定量免疫测定不确定度评定方法探讨

Discussion on evaluation method of uncertainty in quantitative immunoassay

目的通过对肿瘤标志物五项(CA125、CA153、CA199、AFP、CEA)的不确定度进行评定,探讨定量免疫测定不确定度的评定方法。方法采用自上而下(up-down)的方法评定测量不确定度。收集本实验室CA125、CA153、CA199、AFP、CEA五项肿瘤标志物12个月室内质控数据计算精密度用于评价A类不确定度,分别采用校准品的不确定度和卫生部临床检验中心(NCCL)肿瘤标志物能力比对(PT)数据评定B类不确定度,由A类和B类不确定度合成测量不确定度。结果用室内质控不精密度和校准品不确定度评定CA125、CA153、CA199、AFP、CEA五项肿瘤标志物的最大扩展不确定度分别为10.3%,9.1%,13.0%,13.1%,13.7%;用室内质控不精密度和NCCL室间质评数据评定CA125、CA153、CA199、AFP、CEA五项肿瘤标志物的最大扩展不确定度分别为16.02%,14.24%,15.14%,14.88%,20.40%。结论在相同室内质控不精密度的情况下,采用校准品不确定度评定B类不确定度合成的不确定度较用室间能力比对数据评定的B类合成的不确定度小,但评定结果均达到质量目标要求。采用长期室内质控不精密度和校准品的不确定度来评定定量免疫测定项目的不确定度更为简便。

Objective The uncertainty of five tumor markers(CA125,CA153,CA199,AFP and CEA)was evaluated to explore the evaluation method of the uncertainty of quantitative immunoassay. Methods The up-down method was used to evaluate the measurement uncertainty. The internal quality control(IQC)data from 12 months of tumor markers CA125、CA153、CA199、AFP、CEA were collected and were used to calculate the imprecision for evaluated the A group measurement uncertainty,the B group measurement uncertainty was evaluate by using the measurement uncertainty of the tumor markers calibration and the proficiency testing(PT)data of tumor markers from NCCL,then the measurement uncertainty was combined with A group and B group measurement uncertainty. Results The maximal expanded uncertainty of the five tumor markers were evaluated by the imprecision from IQC and uncertainty of calibration were10.3%,9.1%,13.0%,13.1%,13.7% respectively,and the maximal expanded uncertainty of the five tumor markers were evaluated by the imprecision from IQC and PT data were 14.88%、14.24%、15.14%、14.88%、20.4%respectively. Conclusion In the same imprecision of IQC,the uncertainty combined by B group uncertainty that was evaluated by uncertainty of the tumor calibration was less than combined by B group uncertainty that was evaluated by PT data,but both results were meet the quality requirements. To using the long-term imprecision of IQC and uncertainty of calibration to evaluate measurement uncertainty is more simple.