Evaluation of commercial point-of-care glucometers for detection and monitoring of neonatal hypoglycemia in resource-constrained settings

Accurate identification and treatment of hypoglycemia or low blood glucose is important for neonates, especially premature, small for gestational age, and intra-uterine growth restricted (“small and sick”) newborns. Existing point-of-care (POC) adult glucometers, intended for home management of adults or children with diabetes, are widely available and less costly than machines designed and validated for neonates, but they must be validated at the lower glucose levels relevant for neonates and at appropriate hematocrit levels before being used for neonates. We simulated neonatal blood by altering hematocrit (to 18%, 40%, and 55%) and glucose concentration (to 40 mg/dL and 150 mg/dL) of normal adult blood. We then used 11 POC glucometers to measure glucose levels in these samples and compared results to values measured with a clinical chemistry analyzer (YSI 2300). We report mean percent bias, mean absolute relative difference, and standard deviation for each tested condition and glucometer. We ranked the tested glucometers based on a combination of accuracy and precision for measuring neonatal glucose levels and consumable cost. From best-performing to worst, they are Group 1: StatStrip and StatStrip Xpress 2; Group 2: AccuChek Guide, AccuChek GuideMe, AccuChek Instant, and AccuChek Performa; Group 3: AccuChek Active, GlucoNavii; Group 4: HemoCue RT, Nipro Premier, and OneTouchUltra 2. This paper describes a simple, laboratory-based method to test glucometers across a wide range of hematocrit values and glucose concentrations and contributes up-to-date testing data on currently available POC glucometers at ranges relevant to neonatal use.