Variability of point-of-care testing blood glucometers versus the laboratory reference method in a tertiary teaching hospital

Background: Self-monitoring blood glucose (SMBG) has been established as a part of the point-of-care testing (POCT) in the management of patients diagnosed with diabetes mellitus. Various glucometers in the market have been used for this purpose but most are calibrated to whole blood. The recent recommendation by the International Federation of Clinical Chemistry (IFCC) is for glucose to be reported in plasma equivalent units which should correlate well with the clinical laboratory reference method. This is to minimize as much as possible the differences normally seen in the capillary whole blood glucose levels obtained during SMBG and routine plasma glucose levels reported by the laboratory. Objective: We evaluated the accuracy of four commercial blood glucose strips currently available in the local hospital setting against the laboratory reference method. Methods: We analyzed 145 whole blood samples collected into plain tubes from the walk-in patients in the hospital. A simultaneously whole blood sample was collected in sodium fluoride tubes and analyzed by the routine laboratory hexokinase method (RxL, Siemens). The four different types of glucometer strips used were AccuChek Performa, AccuChek Advantage, Abbott Optium and Johnson & Johnson Surestep. Results: The overall correlation coefficient for the four strips showed that the Accuchek Performa compared closest to the laboratory method which is R-2=0.9761. This was followed by the Accuchek Advantage R-2=0.9726. The Optium and Surestep systems showed R-2=0.9376 and R-2=0.9277 respectively. The coefficient variance (CV) for the Accuchek Performa is also <5% for both the control and sample precision. The ISO15197.2003 (E) acceptance criteria: percentage of individual results falling outside 0.8 mmol/L at glucose concentrations <4.1mmol/L and outside 20% of the reference result at glucose concentrations >4.1 mmol/L will not be higher than 5% showed that the compliance for Performa is 100% compared to the rest (for percentage of results within 95% compliance limit). Conclusion: The results showed that the Accuchek Performa system compared the closest to the laboratory method and complies to the ISO15197.2003 (E) with acceptable CV's.