WIP 126 - The accuracy of Fenton compared to WHO growth charts in detecting hypoglycemia in susceptible term neonates

Poster Number: WIP 126
Publication Number: WIP 126.2424

Background: Infants classified as small-for-gestational-age (SGA; birth weight < 10th percentile) or large-for-gestational-age (LGA; birth weight >90th percentile) are at risk for hypoglycemia. Due to the differences in the WHO and Fenton growth charts, infants evaluated with both curves will not be consistently categorized as SGA or LGA. It is important to accurately identify infants at risk for hypoglycemia to avoid any potential serious morbidities and to avoid unnecessary interventions in infants not at risk for hypoglycemia. (Figure 1)

Objective: Our goal is to compare the accuracy of anthropometric measurements using Fenton Growth Chart 2013 (FGC) and WHO Growth Chart 2006 (WGC) as screening tools in identifying infants who are at risk for hypoglycemia.

Design/Methods: In a single center retrospective chart review, 300 infants admitted to the well nursery (WBN) from January 1, 2012 to December 31, 2023 with a gestational age of 37 0/7 wk - 41 6/7 wk will be included to achieve 80% power with a significance threshold of 0.05. Infants classified as SGA or LGA for sex according to the WGC and/or FGC will be included. Infants will be excluded if other risk factors for hypoglycemia are present such as maternal diabetes, perinatal depression, family history of hypoglycemia, major chromosomal/congenital anomalies or if point of care blood glucoses are not available. Information about maternal pregnancy and labor and delivery courses and neonatal nursery courses will be abstracted.

A confusion matrix will be used to visualize the performance of each growth curve in classifying infants as at risk for hypoglycemia (either SGA or LGA) or not at risk (AGA). Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) will be calculated. Non-inferiority of the WGC growth curve to the FGC will be determined by calculating relative PPV and NPV along with the respective 95% confidence interval for each growth curve to see if the confidence intervals are bounded below 1.