Comparison of blood glucose meters in a New Zealand diabetes centre

Abstract

Background To assess the accuracy and precision of the Roche Performa and Medisense Optium Xceed (5 and 10 s reading) blood glucose meters. Methods Capillary blood samples were taken from 100 patients attending a diabetes centre and blood glucose measured on Roche Performa ( n = 4) and Medisense Optium Xceed 5 s ( n = 2) and 10 s reading ( n = 2) meters. Venous plasma glucose from samples taken simultaneously was measured by the laboratory hexokinase method as reference standard. Imprecision was determined on the meters by replicate analysis ( n = 20) of control solutions provided by the manufacturers and also patient venous whole-blood samples. Accuracy was assessed relative to the reference method by Bland–Altman plots, Passing and Bablok regression analysis, and both Clarke and consensus error grid analysis. Coefficients of variation (CVs) were calculated to determine imprecision. Results Bland–Altman and Passing–Bablok analysis confirmed significant systematic bias for all meters, with relative under-reading of higher glucose concentrations. Error grid analysis showed that <5% readings exceeded ±20% (or ±0.83 mmol/L for readings <4 mmol/L) deviation from the reference method (1%, 2% and 4% for the Roche, Optium 5 and 10 s meters, respectively). CVs were all <4% for the control solutions and <6% for patient samples. Conclusions Both Roche Performa and Medisense Optium glucose meters (5 and 10 s readings) perform satisfactorily and are acceptable for operational use.