Evaluation of the precision and accuracy of the results from the 6 systems using 4 serum glucose analysis methods

LIU Ou,XU Guo-bin,WANG Lan-zhen,XIA Chang-yu,LIU Yan

DOI: https://doi.org/10.13602/j.cnki.jcls.2010.03.015

2010-01-01

Abstract:Objective To evaluate the precision and accuracy of the results from six systems using four serum glucose analysis methods.Methods The glucose value of five frozen human serum pools was determined by the glucose hexokinase reference method which was recommended by the CDC in the USA.The accuracy of the reference laboratories was evaluated by the SRM965a and further validated through attending inter-laboratory comparison organized by IFCC(RELA).Each routine analysis system(n=3) used in the matched calibrators was calibrated by the procedure according to the manufacturer recommended.Bias between the values from routine systems and the target value was used to evaluate the accuracy of routine system.The imprecision of six routine systems was expressed by coefficient variation.Results Two candidate reference laboratories measured four levels NIST′s SRM965a.Eight groups′ results were obtained.Average deviations of four groups were from 2.4%~2.8%,while four other groups′ average deviations were less than 2%.RELA comparison showed that results of the two laboratories were similar to the results from other reference laboratories.Coefficient variation of routine systems was from 1.81% to 2.89%.Relative bias was as follows:-2.47%~-0.32%,1.69%~4.5%,-2.27%~0.27%,3.55%~4.97%and-0.97%~0.32%,respectively.The relative total error of five levels in each routine analysis system was less than 1/2 of CLIA′88 glucose which defined maximum acceptable range(±10%),and also less than of the glucose biological variation total error(±6.9%).Conclusions Two candidate reference laboratories′ results are accuracy.In this study,six systems using four different serum glucose analysis methods have good precision and accuracy.