A Noninvasive Prediction Model With Simple Echocardiographic Variables for Shunts Closure Possibility in Patients With Posttricuspid Valve Shunt Defect

Abstract

Background

As right heart catheterization (RHC) is invasive and not always accessible, this study developed a noninvasive model (P-echo) to predict shunt closure feasibility in adult congenital heart disease (ACHD) patients with post-tricuspid valve shunt defects (PTD), specifically isolated ventricular septal defects (VSD) and patent ductus arteriosus (PDA).

Methods

A retrospective analysis of 1474 VSD or PDA patients from 2012 to 2022 was conducted. Echocardiographic parameters were assessed, and key variables identified via LASSO regression. The P-echo model incorporated left to right velocity (LRv), right to left velocity (RLv), tricuspid regurgitation (TR), pulmonary artery diameter (PA), and RV/LV ratio. Its predictive performance was evaluated using ROC curve analysis.

Results

The P-echo model demonstrated excellent predictive performance with AUC values of 0.975 (95% CI: 0.965–0.984) in the derivation set, 0.963 (95% CI: 0.937–0.989) in the validation set, and high accuracy in both PDA (AUC 0.975, 95% CI: 0.965–0.984) and VSD (AUC 0.958, 95% CI: 0.936–0.980) subsets. In the derivation set, the model categorized patients into low (9.1% closure rate), medium (70.9% closure rate), and high-risk groups (99.7% closure rate) for shunt closure feasibility. Calibration plots confirmed the model's accuracy. Decision curve analysis showed a higher net benefit across a range of threshold probabilities, indicating the clinical usefulness of the model.

Conclusions

The P-echo model is a robust and reliable tool for predicting the feasibility of shunt closure in patients with PTD, offering a noninvasive alternative to RHC. This model can guide clinical decision-making and support individualized treatment strategies in ACHD management.