Development and validation of a prognostic risk model for pediatric patients with left-to-right shunt congenital heart disease and heart failure

This retrospective study aimed to develop a reliable prognostic model for pediatric patients with left-to-right (L-R) shunt congenital heart disease (CHD) complicated by heart failure (HF), based on clinical data collected from 407 cases between August 2012 and June 2024. The cohort included 63.4% male patients with a median age of 4.33 months, and participants were randomly assigned to a training set (<i>n</i> = 284, 69.8%) and an internal validation set (<i>n</i> = 123, 30.2%). Univariate and multivariate Cox regression analyses combined with LASSO regression were used to identify key independent prognostic variables for overall survival (OS): modified Ross classification, N-terminal pro-brain natriuretic peptide (NT-proBNP) levels exceeding 5,743 pg./mL, elevated blood urea nitrogen (BUN), presence of shock, and history of thoracotomy. These factors were integrated to construct a nomogram, which showed strong prognostic performance-concordance indices (C-index) were 0.829 in the training set and 0.850 in the validation set. Survival analysis revealed significant differences in 3-, 6-, and 12-month OS between high-risk and low-risk groups stratified by nomogram scores. These findings suggest that the proposed nomogram, when applied to pediatric populations with L-R shunt CHD and HF who share baseline characteristics similar to those in our study, could serve as a promising auxiliary tool for early risk stratification and clinical decision-making.