Transcatheter Aortic Valve Replacement: Assessment of the Learning Curve Based on the PARTNER Trial

Background: Transcatheter aortic-valve replacement (TAVR) is a new procedure for the treatment of severe aortic stenosis. The number of cases require to gain proficiency “learning curve” is unknown. The aim of this study is to assess the learning curve of the team involved in the performance of TAVR in the PARTNER trial. We hypothesized that a) a quantifiable learning curve exists for TAVR and b) process measures of procedural proficiency would improve with increasing case volumes Methods: 2,621 patients underwent TAVR with the SAPIEN heart valve via TF (N=1521) or TA (N=1100) delivery routes between April 2007 and January 2012. The study was divided into 2 time domains for this analysis, 1st half (2007 - 2009) and 2nd half (2010 - 2012). The Shapes of the learning curves were assessed using a semi-parametric mixed effects model. Results: Figure 1A details the outcomes of the technical performance variables. Using the date of implant, there appeared to be a significant downward trend in procedure times (Figure 1B), fluoroscopy times and contrast volume. (Figure 1C and D) The improvement in these parameters was particularly striking in TF cases p < 0.005. (Figure Presented) Conclusions: Our results demonstrate that in the PARTNER trial there is an important TAVR learning curve with significant improvement in procedural and performance variables over time. Although the differences in primary operator skill sets between TA and TF approaches may explain some of the observed heterogeneity in learning curves, a critical volume of cases appears necessary to become proficient in the safe performance of TAVR, particularly via the TF route.