By Jeffrey Zimmet, MD, PhD

Associate Professor of Medicine, University of California, San Francisco; Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center

Dr. Zimmet reports no financial relationships relevant to this field of study.

SYNOPSIS: This large-scale retrospective analysis of all balloon-expandable transcatheter aortic valve replacement valves revealed no apparent relationship between procedure volume and outcomes after an initial learning curve of 200 procedures.

SOURCE: Russo MJ, McCabe JM, Thourani VH, et al. Case volume and outcomes after TAVR with balloon-expandable prostheses: Insights from TVT Registry. J Am Coll Cardiol 2019;73:427-440.

Practice makes perfect, or so the saying goes. In the realm of complex cardiovascular interventions, many assume there is a relationship between procedural experience and patient outcomes. This so-called volume-outcome relationship (VOR) has been demonstrated previously in cardiac surgery and in percutaneous coronary intervention. Prior studies of transcatheter aortic valve replacement (TAVR) have produced conflicting results, confounded in part by the relative newness of the procedure, issues associated with the learning curve, and rapid advances in device technology.

Russo et al set out to first determine whether the VOR for TAVR applies beyond the initial learning curve. For the purposes of the analysis, the authors chose to focus only on the three generations of the balloon-expandable valves made by Edwards Lifesciences: the Sapien, Sapien XT, and Sapien 3 (S3). The Edwards balloon-expandable valve was the first to be approved for use in the United States and comprises the bulk of U.S. TAVR experience.

Looking first at all three Edwards balloon-expandable valves, the analysis showed a significant association between early case volume and 30-day mortality after adjustment for Society of Thoracic Surgery mortality score and site random effects. Outcomes leveled after a site’s first 55 cases, suggesting that this experience level produces a significant effect. However, after the first 200 cases, there was no longer a difference in composite 30-day mortality or stroke. Therefore, the authors denoted case #201 as the learning curve termination. After the initial learning curve, examination of data from all three balloon-expandable valves showed a significant association between major vascular complications and hospital implant frequency; however, no significant association remained with 30-day mortality or stroke.

The authors analyzed learning curve data separately for the current-generation balloon-expandable valve (the S3). They hypothesized that this valve, by virtue of device enhancements in combination with improvements in procedural techniques and proctoring, would result in a shortened learning curve for those relatively late adopters who began their TAVR experience with this platform. Interestingly, the analysis showed no trend for association between 30-day mortality, stroke, or major vascular complications and case sequence number. Only for centers with no prior TAVR experience was there a trend toward more vascular complications, and only at those centers performing two or fewer S3 implants/month (P = 0.052). In other words, even for sites with no prior experience with the balloon-expandable TAVR valves, this analysis did not demonstrate increased hazard in the early learning phase nor did it show a VOR.

The authors concluded that in U.S. cases involving all versions of the Edwards balloon-expandable TAVR valve, learning curve termination was observed after approximately 200 cases. Subsequent to that, the authors discerned no significant differences in hard outcomes related to ongoing procedure volume. With the S3 valve, they saw no learning curve at all and reported that centers should expect good outcomes “even during early case experience.”


Inherent in the volume-outcome relationship discussion is a tension between those who wish to limit procedures to high-volume centers and those who seek to expand access. What Russo et al found is that quality is not easily linked to a simple metric such as case volume. Expansion of the TAVR procedure has required training of new sites and new operators. The results of earlier studies suggested a significant learning curve resulting in poorer outcomes early in a site’s experience.

The lack of a clear learning curve in the analysis of the S3 device comes during a constellation of changes that have improved procedure safety. These include moves toward more transfemoral access and more conscious sedation, the establishment of specialized structural heart training programs, and intensive industry-sponsored proctorships. Movement of TAVR to the mainstream has promoted discussion of techniques and best practices, with effective knowledge transfer from more-experienced to less-experienced operators and hospitals. Improvements in the devices themselves likely play an important but less significant role. As TAVR evolves, the results of this study suggest that focusing on simple volume metrics as a surrogate for quality is likely to be misleading and may unnecessarily limit access.