CMS Coverage Plan Keeps Volume Requirements for Existing TAVR Programs but Lowers Bar for New Centers

The proposed national coverage determination is already drawing a mix of strong reactions seen in previous debates over volume, access, and quality.

CMS Coverage Plan Keeps Volume Requirements for Existing TAVR Programs but Lowers Bar for New Centers

(UPDATED) An eagerly awaited and much debated Centers for Medicare & Medicaid Services (CMS) proposal for TAVR reimbursement has finally been released. While the national coverage determination (NCD) relaxes some requirements, particularly for hospitals looking to get into the game, CMS wants to keep the relatively strict volume requirements in place for centers looking to maintain an existing program.

As reported by TCTMD, the proposed update to the original NCD for TAVR has been a protracted process, drawing mixed opinions both for and against the need for hospitals and practitioners to maintain a certain volume of procedures while guaranteeing adequate access to care for all patients.

In the plan released late yesterday, CMS proposed that a hospital without TAVR experience looking to start a program must meet all of the following requirements to be reimbursed for the procedure:

  • 50 open-heart surgeries in the year prior to starting the TAVR program
  • ≥ 20 aortic valve-related procedures in the 2 years prior to starting the TAVR program
  • ≥ 2 physicians with cardiac surgery privileges
  • ≥ 1 more physician with interventional cardiology privileges
  • ≥ 300 PCIs annually

Moreover, the cardiovascular surgeon on the heart team is required to have completed 100 or more open-heart surgeries, of which 25 or more are related to the aortic valve. For the interventional cardiologist, he or she must have done 100 or more structural heart disease procedures (or 30 left-sided structural procedures per year), as well as device-specific training by the valve manufacturers.

For hospitals with existing TAVR programs, CMS has determined that the following requirements are maintained as part of the NCD:

  • ≥ 50 AVRs (TAVR or SAVR) per year, including ≥ 20 TAVR procedures in the year prior, or;
  • ≥ 100 AVRs (TAVR or SAVR) every 2 years, including ≥ 40 TAVR procedures in the 2 years prior, and;
  • ≥ 2 physicians with cardiac surgery privileges, and;
  • ≥ 1 physician with interventional cardiology privileges, and;
  • ≥ 300 PCIs annually

One notable change will satisfy practitioners who’ve vocally criticized the need for a two-surgeon sign-off before a patient can be considered for TAVR. In an update from the current NCD, which was issued in 2012 when TAVR was just emerging in clinical practice, a single surgeon is now all that’s needed to evaluate a patient for SAVR, TAVR, or palliative therapy.

CMS also wants valve-center heart teams and hospitals to participate in a prospective, national, audited registry that tracks clinical outcomes, including stroke, all-cause mortality, transient ischemic attack, major vascular events, acute kidney injury, repeat valve procedures, new permanent pacemaker implantation, and quality of life.

Is Volume a Suitable Metric for Quality?

Peter Pelikan, MD (Providence Saint John’s Health Center, Santa Monica, CA), one of the authors of the letter that spurred CMS to reexamine the requirements, conceded that the proposed NCD is less restrictive than the 2012 coverage decision, but said he believes volume is no longer a suitable metric for the assessment of TAVR quality.

In the early days with TAVR, “there was clearly a learning curve,” but that’s no longer the case with the latest-generation devices, he said. To support his case, Pelikan pointed to recently published data with Sapien S3 (Edwards Lifesciences) valve that failed to find any link between procedural volume and outcomes, even early in a center’s clinical experience.

In addition to the latest-generation valves being more deliverable and working better, “there is a massive infrastructure in this country, if not around the world, for colleagues to help each other learn,” said Pelikan. Not only that, he argued, but young interventional cardiologists hired today have spent at least 2 years in training programs learning how to do the procedure.

“I wish what [CMS] had done was look at quality instead of volume,” said Pelikan.

Recently, the PARTNER 3 and the Evolut TAVR in Low-Risk Patients trials showed that the procedure was equivalent to, if not better than, surgery in the treatment of low-risk patients with aortic stenosis. As a result, said Pelikan, TAVR is set to explode and the volume requirements could have an adverse impact on patient access.

“We’re going to have patients on waiting lists,” said Pelikan. “I can tell you that in Los Angeles, the patients who are poor and in need are not going be attended to because their hospital in their part of town is not going to make these numbers.”  

For Joseph Bavaria, MD (Penn Medicine, Philadelphia, PA), a co-chair of the updated American Association for Thoracic Surgery/American College of Cardiology/Society for Cardiovascular Angiography and Interventions/Society of Thoracic Surgeons expert consensus statement on operator and institutional requirements for TAVR, the proposed NCD is “reasonable” and consistent with their recommendations. He believes CMS did a good job navigating the complex interplay between access to care and the known relationship between volume and clinical outcomes.

Overall, the proposed volume requirements “walk the fine line between too much and too little,” he told TCTMD. 

He pointed out that for maintaining a TAVR program, the CMS proposal is roughly similar to the old NCD and a little less stringent than in the past for hospitals looking to start a program. He defended the mandatory volume requirements not only because of the volume-outcome relationship, but because a minimum volume threshold is necessary to assess quality. Without a sufficient number of procedures, any attempt to directly measure clinical outcomes at TAVR-capable centers would be fruitless, said Bavaria.

David Cohen, MD (University of Missouri-Kansas City School of Medicine, MO), pointed out that it can be challenging to assess quality of TAVR care at low-volume centers, noting it’s nearly impossible to distinguish between excellent clinical results or good luck in such settings. On the other hand, in-hospital and 30-day mortality rates with TAVR are now low enough where even more than one patient death each year would adversely affect outcome-based quality metrics for low-volume centers.      

Regarding the requirement of 50 or more AVRs per year, Cohen said it’s not a “perfect number to ensure top-tier quality because I still believe that practice makes perfect in almost anything that you do,” but it’s a reasonable bar for centers to meet to maintain their TAVR program and to report their clinical results. Nonetheless, “if I needed a TAVR, or my parents needed a TAVR, I would find a place that does more procedures than the minimum requirement,” said Cohen.

The bottom line, he said, is that hospitals initiating, or even maintaining TAVR as a treatment option need to be committed to their programs. “As those of us who have been doing it for a decade know, everything is wonderful until it isn’t.” Only when things go south do physicians and their teams find out if their systems and skills are as good as they believed them to be, said Cohen.

Finally, Cohen praised the CMS decision to combine both TAVR and SAVR as a volume metric for centers to maintain their existing programs. With the growing evidence base for TAVR, it doesn’t make sense to mandate centers perform a given number of SAVRs each year (or every 2 years). “It’s pretty clear that most places that do TAVR already are doing more than surgical aortic valve replacement,” he said. “Forcing smaller programs to do things that might not be in their patient’s best interests just to maintain a TAVR program doesn’t make a lot of sense.”

PCI to Maintain TAVR Status

For Pelikan, however, one of the big sticking points of the proposed NCD is the number of annual PCI procedures needed to start or maintain the structural heart program. “There’s no data correlating PCI volume with TAVR quality,” he said. “This is a number picked out of a hat. You have to wonder if this is going to serve the public well.”

For example, Pelikan estimated that among the approximately 380 hospitals in California, roughly 90% would be excluded from performing TAVR based on the PCI requirements. Of the approximately 170 California hospitals who performed PCI, just one in four exceeded 300 or more procedures in 2017, according to data from Office of Statewide Health Planning and Development. The hospitals excluded from TAVR based on the PCI requirements include county hospitals caring for the indigent, community hospitals, and prominent teaching hospitals, said Pelikan.

The unintended effect of the PCI requirement could force hospitals to increase their PCI volume to maintain their TAVR programs. In a worst-case scenario, some patients might be inappropriately treated with PCI when medical therapy or CABG surgery is the better option. To TCTMD, Pelikan stressed he’s in favor of “high-quality care” and not advocating every hospital start a TAVR program, but that the main issue is having volume stand in as a surrogate for quality.

“Measuring quality is difficult,” he said. “Measuring volume is easy.”

To TCTMD, Bavaria noted that the 2012 NCD required hospitals starting a TAVR program to perform 1,000 or more catheterizations, including more than 400 PCIs, so that particular requirement has been reduced. At the Medicare Evidence Development and Coverage Advisory Committee (MEDCAC), which was held last year and allowed experts to weigh in on whether PCI, SAVR, and TAVR volumes influence TAVR outcomes, some had suggested that 250 PCIs annually might be reasonable.

“I think the main thing with PCI, what CMS is trying to do is make sure these TAVR cases are done at ‘functional’ cardiac surgery and interventional cardiology programs,” said Bavaria. “If you’re not doing a certain amount of PCIs, then the robustness of your interventional cardiology program is probably not that great.” A similar logic applies to setting a requirement for aortic valve replacement in general, he added. “You might not have a robust interventional cardiology program, or you might not have a reasonable cardiac surgery program, and really if you want reasonable outcomes you have to have both.”

To TCTMD, Paul Teirstein, MD (Scripps Clinic, La Jolla, CA), said the proposed NCD is an improvement from the 2012 iteration, but that he would make two recommendations. First, for patients undergoing SAVR, he would like to see it mandated that the patient also be assessed by an interventional cardiologist to evaluate the patient’s suitability for surgery, TAVR, or palliative care.

Secondly, he doesn’t think it’s necessary for both an interventional cardiologist and surgeon to jointly participate in the procedure. “Physician scheduling is a huge challenge for busy TAVR programs,” said Teirstein in an email. “Many patients sit in the hospital for many extra days or wait weeks at home for the availability of both a surgeon and an interventional cardiologist.”

Instead, the requirement should be for a well-trained, experienced physician—either a surgeon or interventional cardiologist—to perform the procedure. “Requiring both an interventional cardiologist and a surgeon is redundant, and vastly increases costs due to . . . patients waiting for both physicians to be available and two physician fees instead of one,” he said. 

The new NCD is not yet finalized, and CMS is now seeking public comments on their recommendations, which can be submitted through their website. CMS will post the final NCD no later than 60 days after the 30-day public commenting period closes.  

Michael O’Riordan is the Managing Editor for TCTMD. He completed his undergraduate degrees at Queen’s University in Kingston, ON, and…

Read Full Bio
  • Pelikan reports no conflicts of interest.