目的基于Nordtest准则,结合不同来源的数据评定测量不确定度,评价其在医学实验室的适用性。方法:以心肌标志物为例,方案1:收集本实验室2015年8月到2016年1月的室内质控数据与2013-2015年共6次参加卫生部临检中心(NCCL)心肌标志物室间质评的结果;方案2:收集本实验室2015年8月到2016年1月的室内质控数据与BIO-RAD室内全球数据报告(Unity TM Interlab Report Woldwide Report)。按测量不确定度评定过程,分析不确定度分量的来源,并按这两种数据来源制定方案进行评定,根据这两种方案合成标准不确定度,计算扩展不确定度。结果:低浓度,方案1中CKMB,MYO,TNI相对扩展不确定度为14.02%,18.78%,20.54%,方案2中CKMB,MYO,TNI相对扩展不确定度为11.74%,12.15%,16.06%;高浓度,方案1中CKMB,MYO,TNI相对扩展不确定度为13.78%,16.39%,15.69%,方案2中CKMB,MYO,TNI相对扩展不确定度为10.87%,14.15%,15.91%。结论:两种评定测量不确定度的方法均适用于医学实验室,但方案2利用伯乐全球数据计算偏倚相对于室间质评计算偏倚的方法来说,用时更短,更方便,实用性更广。 OBJECTIVE Based on the Nordtest criterion and the data from different sources, the measurement uncertainty is evaluated and its applicability in medical laboratories is evaluated. Methods: Take myocardial markers as an example. Option 1: Collect indoor quality control data from August 2015 to January 2016 in our laboratory and 6 myocardial markers from the Ministry of Health Checkpoint (NCCL) during 2013-2015 Scenario 2: Indoor laboratory data from August 2015 to January 2016 was collected together with the BIO-RAD Indoor Global Data Report (Unity TM Interlab Report Woldwide Report). According to the measurement uncertainty evaluation process, the source of the uncertainty component is analyzed, and the two data sources are used to make a plan for evaluation. According to the two schemes, the standard uncertainty is calculated and the expansion uncertainty is calculated. The results showed that the relative expanded uncertainty of CKMB, MYO and TNI in program 1 was 14.02%, 18.78% and 20.54%, respectively. The relative expanded uncertainty of CKMB, MYO and TNI in program 2 were 11.74%, 12.15% and 16.06% The relative expansion of CKMB, MYO and TNI were 13.78%, 16.39% and 15.69%, respectively. The relative expansion of CKMB, MYO and TNI were 10.87%, 14.15% and 15.91% respectively. CONCLUSIONS: Both methods of assessing measurement uncertainty are suitable for use in medical laboratories, but scenario 2 uses Bole Global Data to Compute Bias Relative to Room Bias Assessment Bias, which takes less time, is more convenient, and more practical wide.