Reproduction of reference method for serum glucose detection and comparison with routine glucose kits / 中华检验医学杂志

ABSTRACT

Objective To reproduce and validate the hexokinase reference method for glucose detection and compare other 5 routine glucose kits with this reference method. Methods The CDC hexokinase reference method for glucose detection was established and the performance was validated through testing a standard reference material (SRM) and participating in the IFCC ring-trial for reference laboratories for glucose evaluation. The CLSI EP 9-A2 protocol was used to compare the 5 routine glucose kits with the hexokinase reference method. Forty serum samples were analyzed by 5 routine kits and the hexokinase reference method. Results When SRM 965a was determined by the reference method,the bias of level 2 and level 3 were 0. 93%, -0. 23% respectively. The results for IFCC ring trial were within the accepted range. For the 5 routine kits, the confidence intervals of the predictive bias at the medical decision point Xc (Xc = 6. 11 mmol/L) were all within the range of defined acceptable error (10%) and the range of biological variation bias (6.9%). Conclusions The hexokinase reference method for serum glucose was reproduced in our lab. The serum glucose results measured with 5 routine kits were different from results detected with the reference method, but the bias was acceptable,and it will not affect the detection results.