Résumé
Objective:The allowable total error ( TEa),allowable imprecision ( CV)and allowable bias( Bias)were recommended for 34 routine chemistry analytes in China. Methods:According to the performance specification setting mode newly determined at the Milan conference in Italy,the performance specification was derived based on components biological variation (BV)and current state of the art mode. The data(including EQA data and IQC data)of laboratories participating in the routine chemistry and lipids and lipoproteins EQA activities of the national center for clinical laboratories from 2019 to 2021 was collected through clinet-EQA. For the analytes with biological variation(BV)data,compared the'percentage difference′ of EQA data and the'in-control coefficient of variation of the month′ of IQC data of each research analyte with the three levels evaluation criteria derived based on BV,and calculated the percentage difference passing rate and CV passing rate of all batches in each year. When the passing rate reaches 80%,the performance specifications of this level met the requirements of the recommended performance specifications of the analyte. For the analytes without BV data or analytes whose performance specifications at three levels derived based on BV could not be used as recommended standards,the recommended performance specifications are derived based on the current state of the art. After obtaining the recommended TEa and allowable CV for each analyte,used the formula | Bias|≤ TEa-z? CV to derive the recommended allowable bias. Results:The results of TEa ( CV)% recommended by 34 analytes are as follows:K4.7(2),Na4(1.5),Cl4(1.4),Ca5(2),P9.6(3.9),Glu6.4(2.5),Urea8(3),UA12(4.1),Cre11(3.3),TP5(2),Alb5.2(2.4),TC8.6(2.7),TG13.5(5),HDL-C16.5(4.3),LDL-C20.5(6.2),ApoAⅠ16(5.3),ApoB 17.1(5.5),Lp(a) 24.1(10.4),TBil 12.4(5),DBil 20(7.3),ALT16(5),AST13.5(4.8),ALP17.5(4.8),AMY13.1(3.3),CK11.3(3.8),LDH11(3.9),CHE13.4(5.3),LIP20(6.9),Fe13.3(5.2),Mg14(4.5),Cu17.9(6.8),Zn15.1(6.4),γ-GGT10(3.3),α-HBDH18(5.8).The formula | Bias|≤ TEa-z? CV is used to derive the allowable bias of 34 analytes. Conclusions:For 34 clinical routine chemistry quantitative analytes,the allowable total error,allowable imprecision and allowable bias that meet the current state of the art of Chinese laboratories are recommended.