Key result
Post-TAVR increase in AI ECG AK-AVS score linked to ~78% higher 1-year mortality.
Why the study?
AI ECG analysis is a potential tool for aortic stenosis assessment, but its utility in predicting TAVR outcomes remains unexplored.
Do periprocedural changes in an AI-based ECG model predict 1-year mortality, permanent pacemaker implantation, and length of stay in TAVR recipients?
Cohort (n=1,596)
No
Do periprocedural changes in an AI-based ECG model predict 1-year mortality, permanent pacemaker implantation, and length of stay in TAVR recipients?
Odds Ratio: 1.78 (95% CI 1.09–2.91)
Absolute Event Rate: 7.9% vs 4.3%
p-value: p=0.02
An increase in the AI-based ECG aortic stenosis score immediately after TAVR is associated with higher 1-year mortality, increased risk of permanent pacemaker implantation, and longer hospital length of stay.
Post-TAVR AK-AVS increases identify higher-risk patients; hypothesis-generating for AI-ECG integration in TAVR risk stratification.
Background Artificial intelligence (AI) electrocardiogram (ECG) analysis has emerged as a potential tool for aortic stenosis assessment, but its utility in transcatheter aortic valve replacement (TAVR) outcomes remains unexplored. Objective To evaluate the association between a validated AI ECG-based aortic stenosis model (AccurKardia AK-AVS) and clinical outcomes following TAVR. Methods In this retrospective study of TAVR recipients from 2012-2023 at a tertiary care hospital, we analyzed ECGs immediately before and after TAVR using the AK-AVS model. Baseline scores and post-TAVR score changes were evaluated against clinical outcomes. One-year mortality and permanent pacemaker implantation were modeled using logistic regression, while length of stay was analyzed using negative binomial regression. All models were adjusted for clinical and procedural confounders. Results Of 1,596 patients (mean age 77±9 years, 41% female), the optimal baseline AK-AVS threshold (0.62) was not significantly associated with 1-year mortality (Table), but predicted shorter length of stay (IRR 0.90, 95% CI 0.82-0.98, p=0.01). Following TAVR, 538 (34%) patients showed a decrease in AK-AVS score (mean decrease 21%)while 1,058 (66%) showed an increase (mean increase 23%). Baseline clinical and echocardiographic characteristics were similar between groups, though patients with increased scores had wider QRS duration (123±34 vs 119±32 ms, p=0.02). In unadjusted analyses, patients with increasing AK-AVS scores had higher rates of 1-year mortality (7.9% [83] vs 4.3% [23], p=0.009), permanent pacemaker implantation (15.1%[160] vs 9.1% [49], p=0.001), and length of stay (7 vs 5 days, p=0.002). After adjustment, increases in AK-AVS score remained significantly associated with higher 1-year mortality (OR 1.78, 95% CI 1.09-2.91, p=0.02), increased risk of pacemaker implantation (OR 1.68, 95% CI 1.18-2.40, p=0.004), and longer length of stay (IRR 1.17, 95% CI 1.06-1.28, p=0.001) (Table). Conclusion AI ECG-based aortic stenosis assessment provides prognostic information in TAVR recipients. While baseline scores primarily impact length of stay, post-procedural score changes may identify patients at higher risk for adverse outcomes.
No takes yet. Share an insight, caveat, or question.
Segar et al. (2025) conducted a cohort in Aortic stenosis post-TAVR (n=1,596). Increase in AI ECG-based aortic stenosis score (AK-AVS) post-TAVR vs. Decrease in AK-AVS score post-TAVR was evaluated on 1-year mortality (OR 1.78, 95% CI 1.09-2.91, p=0.02).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: