目的

评价一种运用CLSI EP15-A2文件简便验证生化分析系统测定尿素、肌酐、葡萄糖及尿酸的精密度和正确度性能方法的应用价值，为提高临床实验室质量管理水平提供方法。

方法

采用依据CLSI EP15-A2文件评价实验室生化分析系统性能的实验设计。以高低2个浓度的质控血清C1、C2各5份作为精密度验证样本，采用常规生化系统每天对2011年全国常规化学室间质评第三次活动质评物L1、L2各1份进行批内重复测定3次，共测5 d，计算测定尿素、肌酐、葡萄糖及尿酸的重复性变异系数（coefficient of variation, CV）和实验室内CV，并与厂家提供的指标比较，如实验数值＜厂家数值或以此计算出验证值，则验证通过。用经同位素稀释质谱法（ID-MS）定值的新鲜冰冻血清一级标准物质GBW 09175（FF1）、新鲜冰冻血清（FF2）各5份作为正确度验证样本，在精密度样本测定的同时，每天各取1份FF1、FF2在同一实验批次内重复测定2次，经统计分析4个项目平均值与定值是否等效，等效则表示正确度验证通过。对于不等效的结果，如平均值与定值的偏倚绝对值分别不大于3.0%、5.5%、2.0%和4.5%，则正确度验证仍可以通过。

结果

测定尿素、肌酐、葡萄糖和尿酸4个项目的L1重复性CV分别为0.61%、0.56%、0.64%、0.61%，实验室内CV分别为1.02%、0.66%、0.69%和0.68%，测定L2的重复性CV分别是0.65%、0.29%、0.77%、0.82%，实验室内CV分别为0.97%、0.40%、1.26%和0.86%；与同厂家的对应指标（1.63%、1.36%、0.76%、0.99%、2.86%、1.47%、0.80%、1.44%；0.97%、0.73%、1.20%、0.92%；1.64%、0.77%、1.21%、1.0%）相比，精密度验证通过。正确度验证中，测定FF1时，尿素、肌酐、葡萄糖和尿酸的均值分别为9.8 mmol/L、110.8 μmol/L、6.3 mmol/L和489.6 μmol/L，与参考方法定值比较，偏倚分别为2.3%、-2.7%、0.2%和-1.1%；测定FF2时，尿素、肌酐、葡萄糖和尿酸的均值分别为3.5 mmol/L、348.9 μmol/L、11.1 mmol/L和597.2 μmol/L，偏倚分别是3.2%、0.7%、-2.5%和2.3%；除FF1的尿素外，其余项目参考方法定值皆在均值的验证区间内，通过了正确度验证。FF1尿素测定的偏倚2.3%，符合相关行业标准的要求，被认为通过正确度验证。

结论

运用CLSI EP15-A2文件进行的尿素、肌酐、葡萄糖及尿酸项目精密度及正确度性能验证均获通过，EP15-A2提供了一种简便的精密度和正确度验证方案，其能满足临床实验室对分析系统性能验证的需求。

Objective

To assess the effect of protocol focusing on verification of performance of precision and trueness of a biochemistry analysis system on determination of urea, creatinine, glucose and uric acid based on the approach provided by CLSI EP15-A2, and introduce a simple and practical method to improve management quality for clinical laboratories.

Methods

The CLSI EP15-A2 was announced for verification protocol design. Five aliquots of two control materials C1 and C2 were used as precision verification samples.Each aliquot was measured in triplicate in a run every day during 5 days. The repeatability and within-laboratory coefficients of variation (CV) of urea, creatinine, glucose and uric acid determination were estimated and compared with manufacturer′s claim of CVs. If the estimated CVs were less than the claimed CVs, the claim will be verifiable.Five aliquots of two level frozen fresh sera FF1 and FF2 (values assigned by isotope dilution mass spectrometry)were applied as trueness verification samples, each aliquot was measured in duplicate simultaneously with the precision tests. Trueness verification was carried out by whether it was of equivalence statistically between the average of routine biochemistry analysis system and the target value. If the test result was not equivalent to the target value, the bias was further analyzed.

Results

The CVs of urea, creatinine, glucose and uric acid on L1 were 0.61%, 0.56%, 0.64% and 0.61% respectively and the within-laboratory CVs were 1.02%, 0.66%, 0.69% and 0.68%.The CVs of the 4 analytes on L2 were 0.65%, 0.29%, 0.77% and 0.82% respectively, and the within-laboratory CVs were 0.97%, 0.40%, 1.26% and 0.86% respectively. Compared with the specification claimed by manufactory(1.63%, 1.36%, 0.76%, 0.99%, 2.86%, 1.47%, 0.80%, 1.44%, 0.97%, 0.73%, 1.20%, 0.92%, 1.64%, 0.77%, 1.21%,1.0% respectively), the precision of these 4 tests was verifiable. The results of urea, creatinine, glucose and uric acid averaged 9.8 mmol/L, 110.8 μmol/L, 6.3 mmol/L and 489.6 μmol/L on C1, the biases were 2.3%, -2.7%, 0.2% and-1.1% respectively compared with the values assigned by ID-MS methods. The results of urea, creatinine,glucose and uric acid averaged 3.5 mmol/L, 348.9 μmol/L, 11.1 mmol/L and 597.2 μmol/L on C2, and the biases were 3.2%, 0.7%, -2.5% and 2.3% respectively. All the results are of equivalence with target values but urea on FF1. The bias of urea on FF1 (2.3%) met the specification listed in a criterion issued by the minstry of health in 2012.

Conclusion

The verification of performance for precision and trueness of 4 analytes are passed,the CLSI EP15-A2 provides a simple protocol for verification of performance for precision and trueness, and should be applied for performance verification of analysis systems in clinical laboratories.