中文摘要：

目的：使用Bland-Altman分析法,评估三种化学发光检测系统测定血清糖链抗原125（cancer antigen,CA125）结果之间的一致性。方法：选取1 0 0例患者血清,分别使用Roche、Beckman和Siemens公司的检测系统进行CA125检测,对每两种检测系统之间的检测结果进行相关性分析与Bland-Altman分析,并绘制一致性限度作为评价一致性的指标。结果：Roche、Beckman和Siemens检测系统检测CA125结果范围分别为5.6~279.8、4.5~542.4和4.5~350.0 U/m L;不同检测系统间相关系数如下：Roche与Beckman为0.932（P=0.000）,Roche与Siemens为0.947（P=0.000）,Beckman与Siemens为0.9 4 8（P=0.0 0 0）,每两组之间均存在相关;Bland-A ltman一致性分析结果如下：R o c h e与Beckman、Roche与Siemens和Beckman与Siemens界限内的最大差值与均值的比值分别为46.6%（94.19/201.995）、52.85%（31.98/60.51）和42.02%（119.17/283.59）,均超出了CLIA’88规定的允许误差范围的±25%,三种方法不具有一致性。结论：三种CA125检测系统之间数据不具有一致性,建议对CA125建立标准化的参考系统,现有的检测系统应与标准化的参考系统进行比对。

英文摘要：

Objective： To assess the agreement of three different chemiluminescent detection system of CA125 measurement by using Bland-Altman analysis. Methods： One hundreds serum samples were obtained from 100 patients. CA125 of these samples were measured by three different chemiluminescent detection system, including Roche, Beckman and Siemens. Bland-Airman analysis and Spearman correlation analysis were performed to assess the results of every two systems, and draw the limits of agreement as an evaluation of agreement of different systems. Results： The distribution of CA125 results of Roche, Beckman and Siemens system were 5.6-279.8, 4.5-542.4 and 4.5-350.0 U/mL, respectively. Spearman correlation analysis showed that the results of every two methods had good correlation, the correlation coefficient were 0.932 （P=0.000） between Roche and Beckman, 0.947 （P=0.000） between Roche and Siemens, 0.948 （P=0.000） between Beckman and Siemens. Bland-Altman analysis for the measurements showed bad relativity. The Bland-Altman plot revealed lower agreement between every two methods for CA12S results. Conclusion： Three detection systems do not have the consistency of the CA125 results, it is recommended to establish a standardized reference system of CA125 for comparison.