Loss of Independent Living in Patients Undergoing Transcatheter or Surgical Aortic Valve Replacement: A Retrospective Cohort Study

BACKGROUND: The recommendation for transcatheter aortic valve replacement (TAVR) or surgical aortic valve replacement (SAVR) in patients 65 to 80 years of age is equivocal, leaving patients with a difficult decision. We evaluated whether TAVR compared to SAVR is associated with reduced odds for loss of independent living in patients ≤65, 66 to 79, and ≥80 years of age. Further, we explored mechanisms of the association of TAVR and adverse discharge. METHODS: Adult patients undergoing TAVR or SAVR within a large academic medical system who lived independently before the procedure were included. A multivariable logistic regression model, adjusting for a priori defined confounders including patient demographics, preoperative comorbidities, and a risk score for adverse discharge after cardiac surgery, was used to assess the primary association. We tested the interaction of patient age with the association between aortic valve replacement (AVR) procedure and loss of independent living. We further assessed whether the primary association was mediated (ie, percentage of the association that can be attributed to the mediator) by the procedural duration as prespecified mediator. RESULTS: A total of 1751 patients (age median [quartiles; min–max], 76 [67, 84; 23–100]; sex, 56% female) were included. A total of 27% (222/812) of these patients undergoing SAVR and 20% (188/939) undergoing TAVR lost the ability to live independently. In our cohort, TAVR was associated with reduced odds for loss of independent living compared to SAVR (adjusted odds ratio [OR adj ] 0.19 [95% confidence interval {CI}, 0.14–0.26]; P < .001). This association was attenuated in patients ≤65 years of age (OR adj 0.63 [0.26–1.56]; P = .32) and between 66 and 79 years of age (OR adj 0.23 [0.15–0.35]; P < .001), and magnified in patients ≥80 years of age (OR adj 0.16 [0.10–0.25]; P < .001; P -for-interaction = .004). Among those >65 years of age, a shorter procedural duration mediated 50% (95% CI, 28–76; P < .001) of the beneficial association of TAVR and independent living. CONCLUSIONS: Patients >65 years of age undergoing TAVR compared to SAVR had reduced odds for loss of independent living. This association was partly mediated by shorter procedural duration. No association between AVR approach and the primary end point was found in patients ≤65 years of age.