Study on Heart Block Risks Using Outcome-Based Modeling

In recent years, transcatheter aortic valve replacement (TAVR) has emerged as a minimally invasive surgical approach and has gradually become a standard treatment option for patients with moderate to severe aortic stenosis（AS), particularly those at high surgical risk or deemed ineligible for conventional surgery. TAVR has been widely adopted in clinical practice.

However, post-TAVR cardiac conduction disturbances are among the most frequently observed complications, particularly complete left bundle branch block (CLBBB) and high-degree atrioventricular block (AVB) requiring permanent pacemaker implantation (PPI). These complications are associated with an increased risk of heart failure and cardiac mortality. Identified risk factors include patient-specific characteristics, electrocardiographic features, intraoperative procedural techniques, and the type of device used.

Current research on post-TAVR conduction disorders predominantly focuses on Western populations. However, Chinese AS patients exhibit distinct clinical features-such as higher degrees of valve calcification and a higher prevalence of bicuspid aortic valves-which may lead to differences in risk factors for conduction abnormalities. Moreover, prior studies have largely focused on isolated risk factors rather than developing integrated models to quantify the risk of such complications. Against this backdrop, the purpose of this study is to address the knowledge gap in postoperative risk prediction for Chinese patients. By constructing a risk prediction model tailored to Chinese AS patients for post-TAVR cardiac conduction block, we aim to improve preoperative risk assessment and individualized clinical management, thereby providing actionable guidance for clinical decision-making.

This study included 112 aortic stenosis patients undergoing TAVR at Sun Yat-sen Memorial Hospital (2017-2024), divided into internal (2017-2022) and external (2023-2024) validation cohorts.

Data covered demographics, medical history, ECG, echocardiography, biomarkers, and valve parameters. Outcomes were permanent pacemaker implantation (PPI) ≤48h post-TAVR and complete left bundle branch block (CLBBB).

Univariate logistic regression (p\<0.20 threshold) and stepwise multivariate analysis (R "MASS" package, AIC-based) identified predictors for nomogram development. Model performance was assessed via AUC, C-index, ROC, and Hosmer-Lemeshow test. Internal validation used 1000 bootstrap resamples; external validation employed temporal validation. Analyses were performed in R 4.2.3 (significance: p\<0.05).