512: Are appropriately sized fetuses who “fall off the curve” at increased risk for small-for-gestational age at birth?

Abstract

To evaluate whether fetuses with an estimated fetal weight (EFW) >10th percentile but a downward trend in the EFW percentile between 18-22 and 28-32 weeks have a higher rate of small-for-gestational age (SGA) at birth than those with a non-downward EFW trend. This was a retrospective cohort study of deliveries at our institution from January 1, 2013 to March 31, 2017. We included well-dated, non-anomalous, singleton pregnancies who had an ultrasound (US) exam at 18-22 weeks gestation with a subsequent exam at 28-32 weeks gestation, and whose estimated fetal weight (EFW) remained above the 10th percentile. EFWs and percentiles were calculated using the Hadlock formula. In order to identify the cutoff in EFW percentile downtrend between 18-22 and 28-32 weeks to best predict SGA at birth, we plotted the receiver-operator characteristics and then tested the cutoff which performed best. SGA was defined as a birth weight <10th percentile. The rate of SGA in those with down-trending EFW was compared to the SGA rate in pregnancies with a non-downtrending EFW. Chi square to compare dichotomous variables. We identified 523 patients who had ultrasounds done at both 18-22 and 28-32 weeks gestation with EFWs > 10th percentile, with an overall rate of SGA was 10.3% (n=54). When various cutoffs of decreasing EFW percentile to predict SGA were plotted on a receiver-operator curve, the cutoff with the best performance was a decrease of ≥ 20 points. The rate of SGA at birth among fetuses with an EFW that down-trended by 20 percentile points was 20.0%, versus only 7.1% for fetuses with a non-downtrending EFW percentile (p<0.001). In spite of this strong association, this EFW downtrend was not able to predict SGA well, with an area under the curve of 0.763, sensitivity of 48.1% and positive predictive value of 20.0%. Among fetuses with EFW >10th percentile, a downward trend of ≥ 20 percentile points is associated with an increased risk of SGA at birth, suggesting a need for additional growth surveillance. Used in isolation, however, this test generates a high number of false positives for each correctly identified SGA case.View Large Image Figure ViewerDownload Hi-res image Download (PPT)