Fetal weight estimation by real-time ultrasound measurement of biparietal and transverse trunk diameter.
Measurements of biparietal diameter and transverse trunk diameter with the use of linear array real-time B-mode scanning were performed on 100 fetuses within 72 hours of delivery to evaluate the reliability of Hansmann and associates, nomogram in predicting fetal weight. Under optimal conditions, 82% of predicted birth weights were within 10% of actual birth weight with a mean error of 150 gm (6%) and a correlation coefficient of 0.98. Predictions were most accurate in fetuses weighing less than 1,500 gm and in the 2,000 to 4,000 gm group and were least accurate in the infant weighing less than 1,500 gm and in the 2,000 to 4,000 gm group and were least accurate in the infant weighing more than 4,000 gm. Once the accuracy of the method was determined, It was applied to 100 consecutive cases to evaluate the feasibility of this technique in clinical practice. Measurements were unobtainable in 15%, but in the remainder, predictions were of sufficient accuracy throughout the range of fetal weights to complement clinical assessment in the management of high-risk pregnancies. Seventy-four percent of predicted weights were within 10% of actual birth weight; the mean error was 165 gm (7.2%) and the correlation coefficient was 0.97.