RESUMEN
The study was performed to evaluate the accuracy of the StatStrip (SS) and SureStep Flexx (SF) glucose meters compared to plasma glucose in infants at risk for neonatal hypoglycemia and to determine the effect of bilirubin and hematocrit on the results. A prospective cross-sectional study was conducted on 172 venous blood glucose samples from infants who had initial low point-of-care (POC) glucose tests measured simultaneously by SS and SF. Plasma glucose levels were compared to both POC instruments, and the effect of bilirubin and hematocrit levels on mean glucose differences were analysed. Mean (SD) plasma glucose was 2.12 (0.45) mmol/L; (range, 1.11-3.06 mmol/L). Mean (1.96SD) glucose differences of the SS versus SF were 0.21 (0.70) mmol/L and -0.04 (0.78) mmol/L, respectively. SS sensitivity was 94.7 % with an 86.1 % negative predictive value (NPV) at 2.8 mmol/L, while the SF had a 100 % sensitivity and NPV at the same cut-off level. No correlations were identified between mean glucose differences and either hematocrit or bilirubin levels in both glucose meters. Both the SS and SF glucose meters have limited use when compared to plasma glucose. Hence, they can only be employed as screening tools in at-risk neonates with an appropriate, predetermined cut-off level. Hematocrit and bilirubin levels did not affect the accuracy of both devices.