Predicting fetal and neonatal demise after fetoscopy for twin-twin transfusion syndrome using recursive partitioning.

ABSTRACT

OBJECTIVE: Donor demise after laser surgery for twin-twin transfusion syndrome (TTTS) is well-characterized, but recipient demise is not, nor is neonatal death. This study aims to characterize factors associated with recipient death, donor death, and dual twin death after laser, both before and after birth. METHODS: This is a prospective cohort study of monochorionic twin pairs who underwent laser ablation for TTTS. Risk factors for fetal and neonatal death of both twins were identified using univariable analysis and recursive partitioning, a novel statistical method to quantify contributions of each factor to outcomes. RESULTS: In 413 twin pairs, death of both twins occurred in 9.2% (38/413), donor death in 12.1% (50/413), and recipient death in 2.4% (10/413). Recursive partitioning showed that gestational age at delivery predicts dual twin death (below 23.7 weeks, likely [p < 0.001], above 28.3 weeks, unlikely [p = 0.004]). Abnormal umbilical artery Doppler and weight discordance predict donor demise (p < 0.001 and p = 0.033, respectively). Cervical length under 16 mm predicts neonatal death of both twins (p < 0.001). CONCLUSIONS: Parents can gain individualized information about the survival of each fetus based on variables available from preoperative and delivery variables. Short cervix and premature delivery cause significant mortality in TTTS.