447: Tried-and-true versus up-and-coming: Which intrauterine growth curve best predicts small-for-gestational age at birth?

Abstract

To compare a traditional U.S. sonographic method for estimated weight (EFW) and intrauterine growth assessment with two recently published methods for the prediction of small-for-gestational age (SGA) at birth. We reviewed deliveries at our institution from January 1, 2013 to March 31, 2017. Singleton, non-anomalous, well-dated fetuses with an ultrasound within two weeks of delivery were included. EFWs were calculated and percentiles assigned using a traditional U.S. method (Hadlock, 1985) and two newer methods: Intergrowth-21st (INTG, 2016) and Salomon (2007). The Intergrowth-21st method is based on a diverse international population, and the Salomon method is derived from the population in Paris, France. Each method used different formulas for both the calculation of EFW and percentile (z score). Fetal growth restriction (FGR) was defined as EFW <10th percentile. We calculated the test characteristics for each method to predict SGA at birth, which was defined as birth weight <10th percentile using a nationwide U.S. cohort. The true positive rates were plotted graphically as a function of the false positive rates for each method to illustrate receiver-operator characteristics and compare each method’s balance of true and false positives. The mean percentile discrepancies between EFW and birth weight measurements were calculated as a measure of method accuracy. Statistical inference was made using ANOVA or Chi-square as appropriate. Of the 831 pregnancies with an ultrasound within 2 weeks of delivery, 138 (16.7%) were SGA at birth. While Hadlock identified more fetuses for FGR than either INTG or Salomon (p<0.01), it also most accurately predicted the birth weight percentile (p<0.001). Compared with the newer methods, the Hadlock method had the highest sensitivity and negative predictive value as well as the best performance on the receiver operator characteristics graph. For each additional case of SGA predicted with the Hadlock method, one false positive was created. The traditional Hadlock method identified more fetuses as having FGR and more accurately predicted birth weight percentile than either the INTG or Salomon methods. In spite of publication more than 30 years ago, it is superior to newer alternatives.View Large Image Figure ViewerDownload Hi-res image Download (PPT)