Comparison of three methods for determination of glucose.

Study refers comparison of three methods for glucose determination - precision (repeatability, reproducibility), traceability to SRM 965a NIST, comparability in blood-pools and in patients' samples: Electrochemical determination on Super GL (DiaSys, Germany) in hemolyzate - GL method, spectrophotometric determination using hexokinase (Glucose System Reagent 800, Olympus) - HKL method - and using glucose dehydrogenase (Glucose Gluc-DH, EcolineS+, DiaSys, Germany) - GDL method - in hemolyzate. For showing differences between the concentration of glucose in hemolyzed blood and corresponding plasma, spectrophotometric determination using hexokinase in plasma was used (Glucose System Reagent 800, Olympus) - HKP method. Coefficients of variation characterizing precision under repeatability and reproducibility conditions are not higher than 3.0% for the GL method, 6.3% for the GDL method and 15.8% for the HKL method with low sensitivity. For glucose concentration less than 8 mmol/l, HKL tends to give lower results than GDL, and GL tends to give higher results than GDL. For glucose concentration about 2 mmol/l, the results of glucose in plasma - HKP method - tend to be significantly lower (by more than ten percent) than in corresponding total (hemolyzed) blood. HKL method can be reasonably used in a high number of parallel determinations. For glucose 8 mmol/l and lower, comparability of results given by HKL, GDL and GL methods gradually worsens, while for glucose between 8 and 34 mmol/l results of the three mentioned methods are well comparable.