taken, and; (iii) biological variation (BV). Appreciation of this last source of variability is the major goal of this review article. Several recent advances have been made to generate, collate, and utilize BV data of biomarker tests within the clinical laboratory setting. Consideration of both prospective and retrospective study designs will be addressed. The prospective/direct study design will be described in accordance with recent recommendations discussed in the framework of a newly-developed system of checklist items. Potential value of retrospective/indirect study design, modeled on data mining from cohort studies or pathology laboratory information systems (LIS), offers an alternative approach to obtain BV estimates for clinical biomarkers. Moreover, updates to BV databases have made these data more current and widely accessible. Principal aims of this review are to provide the clinical laboratory scientist with a historical framework of BV concepts, to highlight useful applications of BV data within the clinical laboratory environment, and to discuss key terms and concepts related to statistical treatment of BV data.

摘要

半个多世纪以来，临床测试结果测量中观察到的变异性归因于三个主要的独立因素：(i) 分析前变异，发生在样本收集和处理步骤；(ii) 进行测量的测试方法的分析变化，以及；(iii) 生物变异 (BV)。了解这最后一个可变性来源是这篇评论文章的主要目标。最近在临床实验室环境中生成、整理和利用生物标志物测试的 BV 数据方面取得了一些进展。将考虑前瞻性和回顾性研究设计。前瞻性/直接研究设计将根据最近在新开发的清单项目系统框架内讨论的建议进行描述。以队列研究或病理实验室信息系统 (LIS) 的数据挖掘为模型的回顾性/间接研究设计的潜在价值提供了一种替代方法来获得临床生物标志物的 BV 估计值。此外，BV 数据库的更新使这些数据更加及时和广泛访问。本综述的主要目的是为临床实验室科学家提供 BV 概念的历史框架，突出 BV 数据在临床实验室环境中的有用应用，并讨论与 BV 数据的统计处理相关的关键术语和概念。BV 数据库的更新使这些数据更具时效性和更广泛的可访问性。本综述的主要目的是为临床实验室科学家提供 BV 概念的历史框架，突出 BV 数据在临床实验室环境中的有用应用，并讨论与 BV 数据的统计处理相关的关键术语和概念。BV 数据库的更新使这些数据更具时效性和更广泛的可访问性。本综述的主要目的是为临床实验室科学家提供 BV 概念的历史框架，突出 BV 数据在临床实验室环境中的有用应用，并讨论与 BV 数据的统计处理相关的关键术语和概念。