P. K. Modi1, M. Oerline1, D. Sukul3, C. Ellimoottil1, V. B. Shahinian2, B. K. Hollenbeck1 1University Of Michigan,Urology,Ann Arbor, MI, USA 2University Of Michigan,Nephrology,Ann Arbor, MI, USA 3University Of Michigan,Cardiology,Ann Arbor, MI, USA
Introduction: While transcatheter aortic valve replacement (TAVR) was initially developed as a treatment option for patients ineligible for surgical aortic valve replacement (SAVR), its indications have expanded to include patients who would be candidates for surgery. As the use of TAVR continues to expand, it is essential to understand the economic impact of this substitution of TAVR for SAVR in real-world clinical practice. Therefore, we examined Medicare payments for TAVR and SAVR in episodes spanning from 90-days before surgery through 90-days after surgery.
Methods: We used a 20% national sample of fee-for-service Medicare beneficiaries who underwent TAVR or SAVR from 2012 through 2015. We used negative binomial regression models adjusted for age, race, sex, baseline health status (using Hierarchical Condition Categories risk score), socioeconomic class, and place of residence to estimate spending differences between TAVR and SAVR. We also examined the components of episodes to identify specific differences between the spending associated with these procedures. Finally, we assessed the effect of patient health status on the association between procedure type and payments.
Results: We identified 6,455 patients who underwent TAVR (34.3%) and 12,349 patients who underwent SAVR (65.7%) during the study period. The use of TAVR increased from 20.5% of all aortic valve replacements in 2012 to 46.7% in 2015. As TAVR replaced SAVR for the highest risk patients, the average baseline health status improved for both groups. Total adjusted TAVR episode payments were approximately 7% lower than payments for SAVR ($55,545 [95% confidence interval {95%CI} $54,643-56,446] vs $59,467 [95%CI $58,723-60,211], p<0.001). TAVR patients had higher pre-operative payments (Incidence rate ratio [IRR] 1.22 [95%CI 1.17-1.26], p<0.001), but lower payments during (IRR 0.96 [95%CI 0.94-0.98], p<0.001) and after the initial hospitalization for surgery (IRR 0.73 [95%CI 0.68-0.77], p<0.001). Episode payments increased with increasing comorbidity score, but this effect was greater for SAVR than TAVR.
Conclusion: After adjusting for patient factors, TAVR is associated with lower episode spending than SAVR due to savings during and after the initial hospitalization. As baseline health status of treated patients improves, the savings associated with TAVR relative to SAVR diminish.
