Accurate assessment of fetal weight by ultrasound examination is critical to obstetric management, particularly close to term. Small-for-gestational-age (SGA) fetuses are vulnerable during the stresses of labor and delivery. Accurate identification of SGA fetuses allows for close monitoring and more informed decision-making about prolonging the pregnancy or operative delivery. Many studies evaluating ultrasound assessment of fetal weight have had poor methodology or statistical analyses. Chien and colleagues compared actual birth weight to that predicted by ultrasonography using four different formulas to calculate birth weight based on several biometric measurements at term.

They studied 50 pregnant women who were perceived to be at low risk of obstetric complications or fetal abnormality. Four ultrasound measurements were made on each fetus within seven days of term delivery. The measurements—biparietal diameter (BPD), femur length (FL), fetal abdominal area (FAA) and abdominal circumference (AC)—were all made by a single observer using standardized techniques. Gestational age was calculated based on the crown-rump length. Birth weight was measured immediately after birth, always using the same scales and techniques of measurement. The ultrasound measurements were used in four different formulas to calculate estimated birth weight. For each newborn, the estimated birth weights from the Aoki, Campbell and Wilkin, Shepard and Hadlock formulas were compared with the actual birth weight. The estimates were adjusted for weight gain between scan and delivery by adding 25 g (1.0 oz) per day.

Delivery occurred on average within four days of scanning. All four methods tended to underestimate fetal weight. The smallest mean difference was seen with the Shepard formula, which had a disparity of 51.4 g (1.8 oz), followed by the Aoki formula, which yielded a 60.5 g (2.1 oz) underestimation. The mean underestimations with the Campbell and Wilkin and Hadlock methods were 141.8 g (5.0 oz) and 190.7 g (6.7 oz), respectively. The range of agreement was narrowest with the Aoki formula and widest with the Campbell formula.

The authors conclude that all four of the formulas currently used in clinical practice provide valid estimates of true birth weight when ultrasonographic measurements are taken at term. The greatest accuracy was obtained with the Aoki formula, which uses three biometric variables (BPD, FAA and FL). Although the Shepard and Hadlock formulas use two variables, they differ from those used by the Aoki formula, which may explain the differences in validity between the two formulas. The relatively low validity of the Campbell and Wilkin formula may result from the use of only one variable.