Suppr超能文献

方法评估中建立与验证分析性主张的概念。

The concept of establishing vs verifying analytic claims in method evaluations.

作者信息

Hartmann A E

机构信息

Department of Pathology, McKennan Hospital, Sioux Falls, SD 57117-5045.

出版信息

Arch Pathol Lab Med. 1992 Jul;116(7):714-7.

PMID:1497443
Abstract

The establishment of analytic claims occurs during the method development stage. For commercial methods, the thorough characterization of analytic performance is the responsibility of the manufacturer. The manufacturer establishes performance claims with a high degree of certainty by utilizing statistically sound experimental protocols with sufficient degrees of freedom. The manufacturer should establish performance claims that state the expected performance of the method for the intended spectrum of patients and laboratory environment. The verification of analytic claims occurs by the end-user of the method. The end-user should not have to necessarily repeat the extensive protocol testing of the manufacturer that was used to establish the performance claims. The goals of the method verification study are not to establish analytic performance or to determine if the analytic performance is clinically acceptable, but, simply, to confirm or verify the stated established performance claims of the manufacturer. This implies an abbreviated method evaluation with statistical confirmation that analytic performance is acceptable. Analytic parameters that should be verified are imprecision, inaccuracy (bias), and linearity including analytic sensitivity. Reference ranges should be determined if the patient population is unique, or if the reference range of the analyte is relatively new or unstudied. Principles of method evaluation are discussed.

摘要

分析性能声明的确立发生在方法开发阶段。对于商业方法,分析性能的全面表征是制造商的责任。制造商通过使用具有足够自由度的统计学合理实验方案,高度确定地确立性能声明。制造商应确立性能声明,说明该方法在预期患者群体和实验室环境中的预期性能。分析性能声明的验证由该方法的最终用户进行。最终用户不必重复制造商用于确立性能声明的广泛方案测试。方法验证研究的目标不是确立分析性能或确定分析性能在临床上是否可接受,而仅仅是确认或验证制造商所述的既定性能声明。这意味着进行简短的方法评估并通过统计确认分析性能是可接受的。应验证的分析参数包括不精密度、不准确(偏差)以及包括分析灵敏度在内的线性。如果患者群体独特,或者分析物的参考范围相对较新或未被研究过,则应确定参考范围。文中讨论了方法评估的原则。

文献检索

告别复杂PubMed语法,用中文像聊天一样搜索,搜遍4000万医学文献。AI智能推荐,让科研检索更轻松。

立即免费搜索

文件翻译

保留排版,准确专业,支持PDF/Word/PPT等文件格式,支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述,25分钟生成高质量综述,智能提取关键信息,辅助科研写作。

立即免费体验