Comparing fetal biometric growth velocity versus estimated fetal weight for prediction of neonatal small for gestational age

Abstract

Objective Growth velocities derived from fetal biometrics have been proposed to improve prediction of small for gestational age (SGA) neonates. We sought to determine if ultrasound growth velocities for abdominal circumference (AC) and estimated fetal weight (EFW) improve the prediction of SGA infants when compared to using EFW alone. Study design This is a secondary analysis from a prospective study of women referred for growth ultrasounds during the third trimester. Growth velocities for AC and EFW were derived from the difference in Z-scores between measurements at the anatomy survey (18–22 weeks gestation) and later growth ultrasound (26–36 weeks gestation). Change in AC and EFW growth velocities <10th percentile were compared with prenatally suspected SGA from Hadlock EFW <10th percentile for prediction of SGA neonates. The primary outcome was defined as the sensitivity and specificity of the growth velocities and Hadlock EFW in predicting SGA neonates. Logistic regression modeling was used to determine if the growth velocities improved prediction of neonatal SGA. Area under the ROC curves (AUC) were determined and compared. Results Of 612 singleton pregnancies meeting inclusion criteria, 68 (11.1%) resulted in SGA neonates. Hadlock EFW <10th percentile had higher sensitivity and specificity when compared to AC growth velocity and EFW growth velocity. Only AC growth velocity and Hadlock EFW had significant odds ratios for association with neonatal SGA. The AUC were 0.54, 0.53, and 0.61 using AC growth velocity, EFW growth velocity, and Hadlock EFW, respectively. The AUC did not significantly improve when the growth velocities were combined with Hadlock EFW (0.63). Adjustment of Z-scores for gestational age at anatomy scan or third trimester growth scan did not significantly change these results (AUC = 0.69). Conclusion EFW determined by Hadlock formula has the highest predictive value in detecting SGA neonates when compared to both AC and EFW growth velocities.