Impact of a Standardized Conduction Management Algorithm on Hospital Length of Stay After Transcatheter Aortic Valve...

Conduction disturbances remain a major barrier to early discharge after transcatheter aortic valve replacement (TAVR), leading to prolonged monitoring and increased hospital length of stay. The ECG-based conduction-management algorithm proposed by Rodés-Cabau stratifies the risk of conduction disturbances after TAVR; however, its impact on hospital length of stay remains insufficiently studied. We aimed to evaluate whether systematic implementation of this algorithm is associated with reduced hospital length of stay after transfemoral TAVR. We conducted a single-center, nonrandomized before-after comparative study including consecutive patients undergoing transfemoral TAVR. Outcomes were compared between a historical control cohort managed without a standardized conduction algorithm and a prospective cohort managed according to the Rodés-Cabau criteria. The primary endpoint was postprocedural hospital length of stay. Multivariable linear regression was used to adjust for baseline characteristics, procedural factors, and postprocedural complications. Propensity score-based sensitivity analyses were performed. A total of 496 patients were included (124 in the control cohort and 372 in the intervention cohort). Mean postprocedural hospital length of stay was significantly shorter in the intervention cohort compared with controls (5.2 ± 2.7 vs 7.6 ± 3.2 days; p <0.001). Implementation of the Rodés-Cabau algorithm remained independently associated with reduced length of stay (adjusted β -1.90 days; 95% CI -2.54 to -1.27). Despite greater use of self-expandable valves, rates of high-grade conduction disturbances and permanent pacemaker implantation were lower in the intervention cohort. In conclusion, systematic ECG-guided conduction management after transfemoral TAVR is associated with a clinically meaningful reduction in hospital length of stay and improved discharge efficiency in real-world practice; however, given the nonrandomized before-after design, temporal confounding cannot be excluded.