Author + information
Cost–Effective Analysis (CEA) from recent RCTs comparing TAVR and SAVR are useful but may not represent outcomes expected in a broad clinical practice. Markov decision analytic modeling may provide useful insights in this setting
We developed a Markov decision analytic lifetime model with Monte Carlo simulations in cohorts at “high risk” for SAVR. We informed the TAVR arm of our model using data from a recent meta–analysis of 16 registries using VARC outcomes (Genereux JACC 2012), and 4 contemporary surgical registries. Utilities for TAVR and SAVR were used from the Partner A data.
Probabilistic sensitivity analysis using Monte Carlo simulation of 10,000 samples estimated that TAVR was associated with a lifetime cost of $81,548 ± $1,240 and an effectiveness of 3.54 ± 0.07 QALYS. Patients undergoing SAVR accumulated $99,641± $3,126 and 3.45 ± 0.08 QALYS. TAVR was found to be a dominant strategy. The model was sensitive to variations in the probability of new dialysis in SAVR patients. If this rate fell below 3.3% in the SAVR arm, SAVR provided a greater net monetary benefit. Our estimated value with perfect information (EVPI) would save $2,919 and gain 0.2 QALYS per patient by using TAVR over SAVR.
In our decision model, TAVR was the dominant strategy over SAVR, though with a significant reduction in cost and small gain in QALYs. Prospective cost–analysis of these patients outside the setting of clinical trials will be beneficial for further evaluation of these therapies.
West, Room 2009
Monday, March 11, 2013, 8:45 a.m.–8:55 a.m.
Session Title: Transcatheter Aortic Valve Replacement (TAVR) I: Evolving Data
Abstract Category: 49. TCT@ACC–i2: Aortic Valve Disease
Presentation Number: 2910–7
- 2013 American College of Cardiology Foundation