NYMC Faculty Publications

Valve-in-Valve Transcatheter Implantation Versus Redo Surgical Aortic Valve Replacement

Journal Title

The American Journal of Cardiology

First Page


Last Page


Document Type


Publication Date




Second Department



Valve-in-valve (ViV) transcatheter aortic valve implantation (TAVI) for a failing prosthesis is an appealing alternative to redo surgical AVR. We utilized data from the US National Inpatient Sample for the period 2012 to 2016 to identify hospitalizations for either ViV-TAVI or redo-SAVR. The primary outcomes of interest were in-hospital adverse events composite outcome (comprising of mortality, myocardial infarction, stroke, or acute kidney injury) and all-cause mortality. We used propensity score matching to adjust for the baseline differences between ViV-TAVI and redo-SAVR cohorts. Survey techniques were employed to compare the 2 groups. Over 5 years, there has been a considerable increase in both interventions for prosthetic aortic valve failure, with significantly higher utilization of ViV-TAVI compared to redo-SAVR (p <0.01). Out of the 3,305 hospitalizations for prosthetic aortic valve failure, 1,420 in matched pairs underwent either ViV-TAVI (n = 710) or redo-SAVR (n = 710). ViV-TAVI was associated with lower in-hospital composite adverse outcomes (14.1% vs 25.4%, p = 0.018), and numerically lower but statistically insignificant mortality (<1.0% vs 5.2%; p = 0.06). ViV-TAVI was associated with a decreased length of hospitalization (mean 6.6 vs 9.7 days; p <0.01). In the matched cohort, postoperative bleeding and transfusions were significantly lower for ViV-TAVI compared with redo-SAVR (17.6% vs 31.0% and 12% vs 31% respectively, p <0.01 for both). Acute kidney injury, sepsis, permanent pacemaker implantation, and vascular complications, although numerically better, did not differ between 2 strategies. In conclusion, ViV-TAVI is associated with lower in-hospital MACE rates and reduced length of hospitalization compared with redo-SAVR.

This document is currently not available here.