The Learning Curves for Multigenerational Transfemoral Transcatheter Aortic Valve Replacement-A Real World Analysis

Background: There will be an increasing demand for TAVR in the years to come. Defining the learning curve for this procedure carries implications for centers looking to begin TAVR programs. There has yet to be a single center nonPARTNER-I trial TAVR learning curve in the literature. The purpose of this study is to elucidate the efficiency and mastery of one heart team at a single nonPARTNER-I trial institution using consecutive generations of transcatheter aortic valve replacement (TAVR) devices. Methods: In a retrospective, observational, study from 2012-2017, we reviewed data on 920 patients who underwent TAVR at our institution. Excluded were the self-expanding TAVR valves and non-transfemoral access TAVRs. We used three complementary methods (CUSUM analysis, penalized B-splines regression, and generalized linear model) on the remaining 616 patients to define the learning curve for the first 100 consecutive patients using the first (Cohort 1), second (Cohort 2) and third (Cohort 3) generation balloon-expandable TAVR valves. Findings: In Cohort 1, the median procedure time was 178.5 minutes. Efficiency was achieved at case 40 and mastery at case 60. The first 100 consecutive patients in Cohort 2 had a median procedure time of 73 minutes. The first 100 consecutive patients in Cohort 3 had a median procedure time of 92 minutes. There was no additional level of efficiency or mastery after the first generation TAVR valves. Conclusion: In conclusion, given the growing number of eligible patients and expanding indications for TAVR, there will be an increased demand for TAVR. Early experience using the first generation TAVR device carried over to subsequent generations of TAVR devices.