BACKGROUND: The aim of the study was to evaluate the effect of hematocrit (HCT) on glucose meter assays based on different measurement techniques. METHODS: This paper studied glucose meters utilizing the glucose dehydrogenase reaction and four measurement techniques: colorimetry (HemoCue), reflectometry (Accu-Chek Active), amperometry (Optium Xido) and coulometry (Optium Omega). The EDTA venous blood samples HCT were modified by adding or removing defined aliquots of plasma. Glucose concentration was measured using each meter in 27 batches of blood samples, with HCT ranging from 20% to 60% in 10% increments. The data were analyzed using repeated measures models and a linear random effects model. RESULTS: A significant relationship between HCT and glucose reading in all meters was found and, for all meters except Optium Xido, there was a significant modification of this relationship by glucose level. The relative decrease in glucose concentration per 1% increase of the HCT value varied from 0.30% for Optium Omega in samples with glucose concentrations <5.55 mmol/L to 1.37% for Optium Xido in the same stratum (p<0.0001). The 5% glucose meter error (the ADA recommendation) was reached in the <5.55 mmol/L stratum after HCT change by 3.9%-16.7%. CONCLUSIONS: There is a significant continuous effect of HCT on measurement accuracy of glucose meters across its wide range of values and glucose concentrations. The most sensitive to the HCT interference was the system utilizing amperometric technique (Optium Xido) followed by the one with reflectometric technique (Accu-Chek Active), while the systems with the coulometric technique (Optium Omega) or colorimetric measurements in whole blood haemolysate (HemoCue) were less sensitive.
BACKGROUND: The aim of the study was to evaluate the effect of hematocrit (HCT) on glucose meter assays based on different measurement techniques. METHODS: This paper studied glucose meters utilizing the glucose dehydrogenase reaction and four measurement techniques: colorimetry (HemoCue), reflectometry (Accu-Chek Active), amperometry (Optium Xido) and coulometry (Optium Omega). The EDTA venous blood samples HCT were modified by adding or removing defined aliquots of plasma. Glucose concentration was measured using each meter in 27 batches of blood samples, with HCT ranging from 20% to 60% in 10% increments. The data were analyzed using repeated measures models and a linear random effects model. RESULTS: A significant relationship between HCT and glucose reading in all meters was found and, for all meters except Optium Xido, there was a significant modification of this relationship by glucose level. The relative decrease in glucose concentration per 1% increase of the HCT value varied from 0.30% for Optium Omega in samples with glucose concentrations <5.55 mmol/L to 1.37% for Optium Xido in the same stratum (p<0.0001). The 5% glucose meter error (the ADA recommendation) was reached in the <5.55 mmol/L stratum after HCT change by 3.9%-16.7%. CONCLUSIONS: There is a significant continuous effect of HCT on measurement accuracy of glucose meters across its wide range of values and glucose concentrations. The most sensitive to the HCT interference was the system utilizing amperometric technique (Optium Xido) followed by the one with reflectometric technique (Accu-Chek Active), while the systems with the coulometric technique (Optium Omega) or colorimetric measurements in whole blood haemolysate (HemoCue) were less sensitive.