Fetal weight estimation by 2D and 3D ultrasound: comparison of six formulas

To evaluate and compare the accuracy of different formulas to estimate fetal weight using 2D and 3D ultrasound.

We performed a prospective study on unselected singleton pregnancies. All scanned fetuses delivered within 7 days in absence of structural and chromosomal abnormalities were included. The fetal weight was calculated using the 3D Schild, Chang, Liang and 2D Hansmann, Merz and Hadlock formulas. Absolute and mean deviations of estimated fetal weight were calculated.

Of 249 scanned fetuses 200 that fulfilled the criteria were included. Birth weights ranged between 535 and 5020 g. The highest correlation between estimated fetal and birth weight was achieved by applying Schild's equation, and the absolute percentage error was 5.6%. The corresponding values for 2D Hansmann, Merz and four-parametric Hadlock formulas were 7.5%, 7.9%, and 9.2%, respectively, while these were 13.1% and 30.7% for Liang's and Chang's 3D equations. Using the Schild formula, a deviation from birth weight below 10 % was achieved in 80 % of fetuses, with Hansmann's in 73.5% and with Merz in 72.5%, while this parameter was much lower in the remaining equations.

The best option with the highest accuracy for sonographic fetal weight estimation was the 3D Schild equation followed by the 2D Hansmann and Merz formulas. Published data of the accuracy could be reproduced with the exception of the "Asian" 3D equations in our European population. The limited improvement in weight agreement using the 3D technique compared to the 2D technique may be outweighed by the time consumption.