血清淀粉酶分析中不确定度评定方法的比较与选择

The Comparison and Selection of Evaluation Methods of Uncertainty in Serum Amylase Analysis

目的比较血清淀粉酶(AMY)分析中不确定度的评定方法,根据现有数据来源和实际需求选择合适的评定方法。方法参考CNAS-GL037文件中的方法测量2个浓度水平冰冻人血清AMY国家标准物质各5批、每批3次;采用JCTLM参考方法和常规方法同时测量40份浓度为(41.2~1237.0)U/L单人份血清标本各2次;收集2015至2020年我院卫健委正确度验证计划数据;测量2个浓度水平人血清标本各5批、每批5次;收集我院半年内同一批号2个浓度水平室内质控在控数据。参考CNAS-TRL-001:2012文件中的方法,采用以上来源的数据分别计算由测量偏倚引入的不确定度分量[u_(crel)(bias)]和由实验室内测量复现性引入的不确定度分量[u_(rel)(R_(w))]。分别将不同来源数据计算的u_(crel)(bias)和u_(rel)(R_(w))两两组合计算相对合成标准不确定(u_(crel))。比较结果差异,分析实验室如何选择合适的评定方法。结果采用测量标准物质的数据、与参考方法比较的数据、正确度验证计划数据计算低高浓度标本的u_(crel)(bias)分别为2.10%/0.96%、1.93%/3.09%、2.00%/3.49%;采用实验室精密度验证数据、室内质控数据、正确度验证计划数据分别计算低高浓度标本的u_(rel)(R_(w))分别为1.27%/0.92%、0.98%/0.68%、1.18%/0.81%;采用不同来源数据计算的u_(crel)(bias)和u_(rel)(R_(w))两两组合计算的u_(crel)的均值和波动范围分别是2.31%/(2.16~2.45)%;2.68%/(1.18~3.61)%。结论采用不同来源数据计算的u_(crel)均符合行业规定的1/2 TEa(±7.5%)要求,其中低浓度标本的u_(crel)差异不大,但高浓度标本的u_(crel)有差异。实验室在初建AMY检测系统时可采用正确度和精密度验证数据评价所选系统的测量不确定度;检测系统使用一段时间后可以采用室间质评和室内质控的数据计算,更符合实验室的实际情况。

Objective To compare the evaluation methods of uncertainty in the analysis of serum amylase(AMY)and to select the appropriate evaluation method according to the data sources and actual needs.Methods According to CNAS-GL037 CNAS-TRL-001:2012 document,5 batches of frozen human serum AMY national reference materials of 2 concentration levels were measured,with 3 times for each batch;40 single serum samples with concentration of(41.2-1237)U/L were measured by JCTLM reference method and routine method,with twice every sample;The trueness verification plan data of National Health Commission from 2015 to 2020 was collected;5 batches of human serum samples of 2 concentration levels were measured,with 5 times for each batch;The internal quality control data of 2 concentration levels of the same batch in our hospital within half a year were collected.The uncertainty component(u_(crel)(bias)u_(rel)(R_(w)))introduced by measurement bias and reproducibility in the laboratory were calculated respectively using the data from the above sources.The relative combined standard uncertainty(u_(crel))was calculated by combining the uncertainty component from different sources in a two-by-two manner.We evaluated the differences between the results,and analyzed how laboratory could select the appropriate evaluation method.Results According to the measurement data of reference materials,the one compared with reference methods and the trueness verification plan,u_(crel)(bias)of low and high concentration samples were 2.10%/0.96%,1.93%/3.09%and 2.00%/3.49%,respectively.According to the laboratory precision verification data,the data from IQC and trueness verification plan,u_(crel)(R_(w))of low and high concentration samples were 1.27%/0.92%,0.98%/0.68%,1.18%/0.81%respectively.The average value and fluctuation range of u_(crel) calculated by data from different sources were 2.31%/(2.16-2.45)%,and 2.68%/(1.18-3.61)%.Conclusion u_(crel) calculated from different sources can meet the requirements of 1/2 TEa(±7.5%).The difference of low concentration samples is not significant,but the difference of high concentration samples is significant.When the laboratory builds AMY measurement system initially,the trueness and precision verification data can be used to evaluate the measurement uncertainty.After using the measurement system for a period of time,the data of internal quality control and external quality assessment can be used to evaluate the measurement uncertainty,which can fit better the actual needs of the laboratory.