Background: This study was undertaken to evaluate the quality of ultrasound estimation of fetal weight when performed by midwives experienced in ultrasound examinations. We also examined whether the accuracy was affected by fetal presentation, twin pregnancy or birth weight category. The results of 5 different formulas were compared to determine which was most accurate in our study population.
Methods: The study population consisted of 620 fetuses in 607 pregnancies, on whom fetal weight estimations had been performed within 3 days prior to delivery. The group of twins (n=27) was analysed separately. Results achieved by Hadlock 2 formula used in our unit were compared with 4 other widely used formulas for estimation of fetal weight.
Results: With Hadlock 2 formula, mean absolute percent error was 6.2% and SD of error was 7.6% of mean birth weight. A total of 81% of estimates were within 10% of the actual birth weight. All the formulas tended to overestimate the weight of twins and fetuses weighing <2,500 g, and underestimate the weight of fetuses >4,000 g. Presentation of the fetus did not significantly influence the accuracy. The formula Hadlock 2, using 3 parameters (biparietal diameter, abdominal circumference and femur length) gave the highest ICC of 0.910.
Conclusions: Ultrasound estimation of fetal weight performed by midwives is feasible and of similar accuracy as in the original studies. Major errors may occur both in small and large birth weight groups.