Accuracy of sonographic fetal weight estimation and prediction of birth-weight discordance in twin pregnancy: large single-center study

Objectives To determine the accuracy of sonographic fetal weight estimation in predicting birth weight (BW) and BW discordance in twin gestations, and to evaluate maternal and fetal characteristics that may affect the accuracy of this assessment. Methods This was a retrospective cohort study of all twins delivered at a single tertiary medical center between 2010 and 2021. Twin gestations for which sonographic estimation of fetal weight was performed within the week preceding delivery were included. Statistical analysis was performed to evaluate the strength of the correlation between sonographic estimated fetal weight (EFW) and BW, and to determine the impact of maternal and fetal factors on the accuracy of sonographic estimation. Results The study included 2154 twin pregnancies. There was a strong correlation between sonographic EFW and corresponding BW for all twins (r=0.922; P<0.001). Strong correlations were observed for both the presenting and non-presenting cotwin (r=0.921 and r=0.922, respectively; both P<0.001), as well as the larger and smaller cotwin (r=0.928 and r=0.934, respectively; both P<0.001). The overall mean +/- SD absolute error of sonographic EFW was 7.41 +/- 6.81%. This error was greater for the non-presenting cotwin compared with the presenting cotwin (7.99 +/- 6.12% vs 7.17 +/- 5.64%; P<0.001), and for the smaller cotwin compared with the larger cotwin (8.56 +/- 7.50% vs 6.58 +/- 5.47%; P<0.001). Advanced gestational age at scanning was correlated inversely with the mean absolute error of sonographic EFW. Multivariate logistic regression indicated that an earlier gestational age at scanning, being the non-presenting cotwin and being the smaller cotwin were independent risk factors for sonographic EFW inaccuracy. Pregnancies in which the presenting twin was estimated to be the smaller cotwin had twice the rate of false-positive BW discordance compared with pregnancies in which the presenting twin was estimated to be the larger cotwin (36.0% vs 13.0% for BW discordance > 15%, 35.0% vs 17.0% for BW discordance > 20% and 37.7% vs 12.1% for BW discordance > 25%; all P<0.001). The error in sonographic EFW discordance was not related to chorionicity, the position of the presenting fetus or gestational age at the time of fetal weight estimation. Conclusions Sonographic estimation of fetal weight within 7 days before delivery accurately predicts BW in twin pregnancy. Sonographic EFW accuracy is reduced for the non-presenting twin, the smaller cotwin and when delivery occurs at an earlier gestational age. Sonographic estimation of fetal weight discordance is less accurate when the presenting twin is the smaller cotwin. (c) 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.