Comparison of Four Point-of-Care Glucometers for Neonatal Glucose Testing-A Diagnostic Study.
PURPOSE Point-of-care (POC) glucometers used for neonatal glucose monitoring were originally designed for adults and may be affected by neonatal-specific factors such as high hematocrit and oxygen therapy. Although newer-generation glucometers incorporate features to minimize such errors, their diagnostic performance in neonates remains inadequately evaluated. This study aimed to assess the accuracy and precision of four newer-generation glucometers for neonatal glucose screening and evaluate the influence of factors such as hematocrit and oxygen use on measurement accuracy, comparing performance against International Organization for Standardization (ISO) 15197:2013, CLSI POCT12-A3, and United Nations Children's Fund (UNICEF) standards. METHODS A prospective diagnostic accuracy study was conducted among 199 neonates requiring glucose monitoring after obtaining parental informed consent. Whole blood obtained from venous sampling was analyzed using four glucometers: Contour Plus, Accu-Chek Instant, OneTouch Select Plus, and FreeStyle Optium Neo. Simultaneously, 1 mL of blood was collected in a fluoride-oxalate vial for laboratory plasma glucose estimation using standard methods. Agreement between glucometer and laboratory values was assessed using Bland-Altman percentage-difference plots. Glucometer performance was evaluated against ISO 15197:2013 and CLSI POCT12-A3 criteria, as well as the UNICEF evaluation framework (optimal ±6% and minimal ±8% deviation from laboratory values). FINDINGS For laboratory plasma glucose values ≥100 mg/dL, 100% of Accu-Chek Instant, 100% of Contour Plus, and 96.7% of OneTouch Select Plus readings were within ±15% of laboratory values, meeting ISO 15197:2013 criteria requiring ≥95% agreement. However, only 88% of readings met CLSI/ISO accuracy standards at this threshold. For glucose values <100 mg/dL, all glucometers demonstrated readings within ±15% of laboratory values and fulfilled ISO criteria, while Accu-Chek Instant and Contour Plus additionally met CLSI standards. None of the glucometers met UNICEF performance benchmarks. Hematocrit showed minimal association with measurement differences between glucometers and laboratory values (Pearson's r ranging from -0.258 to +0.196), and patient characteristics including sex, gestational age, oxygen therapy, and inotrope use had negligible effects on measurement discordance. IMPLICATIONS All four newer-generation glucometers demonstrated good precision in neonatal glucose measurement, particularly for glucose values <100 mg/dL, with performance generally aligning with ISO standards. However, none fulfilled the more stringent UNICEF criteria, highlighting the need for neonatal-specific validation standards for POC glucose monitoring devices.