六西格玛质量标准在酶学参考测量室间比对中的应用

摘要

Objective To illustrate the application of six sigma metrics in Enzymes External Quality Assess-ment for Reference Laboratories (REEQA) .Methods σvalues and quality goal index (QGI) of different tests meas-ured by 2013 REEQA participating laboratories were calculated .Distributions of σvalues of different tests were com-pared .Quality of enzymology reference measurement and the state of performance were evaluated .Quality control (QC) rules for different tests were chosen in accordance with σvalues and rationality of existing practical QC proce-dures was assessed .Expected total error of Ring Trials for Reference Laboratories (TERELA ) and expected coefficient of variation within laboratory (CVE) of enzymology reference measurement were calculated .Whether TERELA meeting the demand of traceability of routine methods or not was analyzed by comparing TERELA with two kinds of total error induced from biology variation .Feasibility of CVE according to CVE coincident rate of actual coefficient of variation (CV) was assessed .Results When 1/2TEBO and 1/2TEWS/T were taken as tolerance limit ,sigma(σ) values of differ-ent tests were alanine aminotransferas (ALT ,0 .0 -16 .2 ,0 .1 -21 .0) ,aspartate aminotransferase(AST ,0 .0 -9 .9 ,3 .9-20 .2) ,γ-glutamyltransferase(GGT ,0 .0 -26 .6 ,0 .4 -26 .3) ,lactate dehydrogenase(LDH ,0 .0 -18 .9 ,0 .8 -37 .5) ,creatine kinase(CK ,3 .8 -34 .5 ,3 .7 -34 .1) ,amylase(AMY ,0 .0 -10 .1 ,3 .3 -25 .2) ,alkaline phosphatase (ALP ,0 .0-7 .9 ,3 .1-42 .9) respectively .When being judged by 1/2TEWS/T ,the proportions of σ>5 of AST ,CK , AMY and ALP were more than 80 .0% and no case of σ< 3 was found ,the proportion of σ> 4 of ALT exceeded 80 .0% and that of σ<3 was 10 .3% ,and the proportions of σ>4 of GGT and LDH were 77 .8% and 76 .9% respec-tively and those of σ<3 were 11 .1% and 17 .9% respectively .When being judged by 1/2TEBO ,only the proportion ofσ>5 of CK exceeded 80 .0% ,and the proportions ofσ>4 of remaining items were less than 80 .0% .The ratios ofσ>6 decreased from 56 .4% -83 .3 % to 13 .3% - 48 .7% and that of σ<3 increased from 0 .0% -17 .9% to 23 .1% -60 .0% .To select QC procedures based on the items′σvalue ,the QC rules of 13s ,n=2 and 12 .5s ,n=2 were appropri-ate to the conditions of σ≥6 and 5≤σ< 6 respectively .The existing practical QC procedures were equal to these rules .Under the condition of 4≤σ<5 ,the appropriate QC rules were 13s/22s ,n=4 .The performance of any items ofσ<3 must be reconfirmed .The ratios of favorable performance of all participating laboratories were 20 .0% -94 .4% .The ratios of imprecision were 0 .0% -46 .7% ,the proportions of inaccuracy were 0 .0% -33 .3% ,and the proportions of both were 0 .0% -16 .7% .The range of TERELA of 7 enzymes was 6 .3% -7 .1% .TERELA of ALT , AST ,AMY and ALP were between 1/2TEBO and 1/2TEWS/T ,which of GGT and LDH were higher than 1/2TEWS/T , which of CK was lower than 1/2TEBO .The CVEBO and CVEWS of 7 enzymes were 0 .5% -1 .3% and 0 .9% -1 .5% , the CVE coincident rate of participating laboratories′actual CV were 37 .8 % -97 .0% and 76 .9% -100 .0% .Con-clusion Through application of six sigma metrics ,REEQA could provide not only assessment of accuracy ,but also information about quality improvement to reference laboratories ,which could be an effective means to ensure perfect performance of enzymology reference measurements .%目的：采用六西格玛质量标准分析2013年全国酶学参考实验室室间比对（REEQA）数据，说明六西格玛质量标准在REEQA中的应用方式及效果。方法计算2013REEQA参评实验室各项目的西格玛（σ）值及质量目标指数，分析比较不同项目的σ值分布，评价酶学参考测量水平与实验室检测性能状况。参照各项目σ值选择室内质控程序并评价现行经验方法的合理性。计算国际参考实验室室间比对（REL A ）预期总误差（ T EREL A ）及酶学参考测量预期室内变异系数，将 TERELA与两种由生物学变异导出的总误差指标相比较，分析 TERELA能否满足常规检测溯源性需求，计算实验室实际室内变异对预期指标的符合率，判断预期指标的可行性。结果总误差为1/2最佳总误差（TEBO ）与1/2总误差指标（TEWS/T ）时，各项目的σ值范围分别为：丙氨酸氨基转移酶（ALT ）（0．0～16．2、0．1～21．0）、天门冬氨酸氨基转移酶（AST ）（0．0～9．9、3．9～20．2）、γ-谷氨酰转移酶（GGT ）（0．0～26．6、0．4～26．3）、乳酸脱氢酶（LDH）（0．0～18．9、0．8～37．5）、肌酸激酶（CK）（3．8～34．5、3．7～34．1）、α-淀粉酶（AMY）（0．0～10．1、3．3～25．2）、碱性磷酸酶（ALP）（0．0～7．9、3．1～42．9）。总误差为1/2TEWS/T时，AST、CK、AMY、ALP的σ＞5比例超过80．0％且没有σ＜3的情况，ALT 的σ＞4比例超过80．0％，σ＜3比例为10．3％，GGT 与LDH的σ＞4比例分别为77．8％与76．9％，σ＜3比例分别为11．1％与17．9％。总误差为1/2TEBO时，CK的σ＞5比例超过80．0％，其余项目σ＞4的比例均不足80．0％。除GGT、CK外其余项目的σ＞6比例从56．4％～83．3％下降至13．3％～48．7％，σ＜3比例从0．0％～17．9％上升至23．1％～60．0％。项目的σ≥6与5≤σ＜6时，分别采用13s ，n＝2与12．5s ，n＝2规则，经验性室内质控程序与之等效。4≤σ＜5时选用13s/22s ，n＝4规则。σ＜3时必须重新确认项目性能。全部参评实验室检测性能良好比例为20．0％～94．4％。各实验室精密度不佳比例为0．0％～46．7％，准确度不佳比例为0．0％～33．3％，二者均不佳比例为0．0％～16．7％。7种酶学项目的 T ERELA范围为6．3％～7．1％，其中ALT、AST、AMY、ALP的 TERELA介于1/2TEBO与1/2TEWS/T之间，GGT、LDH的 TERELA高于1/2TEWS/T ，CK的 TERELA则低于1/2TEBO 。7种酶学参考测量预期室内变异系数（CV）EBO及CVEWS范围分别为0．5％～1．3％与0．9％～1．5％，参评实验室实际室内变异系数与各预期指标相符比例分别为37．8％～97．0％与76．9％～100．0％。结论六西格玛质量标准在REEQA中的应用使该比对计划不仅限于评价的准确性，还可向实验室提供更多有关质量改进的信息，成为保证酶学参考测量长期稳定高质量运行的有效手段。