Abstract
BACKGROUND: Pulmonary valve-sparing repair in Tetralogy of Fallot is associated with better long-term outcomes, but its feasibility remains uncertain in pediatric humanitarian patients, who often present late, with complex anatomy and comorbidities. OBJECTIVE: To develop a simple and interpretable decision tree model to predict the likelihood of pulmonary valve-sparing repair based on preoperative data in a humanitarian pediatric cohort. METHODS: A post-hoc classification and regression tree analysis was conducted on 115 pediatric humanitarian patients with Tetralogy of Fallot who underwent surgical correction at our center, between 2019 and 2023. Predictor variables included demographic, anthropometric, and anatomical parameters. RESULTS: The model achieved an overall accuracy of 80.0%, with a sensitivity of 88.2% and specificity of 56.7%. The pulmonary valve annulus diameter (> 0.855 cm) was the primary discriminator for pulmonary valve-sparing repair. CONCLUSION: A simple decision tree model using preoperative anatomical variables may support surgical planning in humanitarian contexts. Its interpretability and high sensitivity suggest potential value as a preoperative triage and planning aid. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13019-026-03882-8.