Sonographic fetal weight estimation prior to delivery: analysis of accuracy and interobserver variability between obstetrician and radiologist

Abstract

The aim of this study was to evaluate the interobserver reliability of sonographic fetal weight estimation between experienced obstetrician and radiologist immadiately prior to delivery and to compare the sonographic accuracies of estimated fetal weights (EFW) between the sonographers. This was a prospective study of singleton pregnancies that underwent sonographic fetal weight estimation prior to scheduled delivery. One obstetrician and one radiologist, performed all examinations prior to delivery. Sonographic fetal weight estimations were obtained from each ultrasonographic examination. EFWs were compared with the actual birth weight (ABW) and interobserver reliability of sonographic fetal weight estimation between two sonographers was assessed. A total of 245 subjects were included. Inter-observer reliability coefficient for EFW was 0.864 (95% confidence interval: 0.825-0.894, P < 0.001). Mean values of EFWs by radiologist and obstetrician did not significantly differ (P = 0.480) while both did significantly differ from the ABW (P < 0.001, for both comparisons). The mean absolute error for EFW obtained by radiologist and obsterician was 227 (+/- 160) and 197 (+/- 138), respectively (P < 0.01). The mean absolute percentage error for EFW obtained by radiologist and obsterician was 7.09 (+/- 5.17) and 6.15 (+/- 4.38), respectively (P < 0.01). EFWs measured by radiologists and obstetricians are similar but both are significantly distant from the ABW. However, obstetricians measure EFW significantly more accurate than the radiologists do.