Comparing fetal ultrasound biometric measurements to neonatal anthropometry at the extremes of birthweight.

OBJECTIVE:Error in birthweight prediction by sonographic estimated fetal weight (EFW) has clinical implications, such as avoidable cesarean or misclassification of fetal risk in labor. We aimed to evaluate optimal timing of ultrasound and which fetal measurements contribute to error in fetal ultrasound estimations of birth size at the extremes of birthweight. STUDY DESIGN:We compared differences in head circumference (HC), abdominal circumference (AC), femur length (FL), and EFW between ultrasound and corresponding birth measurements within 14 (n=1290) and 7 (n=617) days of birth for small (SGA, <10th percentile), appropriate (AGA, 10th-90th), and large-for-gestational age (LGA, >90th) newborns. RESULTS:Average differences between EFW and birthweight for SGA neonates were: -40.2g (CI -82.1, 1.6) at 14 days vs. 13.6g (CI -52.4, 79.7) at 7 days; for AGA, -122.4g (-139.6, -105.1) at 14 days vs.-27.2g (-50.4, -4.0) at 7 days; and for LGA, -242.8g (-306.5, -179.1) at 14 days vs. -72.1g (-152.0, 7.9) at 7 days. Differences between fetal and neonatal HC were larger at 14 versus 7 days, and similar to patterns for EFW and birthweight, differences were largest for LGA at both intervals. In contrast, differences between fetal and neonatal AC were larger at 7 versus 14 days, suggesting larger error in AC estimation closer to birth. CONCLUSIONS:Using a standardized ultrasound protocol, SGA neonates had ultrasound measurements closer to actual birth measurements compared to AGA or LGA neonates. LGA neonates had the largest differences between fetal and neonatal size, with measurements 14 days from delivery showing 3-4-fold greater differences from birthweight. Differences in EFW and birthweight may not be explained by a single fetal measurement; whether estimation may be improved by incorporation of other knowable factors should be evaluated in future research.