Background: Adult congenital heart disease (ACHD) patients pose unique challenges in identifying the time for transplantation and factors influencing outcomes.
Objective: To identify hemodynamic, functional, and laboratory parameters that correlate with 1- and 10-year outcomes in ACHD patients considered for transplantation.
Methods: A retrospective chart review of long-term outcomes in adult patients with congenital heart disease (CHD) evaluated for heart or heart + additional organ transplant between 2004 and 2014 at our center was performed. A machine learning decision tree model was used to evaluate multiple clinical parameters correlating with 1- and 10-year survival.
Results: We identified 58 patients meeting criteria. D-transposition of the great arteries (D-TGA) with atrial switch operation (20.7%), tetralogy of Fallot/pulmonary atresia (15.5%), and tricuspid atresia (13.8%) were the most common diagnosis for transplant. Single ventricle patients were most likely to be listed for transplantation (39.8% of evaluated patients). Among a comprehensive list of clinical factors, invasive hemodynamic parameters (pulmonary capillary wedge pressure (PCWP), systemic vascular pressure (SVP), and end diastolic pressures (EDP) most correlated with 1- and 10-year outcomes. Transplanted patients with SVP < 14 and non- transplanted patients with PCWP < 15 had 100% survival 1-year post-transplantation.
Conclusion: For the first time, our study identifies that hemodynamic parameters most strongly correlate with 1- and 10-year outcomes in ACHD patients considered for transplantation, using a data-driven machine learning model.
Keywords: classification systems; clinical decision-making; heart disease: congenital.
© 2023 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.