Identification of pathologically small fetuses using customized, ultrasound and population-based growth norms

Objectives Fetal growth restriction is a strong risk factor for stillbirth. We compared the performance of three fetal growth curves - customized, ultrasound (Hadlock) and population - in identifying abnormally grown fetuses at risk of stillbirth. Methods We performed a case-control study of singleton stillbirths (delivered between 2000 and 2010) at one center. Four liveborn controls were randomly identified for each stillbirth. Ultrasound-estimated fetal weight within 1 month prior to delivery was used to calculate growth percentiles for each fetus using three fetal growth norms. Sensitivities and odds ratios for stillbirth, as well as odds of abnormal growth according to formula, were calculated. Results There were 49 stillbirths and 197 live births. Using the customized norms, growth of the fetuses destined to be stillborn was bimodal, with both more small-for-gestational-age (SGA; < 10th percentile) and large-for-gestational-age (LGA; >= 90th percentile) fetuses. Odds of being abnormally grown were significantly higher using ultrasound compared with population norms (P = 0.02) but were not statistically different using ultrasound and customized norms (P = 0.21). Sensitivity for identification of SGA on ultrasound as a predictor of stillbirth was higher using customized (39%; 95% CI, 24-54%) or ultrasound (33%; 95% CI, 19-47%), rather than population (14%; 95% CI, 4-25%), norms. Conclusions Among fetuses destined to be stillborn, customized and ultrasound norms identified a greater proportion of both SGA and LGA estimated fetal weights. The customized norms performed best in identifying death among SGA fetuses. These results should be interpreted within the limitations of the study design. Copyright (C) 2014 ISUOG. Published by John Wiley & Sons Ltd.