Accuracy of sonographic estimation of fetal weight before induction of labor in diabetic pregnancies and pregnancies with suspected fetal macrosomia.

Abstract

To evaluate the accuracy of sonographic estimation of fetal weight (EFW) in diabetic pregnancies and pregnancies with suspected fetal macrosomia. 63 women with diabetic pregnancies, 74 nondiabetic women with suspected large-for-gestational-age (LGA) infants, and 161 controls underwent ultrasound assessment prior to induction of labor. EFW was compared to the weight at birth, 1-3 days later. EFW was highly correlated to birth weight. Absolute or actual weight differences between the birth weight and the EFW, and the rate of EFW within 10% of birth weight were not different between the groups. A linear regression model controlling for maternal and gestational age, diagnosis of gestational or pregestational diabetes, birth weight, gravidity, parity, nulliparity, placental location and AFI was not significantly correlated to the absolute or actual weight differences. In pregnancies with suspected LGA, higher birth weight was an independent and significant predictor of high weight difference inaccuracy. The ultrasonographic EFW 1-3 days before delivery is highly correlated with birth weight, reaffirming the clinical use of abdominal circumference and femur length in estimating fetal weight near labor at term. In pregnancies with suspected LGA fetuses and higher prevalence of macrosomia, ultrasound has higher sensitivity but lower specificity than the controls.