Transcatheter and Surgical Aortic Valve Replacement in Dialysis Patients: A Propensity-Matched Comparison

Background. Transcatheter aortic valve replacement (TAVR) clinical trials in North America excluded patients on dialysis and, consequently, the outcomes of TAVR in dialysis-dependent patients remain unknown. Methods. All Medicare fee-for-service patients undergoing TAVR (n=5,005) or surgical aortic valve replacement (SAVR) (n=32,634) between January 1, 2011, and November 30, 2012, were identified using procedural codes collected by the Centers for Medicare & Medicaid Services. Dialysis status and comorbidities were identified using diagnosis codes present on arrival for TAVR hospitalization. Patients supported on dialysis who underwent TAVR (n=224) were compared with non-dialysis patients who underwent TAVR as well as a propensity-matched group of contemporaneous dialysis patients who underwent SAVR (n=194 pairs). Results. The TAVR patients on dialysis were younger than non-dialysis TAVR patients (79.2 years vs 84.1 years; p < 0.01) but had higher prevalence of comorbidities. Dialysis TAVR patients had increased mortality at 30 days (13% vs 6%, p < 0.01) and significantly worse survival by Kaplan-Meier analysis. Multivariable regression found dialysis to be independently associated with worse survival (hazard ratio, 1.73; 95% confidence interval, 1.33% to 2.25%, p < 0.01) in TAVR patients. Propensity-matched dialysis SAVR and dialysis TAVR patients had no significant differences in demographic or risk factors. Matched dialysis TAVR patients had shorter length of stay (6 interquartile range, 4 to 10] vs 10 [IQR 7 to 18] days; p < 0.01) and comparable survival. Conclusions. TAVR in dialysis patients is associated with decreased survival compared with non-dialysis patients; however, it is comparable with SAVR in high risk dialysis patients based on a propensity-matched comparison. (C) 2015 by The Society of Thoracic Surgeons