Abstract

Objective To compare the predictive performance of intertwin estimated fetal weight (EFW) discordance and EFW centile calculated according to either the Fetal Medicine Foundation (FMF) singleton or twin‐specific fetal growth charts for adverse perinatal outcome in dichorionic and monochorionic twin pregnancies. Methods This was a retrospective multicenter cohort study of twin pregnancies managed between January 2013 and December 2023 at three tertiary fetal medicine centers in the UK, Italy and Belgium. Twin pregnancies in which an obstetric ultrasound exam was performed to obtain fetal biometry within 2 weeks before live birth or diagnosis of intrauterine fetal demise of one or both twins were included. Cases with anastomotic complications were excluded. The primary outcome was a composite adverse perinatal outcome (CAPO), defined as stillbirth (intrauterine fetal demise ≥ 22 weeks) of at least one cotwin and/or iatrogenic early preterm birth (delivery < 34 weeks) for fetal indications. The predictive performance of intertwin EFW discordance and of the EFW centile of the smaller twin, calculated using either singleton or twin‐specific FMF fetal growth charts, was assessed for CAPO. Predictive models were developed using logistic regression analysis and evaluated using the area under the receiver‐operating‐characteristics curve (AUC); pairwise comparisons between models were performed using DeLong's test. Results The study analyzed 1294 dichorionic and 487 monochorionic twin pregnancies. For the prediction of CAPO, intertwin EFW discordance in dichorionic twins achieved an AUC of 0.93 (95% CI, 0.89–0.98), compared with 0.83 (95% CI, 0.77–0.90) (P = 0.001) and 0.87 (95% CI, 0.81–0.93) (P = 0.017) for EFW centile based on singleton and twin‐specific growth charts, respectively. Likewise, intertwin EFW discordance achieved an AUC of 0.95 (95% CI, 0.92–0.97) for predicting CAPO in monochorionic twins, outperforming EFW centile based on singleton charts (AUC, 0.80 (95% CI, 0.73–0.87); P < 0.001) and twin‐specific growth charts (AUC, 0.83 (95% CI, 0.76–0.90); P < 0.001). In clinical terms, at a 20% false‐positive rate (FPR), the sensitivity for CAPO in dichorionic twin pregnancies was 74%, 81% and 93% using singleton charts, twin‐specific charts and intertwin EFW discordance, respectively. Similarly, in monochorionic twin pregnancies, the sensitivity at a FPR of 20% was 75%, 79% and 98% using singleton charts, twin‐specific charts and intertwin EFW discordance, respectively. Conclusions Intertwin EFW discordance is a more reliable predictor of the composite outcome of stillbirth in at least one cotwin and/or iatrogenic early preterm birth in twin pregnancies than is EFW centile based on either singleton or twin‐specific growth charts. This approach addresses the limitations of using different fetal growth reference charts that rely on arbitrary cut‐offs. Following external validation, the use of intertwin EFW discordance prediction algorithms may potentially enhance risk stratification to improve clinical outcomes in twin pregnancies.