Small and large foetuses: Identification and estimation of foetal weight at delivery from third-trimester ultrasound data

Abstract Background The estimation of foetal weight (EFW) at delivery is crucial to assess the risk of foetal and neonatal morbidity and mortality, most notably when the foetus is small or large. Aim To accurately predict the EFW at delivery of small foetuses (birth weight [BW] ≤ 2500 g) and large foetuses (BW ≥ 4000 g) identified with third-trimester ultrasound data. Methods We included 1309 foetuses whose age and weight at birth were known and for whom standard data were available from third-trimester ultrasound scans. Small and large foetuses were identified by extrapolation to full term of the estimated foetal weight obtained using Hadlock's equation. We built two regression models for predicting the birth weights of small and large foetuses, respectively. The results obtained with these models were compared to those obtained with Hadlock's equation. Results Third-trimester sonograms were obtained at 33.6 ± 1.3 weeks gestational age [WGA] and birth occurred at 38.7 ± 1.2 WGA. EFW of small foetuses predicted using the regression model showed significantly less systematic bias than the Hadlock estimate (2.3% vs. 7.2%, respectively), whereas random errors were similar. EFW of large foetuses predicted using the regression model showed significantly less random error than the Hadlock estimate (6.2% vs. 10.1%, respectively), whereas systematic bias was similar. Data from an independent validation sample indicate that our regression models are accurate. Conclusions To apply distinct models for accurately predicting the EFWs at delivery of small and large foetuses should prevent adverse events related to newborn size.