Abstract
Background The purpose of this study was to revise the accuracy and precision of three glucometer-strips systems (GSS) in professional routine.
Methods Prospective single-centre study. Triplets of cPG were investigated on 3 GSS: 1. Galileo, 2. Calla, both employing glucoseoxidase strips (Wellion, Austria), 3. Contour Plus (Ascensia, Switzerland) employing glucosedehydrogenase strips. In each GSS three devices and one LOT of strips were employed. Following the cPG readings, venous blood was taken to measure venous plasma glucose (vPG) on Cobas Integra 400 analyzer employing hexokinase. A total of 258 pairs of cPG vs vPG were compared using Wilcoxon signed rank test with Bonferroni correction and software IBM SPSS Statistics for Windows, V. 23.0. Precision was estimated using SD of average differences between individual cPG measurements. P<0.05 was considered significant.
Results Deviations (mean±SD) from vPG≥5.55 mmo/L (100 mg/dL): Galileo 0.25±0.49 mmol/L (5±9 mg/dL), Calla -1.25±1.08 mmol/L (-23±19 mg/dL), Contour -0.63±0.71 mmol/L (-11±13 mg/dL). Number of cPG values within interval ±15 % of vPG: Galileo 196/207 (94.7 %), Calla 162/207 (78.3 %), Contour 200/207 (96.6%). For vPG values <5.55 mmol/L (100 mg/dL) the interval±0.83 mmol/L (15 mg/dL) reached: Galileo 11/51 (21.6 %), Calla 48/51 (94.1 %), Contour 51/51 (100 %). Galileo cPG <7.0 mmol/L (126 mg/dL) after subtraction of 0.7 mmol/L (12,4 mg/dL) reached 51/51 (100 %).
Conclusions Borderline accuracy of Galileo was found for cPG≥7 mmol/L (126 mg/dL) whereas for cPG<7 mmol/L (126 mg/dL) correction –0.7 mmol/L (13 mg/dL) was necessary. Accuracy of Contour but not of Calla was confirmed.