Background: We sought to identify modifiable factors to improve survival of neonatal biventricular repair by analyzing the cause of death and predictors of mortality and reintervention in the last 2 decades.
Methods: Between 1995 and 2016, 991 consecutive neonates were included. The cohort was divided by era: era I was from 1995 to 1999, era II 2000 to 2007, and era III 2008 to 2016. The Kaplan-Meier method was used to estimate freedom from death and reintervention. Univariable and multivariable Cox regression was applied to assess predictors for mortality or reintervention in the contemporary cohorts (2000-2016).
Results: Median age was 8 days (range, 5-13), and median body weight at operation was 3.3 kg (range, 2.9-3.6). The most common diagnosis was transposition with intact ventricular septum (32%), followed by transposition with ventricular septal defect (14.5%), and simple left-to-right shunt lesion (10.9%). There was significant improvement in survival from era I to eras II and III but no difference between eras II and III (1 year: 82.1% vs 89.4% vs 89.6%, respectively; P < .001). The most common cause of death was sudden death in eras I and III and cardiac in era II. Multivariable analysis revealed preoperative (P = .005)/postoperative (P < .001) extracorporeal membrane oxygenation and postoperative renal replacement (P < .001) as independent predictors for mortality. The reintervention rates were comparable between eras II and III (P = .53). Atrioventricular septal defects and common atrial trunk were identified as predictors for reintervention.
Conclusions: Survival after neonatal biventricular repair remained unchanged. Preventing sudden death, myocardial protection, and minimizing residual lesions are potential targets to improve outcomes.
Copyright © 2021 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.