Predictors of cumulative healthcare costs associated with transcatheter aortic valve replacement in severe aortic stenosis

Background — There is wide variation in hospitalization costs for transcatheter aortic valve replacement (TAVR), suggesting inefficiency in care delivery. Our goal was to identify drivers of healthcare costs in TAVR.

Methods — Demographics, procedural details, in-hospital complications, and costs for all adults undergoing first-time TAVR from 2012-2016 in Ontario, Canada, were obtained through linkages of clinical/administrative databases. Total costs included were from initial referral to the first of either death or 1-year post-TAVR. Phase-based costing was performed to empirically estimate the presence, duration and cost/patient for each phase up to 1-year or death. Multivariable regression was used to identify drivers of cost accumulation per phase.

Results — We identified 2,009 first-time TAVR patients (mean age 81.7±7.6, 45.9% female and STS-score of 7.2±5.8). Phases of cost were identified with an early high cost period within 60-days of referral, a second phase from the procedure to 60-days, and a stable phase from 60-360 days post-procedure. The referral phase median cost was $4,527 (Interquartile range [IQR]: 1,708-12,594), the procedure to 60-days phase median cost was $29,518 (IQR: 24,842-40,279) and the post 60-day stable phase median cost was $6,053 (IQR: 3,320-17,048). Predictors of higher cost in the referral phase were in-hospital wait-location, dialysis dependence, and heart failure status. In the second (procedural) phase, predictors were non-transfemoral access, complications of stroke and pacemaker insertion. Predictors of higher cost in the third (stable) phase were predominantly non-modifiable, such as frailty.

Conclusions — This analysis shows that there are 3 distinct phases of cost accumulation from referral to post-TAVR with some potentially modifiable cost drivers in each phase.