ABSTRACT
AIM: Despite targeting newborns at risk of hypoglycaemia based on clinical characteristics, blood glucose measured at 1 and 4 h of age is frequently normal. Identification of at-risk newborns at the greatest risk of hypoglycaemia would allow more targeted, earlier intervention. We aimed to determine the ability of calculated umbilical cord blood glucose extraction to discriminate hypoglycaemia in at-risk newborns in the first 4 h of life. METHODS: Newborns with paired arterial and venous cord blood glucose and 1 ± 4 h capillary or venous blood glucose measured using a blood gas analyser (radiometer) were retrospectively identified (n = 154). Hypoglycaemia was defined as a blood glucose ≤2.0 mmol/L. The ability of calculated umbilical cord blood glucose extraction to discriminate risk of hypoglycaemia was determined by an receiver operating characteristic (ROC) curve. RESULTS: Twenty-seven newborns (18%) had a blood glucose ≤2.0 mmol/L at either time point. Neither arterial nor venous cord blood glucose predicted early hypoglycaemia better than chance. The area under the ROC curve for umbilical cord blood glucose extraction (area under the ROC curve = 0.74, (95% confidence interval, 0.65-0.82)) was significantly better than chance and arterial or venous cord blood glucose. An umbilical cord blood glucose extraction of 16% had the best sensitivity (80%) and specificity (55%) for discriminating the risk of early hypoglycaemia. CONCLUSIONS: Umbilical cord blood glucose extraction discriminates the risk of early hypoglycaemia at 1 or 4 h of age. However, the clinical utility of this test is limited due to the low sensitivity and specificity. Its predictive value may be greater in specific subsets of at-risk newborns and warrants further investigation.
