Accuracy of sonographic estimates of fetal weight in very small infants

Objective Fetal outcome is inversely related to gestational age and birth w eight. Therefore, in very small fetuses, estimated weight may play an impor tant role in clinical management. Our aim was to determine the accuracy of sonographic estimates of fetal weight in very small infants. Design Retrospective chart review. Subjects We retrospectively studied 100 consecutive infants with a birth we ight of < 1000 g, at a gestational age between 24.0 and 34.0 weeks, in whic h biometric data < 2 weeks prior to delivery were available for analysis. Methods We estimated fetal weight with the use of two methods - by those of Hadlock and colleagues(3) and Scott and colleagues(4) - and compared the e stimated values with measured birth weights. Results The infants had a mean birth weight of 742 +/- 173 (SD) g, at a ges tational age of 28.1 +/- 2.0 (SD) weeks. With Hadlock's method, the mean es timated fetal weight (EFW) was 736 +/- 186 (SD) g, which was not significan tly different from birth weight; the mean EFW error was 0.8 +/- 12.7 (SD) % . With Scott's method, the mean EFW was 780 +/- 185 (SD) g, which was signi ficantly increased above birth weight; the mean EFW error was 5.7 +/- 12.5 (SD) %. The accuracy of the weight estimates was not significantly affected by the period between ultrasound examination and delivery if < 2 weeks, or by fetal growth restriction. Conclusion In our population of small fetuses, Hadlock's estimates of fetal weight correlated well with measured birth weight, whereas Scott's method tended to overestimate.