Overall periprocedural death and stroke rates have declined in patients undergoing carotid artery stenting (CAS). Long-term, CAS may be on even footing with endarterectomy and possibly have an edge for event-free survival. On a less-than-positive note, 30-day mortality rates vary by hospital. Given these results, will payers revisit reimbursement policies for CAS?
The current Centers for Medicare & Medicaid Services (CMS) policy limits Medicare reimbursement for CAS to certified facilities; the process requires hospitals to meet minimum standards―physician training, facility device inventory and support and submission of outcomes data.
CAS advocates charge that scientific evidence calls for an expanded policy, but they may not get their wish. Although stenting has lived up to its promise, reimbursement appears unlikely for a variety of reasons. The specter of the sequestration continues to haunt CMS. Compounding the problem is the issue of potential abuse, which could increase under a more favorable reimbursement policy. It may be more likely that CAS is wrapped into general payment reform aligned with the move toward accountable care organizations (ACOs) and bundled payment.
As far as hospital-by-hospital mortality rates, CMS could tap recent data to audit hospitals identified as underperformers.
CAS data at a glance
If CMS were to re-evaluate the National Coverage Determination (NCD), it would need to consider new data relevant to the decision. These data include the CREST dataset (N Engl J Med 2010;363:11-23), which demonstrated comparable four-year outcomes between stenting and endarterectomy and anticipated results from ACT-1. Stent-makers Abbott and Cortis also have gathered prospective post-market data, demonstrating good outcomes with CAS, including in high-surgical risk patients.
SAPPHIRE, a worldwide trial of 20,000 patients, will likely show similar excellent results, predicts William A. Gray, MD, director of endovascular services at Columbia University Medical Center/NewYork-Presbyterian Hospital in New York City. “This is a real-world slice of carotid stenting that includes hundreds of operators in 800 sites. It isn’t randomized, but it’s meaningful.” Overall, multiple studies and clinical experience show outcomes have improved over time; complication rates have been halved over the last decade.
Yet, Gray dubs the mostly supportive data “problematic.” That’s because CMS is likely to avoid revisiting the CAS question.
One option had been coverage with evidence development, akin to the transcatheter aortic valve replacement (TAVR) policy. A similar multispecialty initiative that included vascular surgeons, cardiologists, neurosurgeons, CMS, the FDA and stent-makers had gathered momentum and even generated enthusiasm among CMS representatives, says Gray.
“Unfortunately, when it came time to propose coverage with evidence development, CMS told us it did not have the time or manpower to consider it, blaming the sequestration and overall reduced staffing,” Gray says.
But there may be more to the story than the sequestration. CMS data show approximately 90 percent of carotid revascularizations in the U.S. are performed on asymptomatic patients. Revascularization is recommended for any patient with stenosis greater than 60 to 70 percent. However, research is starting to suggest that the percent stenosis is not the culprit in stroke; the nature of the plaque―whether it’s vulnerable or not―may be the key factor.
Skeptics charge that some of these asymptomatic patients may not need any treatment. CMS approval could open the floodgates to abuse as a wide swath of specialists from vascular surgeons to neurosurgeons to cardiologists float a stent and perform the procedure.
CREST 2 should provide the answer to the riddle, says William H. Brooks, MD, chairman of the neuroscience center at Baptist Health in Lexington, Ky. The trial will randomize asymptomatic patients to best medical therapy, endarterectomy or stenting.
Imaging, performance & expertise
The gold standard to determine whether or not a patient needs revascularization would be a noninvasive way to assess plaque composition and determine whether or not it is vulnerable.
Imaging may be a way to better identify appropriate patient candidates prior to procedures. Baptist Heart and Vascular Institute in Lexington, has deployed optical coherence tomography (OCT) pre-therapy to determine whether a plaque is stable or unstable.
OCT, a high-resolution, light-based imaging technique, could overcome shortcomings of MR, CT and intravascular ultrasound in identifying vulnerable plaque. Specifically, OCT enables visualization of vulnerable plaque features, including thin-cap fibroatheroma, intraluminal thrombus, calcified nodules and vascular inflammation.
Results of a recent retrospective study of 53 patients (J Am Coll Cardiol Intv 2014;7:674-84) identified no association between stenosis severity and the presence of complicated plaque features, and concluded that carotid OCT offers “the potential to alter the understanding and treatment of carotid artery disease.”
However, OCT is invasive and not widely deployed. Large-scale prospective trials of OCT are needed to determine its role and inform decisions about change in treatment strategies. Baptist Heart and Vascular Institute has attempted to find a company to sponsor a trial to get an investigational device exemption. So far, it had not yet secured a taker.
While payers, policy makers and physicians worry about the potential for abuse, others express concerns about expertise with CAS.
“What matters for outcomes?” asks Peter W. Groeneveld, MD, associate professor of medicine at Perelman School of Medicine at the University of Pennsylvania in Philadelphia. “Operator experience plays a role. Clearly, physicians learn to get better at doing these procedures [as they gain experience].”
And there, according to some, is the rub.
Existing registries may not be open enough to encourage operators to participate. “When appropriate patients come to us, we may not be able to stent them,” says Gray. “Volumes have fallen, which has led to a concern about the expertise we can offer to patients both inside and outside of a trial.” Meanwhile, endarterectomy is unrestricted, helping surgeons maintain their expertise.
Are those concerns justified? And what about poor performers? Is there a role for CMS in identifying poor performers?
CMS’ current limited use policy has not restricted access among underserved populations, according to Groeneveld et al (Health Affairs 2011;30: 312-321). Groeneveld explains that there is no definitive evidence demonstrating that patients need to travel to a large hospital for CAS; local community hospitals can provide good service. The next task for CMS may be to determine if the initial criteria used to authorize hospitals to receive payment for CAS suffices to identify underperformers.
“It’s not a bad list; it just isn’t based on data,” explains Groeneveld. Recent data show that a handful of hospitals have worse outcomes, even when researchers controlled for severity and differences in patient risk (Circ Cardiovasc Qual Outcomes 2014;7:574-578).
So what makes a good performer? Operator experience is important, but are there other factors? CAS may or may not follow the typical observed relationship between volume and outcomes. It’s difficult to tease out, says Groeneveld. “It can be harder for a small hospital to fail a quality test than a large one.” A small hospital with an outlying 10 percent death rate may perform only five cases in one year and have a good year. Or it may see a different patient population that is less complex with fewer comorbidities.
Although CMS may be understaffed and lack the resources to audit 500 or 1,000 hospitals, it could manage an audit of the 15 or so hospitals with poor performers. The process would likely entail a deeper dive into registry data and a chart review to determine if patients died due to potentially inappropriate care, and if the trend is worrisome enough to suspend CAS reimbursement at a given hospital.
ACOs carve out a role?
However, ultimately it may not be CMS that plays this role. It could be delegated to a local institution or ACO, with some hospitals having to demonstrate that they can perform stenting as a condition of joining an ACO.
“If the data are there to identify hospitals with suboptimal outcomes, why would we not [use them]?” asks Groeneveld. “As we try to get patients and payers to think about where to go for procedures, outcomes information needs to be accessible. Having cardiologists on staff and being Joint Commission-approved isn’t enough for Medicare to pay $45,000 for procedures.” This model, continues Groeneveld, could extend beyond carotid stenting to include TAVR, left ventricle assist devices, cardiac transplants and any high-dollar cardiovascular intervention.
“The likelihood is that CMS will have the greatest leverage by requiring hospitals to pass a quality bar,” he says. Hospitals, in turn, will have to determine how to get the best surgeons and interventionalists through clinical privileges. Or, in an ACO-driven environment, the ACO will scrutinize performance of individual providers, ousting underperforming physicians because the entire group loses money with one underperformer.
The other piece of the equation is the move away from fee-for-service and toward bundled payment over the next five to 10 years. Essentially, hospitals will be allotted a bundled payment for management of a patient with advanced carotid disease. The hospital will have to determine how to optimize outcomes. As Medicare moves toward this model, it will check patient outcomes, and either cut bundled payments or divert patients away from underperformers, predicts Groeneveld.
Ultimately, it appears unlikely that good results on CAS outcomes will prompt hoped for reimbursement changes. The more likely path may be bundled payments, applications of data and possibly, additional research into effective ways to identify vulnerable plaque prior to carotid revascularization.