IN THE BALANCE: THE ACCURACY OF SONOGRAPHIC ESTIMATION OF FETAL WEIGHT

Dr. Andrew Edwards MBBS FRANZCOG DDU, Fellow in Obstetric and Gynaecological Ultrasound,
Monash Medical Centre

Introduction
An important component of obstetric care is fetal growth and well-being monitoring. In the past this was carried out by clinical assessment of the mother and her in utero fetus. Since the first attempts at sonographic estimation of fetal weight in the early 1970's, clinicians have increasingly utilised ultrasound in the belief that it offers greater accuracy than clinical estimation of fetal weight, and hence may improve pregnancy outcome. It has become a routine investigation in 'high-risk' pregnancies, such as those complicated by maternal hypertension, insulin-requiring diabetes, antepartum haemorrhage or breech presentation, and is often ordered to confirm the clinician's suspicion when palpation suggests macrosomia or growth retardation. Despite this, it cannot be categorically stated that ultrasound estimation of fetal weight is more accurate than clinical assessment. In fact, in a recent study of 1717 singleton pregnancies, clinical estimation of fetal weight was more accurate than sonographic fetal weight estimation for all babies with birth weight between 2500 and 4000g, and equally accurate for babies born weighing greater than 4000g. Overall, 72% of the clinical predictions were within 10% of the actual birth weight, compared to 69% of the sonographic predictions.

In this article I take a step back to look at the entire affair of sonographic estimation of fetal weight. I examine the development of the sonographic methods and the mathematical models, and consider why advances in ultrasound technology have not lead to the type of improvements in this area that we have seen in nearly all other areas of ultrasound over the past two decades. By describing, in detail, the process of sonographic estimation of fetal weight, I hope to allow sonographers to enhance accuracy, and clinicians to be aware of the limitations of the test.

Conclusion
As we have discussed, there are serious problems with calculating and reporting an EFW. The main problems relate to the calculation of an accurate EFW from the biometric information gathered during the ultrasound scan, and the lack of actual fetal weight charts with which to compare the EFW. Both of these issues could be avoided by using the actual measurements of fetal size, as our assessment of fetal well being, rather than the EFW. In other words, we could measure BPD, HC, AC, FL and any other biometric parameter that we think may help assess the well being of a fetus, and compare them with charts of those measurements. Altman and Chitty have published suitable robust charts of the major biometric parameters, developed from a prospective study of 663 unselected fetuses, each scanned once at a randomly allocated time during the pregnancy . The scan report could contain actual fetal biometric measurements, along with accurate comparisons with reference charts expressed as percentiles. This decreases the importance of the EFW because the clinician gets useful, and more accurate, information from the biometry. It depends on what the clinician really wants to know. An EFW may be important for making a decision regarding mode of delivery, but individual biometry measurements plotted on a percentile chart may be more useful in assessing the growth of an 'at risk' fetus. Probably a combination of EFW and biometry is an appropriate compromise, provided that sonographers, sonologists, and clinicians acknowledge the limitations of sonographic estimation of fetal weight.

Citation
Edwards A. In the balance: the accuracy of sonographic estimation of fetal weight. ASUM Bulletin 2001.1:3-6

Keywords
fetal measurement, obstetric, fetal weight, BPD, biparietal diameter, head circumference, abdominal circumference