Glucometers are commonly used in a variety of healthcare settings and use in critically ill patients should not be assumed without appropriate tool validation. The study objective was to evaluate the accuracy of three point-of-care glucometers (POCGs) to assess glucose concentration in human blood sample. The POCGs tested included three different instruments and utilized three factors (hematocrit [Hct], galactose and ascorbic acid) in glucose measurements to determine the glucometers' accuracy and compared to the reference laboratory biochemical analyzer (Cobs 8000, Roche, Basal, Switzerland). In this study, the Nova StatStrip glucometer showed no significant variation compared to the laboratory method at high glucose level with various Hct%. ACCU-Chek Inform II overestimated the glucose results at Hct 22% and underestimated at Hct 62%. The Freestyle glucometer showed lower glucose levels compared to the Cobas 8000 at Hct 62%. The ACCU-Check showed significant increase of blood glucose with low Hct% levels when compared to the laboratory method. The Freestyle showed a decreased level of glucose with high Hct 62% interference compared to the Cobas 8000. Galactose interference 100 and 200 mg/dL dramatically affected the accuracy of ACCU-Chek Inform II. Nevertheless, among all three POCGs in this study, the Nova StatStrip showed the most reliable and stable results for glucose level in the presence of interference. Especially, those in critical care units, whereas the Freestyle Precision Pro and ACCU-Chek Inform II were insufficiently accurate for critically ill patients.
Keywords: Diabetes; glucometers; glucose readings; hematocrit; point-of-care.