Evaluation of six commonly used formulae for sonographic estimation of fetal weight in a Sri Lankan population

Abstract

Objective: To compare six commonly used formulae (Shepard, Campbell, Hadlock I, II, III, and IV) for estimation of fetal weight in Sri Lankan population. Methods: A descriptive cross-sectional study conducted at Ward 9, Sri Jayewardenepura General Hospital, Kotte, Sri Lanka from the October to December 2007.The study included 86 singleton pregnancies. The fetal biometry - biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), femur length (FL) - ultrasonically measured on the previous day of planned delivery, actual birth weights (ABW) taken within 30 minutes of delivery. Estimated fetal weight (EFW) from each equation calculated using basic biometric values for estimation of fetal weight. Statistical analysis done using Karl Pierson correlation. Results: The highest positive correlation between the ABW and the EFW was seen in the Hadlock IV (r =0.836). According to the area under the curve for each ROC curve Hadlock IV equation has the highest (0.849). For fetuses less than 2.5 kg Hadlock IV has the highest sensitivity (75%). All the equations have the high specificity which ranges from93.6"h to 100%. In fetuses more than 3.5 kg the Shepard has the highest sensitivity (90%). The specificity is highest (97.4%) in Hadlock IV. Even though the Hadlock IV is the best formula identified in our study to predict babies less than 2.5 kg out of whole population, it is not the best for prediction of a baby larger than 3.5 kg (Hadlock IV sensitivity 40%) compared to Hadlock I (sensitivity 80% and specificity 89.5%) which is the best formula to predict a baby more than 3.5 kg. Conclusion: All formulae have adequate accuracy for estimating fetal weight in the population tested, Hadlock IV has the best accuracy. Key words: Fetal weight; obstertrics sonography. DOI: 10.4038/sljog.v31i1.1735 Sri Lanka Journal of Obstetrics and Gynaecology 2009; 31: 20-33