Predicting Adverse Perinatal Outcomes in Dichorionic Twin Pregnancies: A Multicentre Cohort Study

Abstract

Objective: Dichorionic twin pregnancies are associated with increased risks of stillbirth or medically indicated early preterm birth (ePTB) to avoid stillbirths. This study evaluated the predictive value of fetal estimated weight (EFW) and Doppler indices before adverse perinatal outcomes.

Design: Retrospective multicentre cohort study.

Setting: Three tertiary centres in the UK, Italy and Belgium.

Population: The study included 1294 dichorionic twin pregnancies managed between 2013 and 2023.

Methods: Univariable and multivariable analyses assessed the association and the predictive accuracy between EFW and Doppler indices taken within 2 weeks of birth or adverse perinatal outcomes.

Main outcome measures: Stillbirths (of one or both twins) or medically indicated ePTB before 34 weeks' gestation for fetal indications.

Results: The study identified 58 pregnancies (4.5%) complicated by adverse perinatal outcomes. There were significant differences (all p < 0.001) between twins with adverse perinatal outcomes and liveborn twins for small for gestational age foetuses (89.5% vs. 59.3%), EFW discordance (31.8% vs. 8.4%), umbilical artery (UA) pulsatility index (PI) discordance (39.7% vs. 12.6%) and middle cerebral artery PI discordance (27.6% vs. 13.3%). These associations remained significant after adjusting for maternal characteristics and gestational age. The best predictive model included EFW discordance and UA PI discordance, with an area under the curve of 0.90.

Conclusions: The integration of intertwin EFW and UA PI discordance can effectively predict stillbirths or the need for medically indicated ePTB. After external validation in larger populations, this model could provide effective risk stratification of dichorionic pregnancies to enable targeted interventions to improve clinical outcomes.

Keywords: adverse perinatal outcome; estimated fetal weight; fetal growth restriction; intertwin discordance; intrauterine demise; middle cerebral artery; multiple pregnancy; stillbirths; twin pregnancies; umbilical artery.

FIGURE 1

ROC curves for EFW discordance versus SGA < 10th centile when used with UA PI discordance to predict adverse perinatal outcomes. AUC, area under the curve; CI, confidence interval; EFW, estimated fetal weight; SGA, small for gestational age; UA PI, umbilical artery pulsatility index.

FIGURE 2

Sensitivity and specificity of the prediction model including EFW discordance and UA PI discordance versus SGA less than 10th centile, 5th centile, and 3rd centile. CI, confidence interval; EFW, estimated fetal weight; SGA; small for gestational age; UA PI, umbilical artery pulsatility index.