诊断试验一致性评价中几种方法的比较及应用

Evaluation and Application on Different Assessment Methods of Consistency of Diagnostic Test

目的对诊断试验中评价一致性的几种方法进行比较。方法对实例及3种模拟情况进行一致性评价,比较几种方法的优劣。结果检验试剂与参考试剂测量游离前列腺特异抗原(FPSA)的实例分析中,除配对t检验外,几种方法均显示出良好的一致性;在系统误差较大的模拟情况下,简单相关分析所得结果相对实例而言没有变化;在随机误差较大的模拟情况下,配对t检验得出两种试剂具有良好的一致性;在系统误差和随机误差均小的模拟情况下,组内相关系数所得结果显示一致性较差。结论配对t检验、简单相关分析用于一致性评价都只适用于部分资料,组内相关系数法适用于存在系统误差和/或随机误差的时候,而Bland-Altman法以及ATE/LER区域可作为一致性评价的优先考虑方法。

Objective To evaluate different assessment methods of consistency of diagnostic test. Methods Different assessment methods were adopted to evaluate the consistency of a true example with good consistency in fact and three simulation situations developed from the true one. Results All assessment methods showed good consistency for the true example except paired t test. Simple correlation analysis was insensitive in the situation of obvious systematic bias. And in the situation with comparatively large random error where low consistency was accepted, paired t test suggested better consistency. In the situation with low range of measurement where both systematic bias and random error were small, in- tra-class correlation analysis failed demonstrate exact results. Conclusion Both paired t test and simple correlation analysis have obvious defects in assessing consistency. Although intra-class correlation analysis has some limitation,it take account of systematic bias and random bias. And Bland- Altman methods and ATE/LER zones are recommended to evaluate the consistency.