International estimated fetal weight standards of the INTERGROWTH-21 super(st) Project

• Published in: Ultrasound in obstetrics & gynecology 2017-04, Vol.49 (4), p.478-486
• Main Authors: Stirnemann, J, Villar, J, Salomon, L J, Ohuma, E, Ruyan, P, Altman, D G, Nosten, F, Craik, R, Munim, S, Cheikh Ismail, L, Barros, F C, Lambert, A, Norris, S, Carvalho, M, Jaffer, YA, Noble, JA, Bertino, E, Gravett, M G, Purwar, M, Victora, C G, Uauy, R, Bhutta, Z, Kennedy, S, Papageorghiou, A T
• Format: Article
• Language: English
• ISSN: 0960-7692; 1469-0705
• DOI: 10.1002/uog.17347

Objective Estimated fetal weight (EFW) and fetal biometry are complementary measures used to screen for fetal growth disturbances. Our aim was to provide international EFW standards to complement the INTERGROWTH-21 super(st) Fetal Growth Standards that are available for use worldwide. Methods Women with an accurate gestational-age assessment, who were enrolled in the prospective, international, multicenter, population-based Fetal Growth Longitudinal Study (FGLS) and INTERBIO-21 super(st) Fetal Study (FS), two components of the INTERGROWTH-21 super(st) Project, had ultrasound scans every 5 weeks from 9-14 weeks' until 40 weeks' gestation. At each visit, measurements of fetal head circumference (HC), biparietal diameter, occipitofrontal diameter, abdominal circumference (AC) and femur length (FL) were obtained blindly by dedicated research sonographers using standardized methods and identical ultrasound machines. Birth weight was measured within 12h of delivery by dedicated research anthropometrists using standardized methods and identical electronic scales. Live babies without any congenital abnormality, who were born within 14 days of the last ultrasound scan, were selected for inclusion. As most births occurred at around 40 weeks' gestation, we constructed a bootstrap model selection and estimation procedure based on resampling of the complete dataset under an approximately uniform distribution of birth weight, thus enriching the sample size at extremes of fetal sizes, to achieve consistent estimates across the full range of fetal weight. We constructed reference centiles using second-degree fractional polynomial models. Results Of the overall population, 2404 babies were born within 14 days of the last ultrasound scan. Mean time between the last scan and birth was 7.7 (range, 0-14) days and was uniformly distributed. Birth weight was best estimated as a function of AC and HC (without FL) as log(EFW)=5.084820-54.06633(AC/100) super(3)-95.80076(AC/100) super(3)log(AC /100)+3.136370(HC/100), where EFW is in g and AC and HC are in cm. All other measures, gestational age, symphysis-fundus height, amniotic fluid indices and interactions between biometric measures and gestational age, were not retained in the selection process because they did not improve the prediction of EFW. Applying the formula to FGLS biometric data ( n =4231) enabled gestational age-specific EFW tables to be constructed. At term, the EFW centiles matched those of the INTERGROWTH-21 super(