Defining intertwin birth weight discordance and optimal delivery time in dichorionic twins: a retrospective cohort study

Abstract

Background A high birth weight discordance (BWD) places twins at high risks of adverse perinatal outcomes; however, there is no commonly accepted definition of intertwin BWD. Given that gestational age (GA) is closely associated with both BWD and adverse neonatal outcomes, the aim of our study is to develop a measure that takes both BWD and GA into account to define BWD and explore the potentially optimal delivery time in dichorionic twins, with the balance against the risk of intrauterine demise and adverse neonatal outcomes. Methods About 99,000 mixed gender twin pairs from the US matched multiple birth file from 1995 to 2000 were enrolled in our study. We set up a two-dimensional matrix model to select the BWD cutoff point related to outcomes. The optimal delivery time was identified based on the risks of adverse neonatal outcomes and fetal death by GA. Results The combination of BWD and GA had a higher predictive performance for adverse neonatal outcomes than BWD alone (area under ROC curve (AUC): 0.816 vs. 0.523, p < .001). When GA was controlled for, the growth discordant cutoff point was BWD at the 40% level. The optimal delivery time suggested in dichorionic twins with BWD higher than 40% was identified at 33–34 gestational weeks. Conclusions The combination of BWD and GA was a better predictor for adverse neonatal outcomes than BWD alone. In dichorionic twins with BWD >40%, pregnancy may be continued till 33–34 gestation weeks if there is no indication for immediate delivery.