Horwitz方程作为方法验证接受标准的局限性探讨

Discussion on the limitations of Horwitz equation as an acceptance criteria for method validation

目的:探讨使用Horwitz方程判断分析方法是否满足其预期用途的局限之处。方法:在介绍Horwitz方程内容的基础上,通过查找收集大量国内外相关文献,从不同方面分析Horwitz方程作为方法验证接受标准的不足之处。结果:(1)Horwitz方程预测值未包含方法所有的变异来源;(2)Horwitz方程不允许用不确定度来评估方法性能;(3)Horwitz方程未与产品限度范围相联系。结论:使用Horwitz方程作为判断标准是不够合理的,应基于分析方法的总变异进行判定。文中介绍了方法的预测区间或容忍区间,以及方法能力指数以用于判断方法是否满足预期用途。

Objective:To discuss the limitations of using the predicted value of Horwitz equation to judge whether the analysis method meets the intended use.Methods:On the basis of introducing the content of the Horwitz equation,the limitations of using the Horwitz equation as accepetance criteria of method validation were analyzed from different points by searching and collecting a large number of domestic and foreign relevant literatures.Results:(1)The predicted value of Horwitz equation did not contain all the method variables;(2)The Horwitz equation did not allow using the uncertainty to evaluate the method performance;(3)The Horwitz equation was not related to product’s limits or ranges.Conclusion:Using Horwitz equation as the judging criteria is not reasonable,it should be depended on the total variability of method.The predictive interval,tolerance interval and method capability index were introduced to determine if the method meets the intended use.