A validation study of ultrasonic foetal weight estimation models for Hong Kong Chinese singleton pregnancies.

OBJECTIVE To validate the published regression models for ultrasonic foetal weight estimation in Hong Kong Chinese singleton pregnancies. DESIGN Prospective cohort study. SETTING Regional hospital, Hong Kong. PARTICIPANTS One hundred and fourteen Hong Kong Chinese women with singleton pregnancy at term (37-42 weeks). MAIN OUTCOME MEASURES The birth weight of the neonate was used to validate the ultrasonic foetal weight estimation models. The ultrasound used to collect the data was performed within 2 days prior to delivery. RESULTS Foetal weight estimation models published by Hadlock and Woo have a high intraclass correlation coefficient of 0.86. Woo's regression model has the least mean difference (0.2 g; 95% limits of agreement, -569.4 to 569.8 g) and Hadlock 3 has the smallest limits of agreement (-114.6 g; 95% limits of agreement, -663.4 to 434.2 g) among the models tested. CONCLUSIONS Woo's regression model of foetal weight estimation gave the least mean difference and the actual birth weight for our local population were within the acceptable limits of agreement.