Hopes and Hype Await New Coronary Revascularization Appropriate Use Criteria in 2016

A new version of the appropriate use criteria (AUC) for coronary revascularization is expected within months, but whether this latest update will address concerns that have dogged these documents since their inception remains an open question. Top of mind for clinicians tackling the “scenarios” at the heart of the AUC are whether specific clinical situations will be upgraded in the new version and whether reimbursement and performance pressures can be uncoupled from the recommendations set forth in the new document.

“One of the biggest challenges” of the AUC concept—and one that it may be impossible to overcome—lies with the term “appropriate,” Ajay J. Kirtane, MD (Columbia University Medical Center, New York, NY), said to TCTMD. “So if something’s not rated as appropriate, colloquially speaking, you view that to be inappropriate. But the reality is there may be things that are [clinically] appropriate but may not be rated ‘appropriate’. And that’s why there’s so much confusion in this area.”

The initial aim of the first coronary revascularization AUC, issued in 2009, was to provide a comprehensive set of clinical scenarios that cardiologists could consult in choosing a course of action for a given patient. Treatment options for each scenario were categorized as appropriate, uncertain, or inappropriate. As a number of critics pointed out at the time, the AUC seemed swiftly outdated, seemingly reflecting a bygone era of cardiology—one without the SYNTAX, FREEDOM, or STICH trial results.

A 2012 update incorporated new data on how best to treat ACS, multivessel disease, and left main disease, but in the minds of many clinicians, did not go far enough.

Now, with the first update in 4 years pending, TCTMD spoke with several experts to find out what they hope to see with the newest take on AUC and what “appropriateness” might mean in years to come.

What’s Coming Down the Pike

Gregory J. Dehmer, MD (Baylor Scott & White Health, Central Texas, Temple, TX), who serves on the writing committee, said that the key task of this group is to develop the individual scenarios, such that the right questions can be asked when the appropriateness of each is debated and voted on by the technical panel. After voting, the document proceeds to all the participating professional organizations—including the American College of Cardiology, American Heart Association, and Society for Cardiovascular Angiography and Interventions (SCAI), among others—for their feedback and approval. The new AUC are meant to be “in harmony with the latest literature” and current guidelines, he commented.

Manesh R. Patel, MD (Duke University Medical Center, Durham, NC), who has chaired the writing group of all three rounds of the coronary revascularization AUC, acknowledged to TCTMD that while the document evolved between 2009 and 2012, gaps remained. “I think a lot of our colleagues are saying, look, there are [common scenarios and] indications in clinical practice that aren’t represented,” he observed.

The most salient gaps in the existing AUC, Patel said, relate to complete revascularization in multivessel disease, invasive physiological assessment such as fractional flow reserve, and what he calls “unique scenarios.” This group includes patients treated either before or after transplantation, TAVR patients, and “other emerging things that we take care of in cardiology,” he explained. “Those are all areas we want to focus on and try to at least get addressed.”

With newer data recently coming out on these key topics—including from the CvLPRIT, PRAMI, DANAMI3-PRIMULTI, and FAME 2 trials—“the overall practice of cardiology and revascularization continues to be informed. But I think most of what you’re hearing people wanting are scenarios that we don’t have a lot of evidence for and are common practice. [Clinicians] are getting pushback but . . . want to make sure they’re doing the best they can,” Patel said.

Complete revascularization in the setting of STEMI is a prime example of the disconnect that can occur without regular revision, Dehmer said. The 2012 AUC accurately reflect the guidelines that were in place at the time but do not capture the 2015 guideline update for STEMI. “So technically speaking, if you did a nonculprit vessel in a stable patient the way current appropriate use criteria are structured, it would be graded as inappropriate,” he explained. “But the latest guidelines give it a Class 2b recommendation. So practitioners are kind of caught in a little bit of a bind here.”

Dehmer noted one major shift that has already taken place: the original categories of “uncertain” and “inappropriate” were revised in 2013 when the ACC task force for AUC methodology decided a change was needed. Instead, the terms are now “may be appropriate” and “rarely appropriate,” he said, “recognizing the fact that even some things which are rarely appropriate, in an individual patient, might be the right thing to do.” This terminology will only now be officially applied to the updated revascularization AUC.

“In my mind, that helps a little bit, but it doesn’t help the fact that the opposite of appropriate, colloquially, is inappropriate,” Kirtane countered.

On the Wish Lists

Clinicians outside the AUC process, however, are outspoken about specific scenarios they believe are not only worthy of inclusion, but also appropriate.

“We have found the most glaring problems with AUC are certain clinical scenarios that are not captured as ‘appropriate’, but we believe to be quite appropriate,” Paul S. Teirstein, MD (Scripps Clinic, La Jolla, CA), commented. “The most common in our practice are PCI for significant coronary disease in the absence of clear-cut angina symptoms in patients who are being worked up for TAVR, patients with asymptomatic nonsustained ventricular tachycardia, and patients with newly diagnosed left ventricular systolic dysfunction. These are three examples of clinical scenarios I believe should be moved into an ‘appropriate’ category.”

TAVR is not mentioned at all in the 2012 AUC, nor is newly diagnosed left ventricular systolic dysfunction. Ventricular tachycardia appears only in passing; sustained ventricular tachycardia is listed as a high-risk feature for short-term risk of death or nonfatal MI among NSTE ACS patients.

For his part, Kishore J. Harjai, MD (Geisinger Wyoming Valley Medical Center, Wilkes-Barre, PA), cited complete revascularization in primary PCI for STEMI and the treatment of chronic total occlusions (CTOs) as two target areas for revision. He agreed that some drawbacks to the 2012 AUC stem from the fact that “clinical research in interventional cardiology moves so quickly that any set of guidelines that comes out every 2 to 3 years cannot keep up.”

According to the 2012 AUC, revascularization of a noninfarct-related artery during index hospitalization is classified as inappropriate in STEMI patients who have successful treatment of the culprit artery by primary PCI or fibrinolysis, normal LVEF, and are asymptomatic (no heart failure, evidence of recurrent or provocable ischemia, or unstable ventricular arrhythmias during index hospitalization). Yet, as mentioned above, “multiple recent studies signal possible benefit from nonculprit PCI and complete revascularization,” Harjai said.

Overall, the prior AUC “downgrade” PCI in CTOs “relative to non-CTOs and to surgical revascularization,” note Bradley H. Strauss, MD (Sunnybrook Health Sciences Center, Toronto, Canada), and colleagues in a 2014 review published in the Journal of the American College of Cardiology. Yet, “over the last few years,” they add, “evidence is increasing that CTO lesions have substantial impact and should be held to similar criteria as nonocclusive lesions.”

Harjai said he believes CTOs should be put on more equal footing with non-CTO lesions. “CTO PCI is technically challenging, has a steep learning curve and higher complication rate, and should only be attempted by trained experts. But this does not make CTO PCI less appropriate than PCI of a high-grade lesion with collaterals,” he stressed. “In regards to CTO PCI, the AUC panel has clearly confused the need for greater expertise with appropriateness of revascularization.”

Gregg W. Stone, MD (Columbia University Medical Center), added another idea to the mix, saying that “hopefully the AUC will expand to . . . more broadly recognize that the requirement to have to be on two anti-ischemic medications for stable coronary artery disease is not considered to be mandatory by many credible physicians.

“There’s very good evidence in patients with angina and documented ischemia that PCI is a valid alternative to medical therapy according to patient choice after a full discussion and informed consent by their physician,” he said in an interview. “There’s been no evidence of harm of PCI compared to medical therapy. And there’s evidence for better quality of life, more rapid resolution of symptoms, and a decrease in requirement for medical therapy.”

Useful Guidance but at a Cost

Everyone who spoke with TCTMD agreed that AUC are useful for practicing physicians at some level. Just how much they are benefiting patients, however—and at what price—remains an open question.

According to Harjai, who says he uses AUC for every case he does, the categories do help in clinical decision making, especially for “major” decisions such as whether someone should receive PCI. “There is clearly a subset of patients with stable angina who do very well with medical therapy,” he noted. “And then there are those patients who clearly need some form of revascularization, whether it is PCI or bypass surgery.”

The act of record keeping is itself important, he said, “because increasingly I suspect the appropriateness data will start becoming more and more transparent. If we don’t carefully document our rationale for doing tests, our tests may sometimes appear inappropriate because of bad documentation. A lot of times the appropriateness rating is incorrect because we just did not document accurately.” One “classic example,” Harjai noted, is when a high-risk stress test is improperly classified as low-risk.

Teirstein says Scripps is “supportive of AUC,” holding at least two AUC conferences per year to review cases coded as “rarely appropriate. “We find it to be an enlightening and helpful conference.”

While it’s possible that the AUC, and the potential of being called out for having made specific decisions, actually affect the behavior of individual doctors, Teirstein said he doesn’t think that’s an issue at his hospital. “I don’t think that happens very much,” he said. “For all that effort, I think it’s very rare.”

Still, AUC do appear to be influencing practice patterns on a broader scale. Teirstein referred to a 2015 paper in JAMA showing sharp decreases nationwide in the use of PCI for stable angina. “The publication was very honest in saying, ‘We’re not sure if this change is due to better care or just doctors and staff learning how to code better.’ Because certainly, that’s what happened at Scripps,” he reported. “We make sure that we document symptoms. We don’t leave that out. We don’t miss that. It’s very important. . . .  The AUC have done a lot to make sure our documentation is better. Is that worth a million dollars a year? I don’t know.”

The “million dollars” Teirstein is referring to covers the many fulltime staff required at Scripps for maintaining the AUC documentation and “pulling and entering” the data required by NCDR and TAVR registries. According to Teirstein, “The question you have to ask is, are there any patients helped by this? Does it result in more appropriate care? Are patients not being treated when they should not be treated and vice versa?”

Stone said he is “of a mixed mind,” as to whether adherence to AUC—and guidelines in general—leads to better results for patients. Much variability in practice exists across the United States, from “very good to very poor,” he pointed out. “So, in general, following guidelines moves the field forward and has led to overall improved outcomes; however, I’m a fierce opponent of ‘cookie-cutter medicine’, and there are many scenarios which are not black and white. Guidelines and AUC sometimes impose too forceful a direction, [with certain] controversial topics being settled one way or the other. My suggestion would be that they do a better job at identifying which areas are more unsettled—representing equipoise—and therefore should be left up to a decision between the patient and their physician.”

Indeed, the “patient preference aspect is nowhere within the AUC,” Kirtane emphasized. “We’re moving toward an era of shared decision making, so why is patient preference not represented?”

The Specter of Denied Reimbursement

One unanimous worry among the physicians interviewed is the threat that AUC will be used by payors to deny reimbursement. The issue has gained enough traction that SCAI recently published a position statement in Catheterization and Cardiovascular Interventions about coverage of PCI.

“Policymakers and payors must be good stewards of the insurance system, and are increasingly challenged to find innovative ways to curb expenditures. Thus, it is tempting for them to view the AUC as a professionally mandated tool for ‘cost-cutting’,” Lloyd W. Klein, MD (Rush Medical College, Chicago, IL), and colleagues assert in the statement. “SCAI and its members recognize the essential need for prudent cost management but are very concerned with this unanticipated and detrimental approach to coverage determinations.” Klein et al add, while “the AUC may be useful in helping to guide insurance coverage, the AUC classification should not be the solitary reason used to deny coverage.”

The reimbursement issue is “definitely there,” Teirstein agreed. “These are clinical scenarios that the originators of the AUC never thought would be hard and fast. . . . There are patients who have reasons for revascularization that aren’t captured by the AUC and issues that just aren’t thought about with the AUC.”

Dehmer emphasized that the AUC’s developers had envisioned them as a tool for examining practice patterns rather than as “a razor sharp look” at each case. “Now, I’m not naive,” he acknowledged. “I know that that’s going on. Insurance companies have seized on this with great vigor that they’re going to use this to decide [whether to pay]. That was never the intent of the AUC, but I do acknowledge that this has happened.”

As a “thought experiment,” Patel said he asks people what would happen if the AUC didn’t exist, and whether, in this scenario, insurers wouldn’t still be trying to find ways not to pay for procedures. “Would external forces try to control reimbursement in cardiovascular practice? The answer of course is yes,” he emphasized. “And how would they do it? They would easily do without our input, without us trying to participate in putting criteria forward.”

The implication is, with the AUC, interventional cardiologists have at least had a strong say in what procedures should and shouldn’t—typically—be performed.

But Patel, too, stressed AUC should not be used as the sole factor in payment decisions and should be as up to date as possible. “We will continue to face challenges with reimbursement pressures given the state of the health economy, and I think we have to be cognizant of that,” Patel said. “In many ways, our patients and our partners look to us to try to be stewards of that.”

Upcoding and Public Reporting

An oft-raised concern is whether AUC could lead to less-than-appropriate behavior when the white coat comes off and the charts come out.

Harjai described, for example, an unpublished analysis that involved interventional cardiologists reviewing each other’s cases. “We found that the operator disagreed with the reviewer’s assessment in three of four cases deemed to be ‘may be appropriate’ or ‘rarely appropriate’,” he said. “Most of the disagreements stemmed from the definition of stable versus unstable angina, maximal medical therapy, and risk stratification of noninvasive testing. As long as such ambiguities persist, the documentation for AUC ratings can be ‘gamed’.”

Another risk is that physicians may become overly cautious, shying away from cases that could actually help complex patients. This would become particularly problematic if AUC choices are made public, something many clinicians believe will ultimately happen in the push for greater transparency and public reporting.

“The pendulum can actually swing too far to the other side, because people are so afraid of being labeled inappropriate that they can stop doing things that are appropriate,” Harjai said. “That is a real possibility, and we have to be wary of something like that.”

Withholding care is not without consequence, Stone stressed. It could lead “to more patients who are developing side effects from medications or having angina and requiring limitations of exercise tolerance that could [instead have been] cured very quickly with a low-risk PCI procedure,” he explained.

According to Kirtane “a zero percent appropriate rate ought to also be a trigger for quality [audits]. Intuitively speaking, it stands to reason that if you’re doing no inappropriately rated cases, that you have a very, very high-quality cath lab. There are programs that advertise this. But what I would personally say is that, because we know these ratings not to be absolute, that to me suggests that people are either gaming the system or are avoiding these somewhat clinically nuanced scenarios.”

There are labs, he said, that have the AUC displayed on the wall, “and if a case comes out to be rated as inappropriate, they won’t let a physician do the case.” For the minority of patients who fit this description, perhaps because their case requires fine-tuned clinical judgment or is simply “poorly mapped,” Kirtane continued, “these poor patients are being held hostage to the AUC.”

Quality, Quantity, and Costs

Asked how he responds to critics of AUC, Dehmer said he believes that documents that help guide—not dictate—physician decision making will prove useful in an ever-evolving healthcare landscape.

“There are a lot of changes that are going to [affect] how medicine is practiced over the next 10 years,” he noted, adding that increasingly reimbursement will based more on quality than quantity. “We got rid of the SGR, but it was replaced by MACRA and MIPS.”

The SGR, known more formally as the Sustainable Growth Rate, was used by the Centers for Medicare & Medicaid Services between 1998 and 2015 to control physician fees. Its fee-for-service model was criticized for rewarding volume over quality care. MACRA, or Medicare Access and CHIP Reauthorization Act of 2015, spelled the end of the SGR and introduced two new approaches: MIPS, the Merit-Based Incentive Payment System, and APMs, shorthand for Alternative Payment Models. Both are meant to encourage better performance.

“Everybody talks about how much the US spends on medical care,” Dehmer noted, “and it’s clear that the cost curve has to be changed or we’ll go bankrupt. I think the value of the appropriate use criteria is that it has started that dialogue and nudged us in that direction.”

Citing the same JAMA paper as Teirstein, he said, “What we’re trying to do is start gradually cutting away at the lowest-hanging fruit [such as doing] PCIs on lesions of marginal hemodynamic importance in asymptomatic patients who are not on medical therapy. Trying to cut that down is going to help shift the cost curve a little bit.

“The other side of the coin, though, also deserves mention, so that we don’t go overboard to the point that patients who really do need revascularization are in some way being directed away from that because the appropriate use criteria don’t address their situation,” Dehmer continued. “AUC are about overuse but also about underuse.”

Kirtane suggested that instances of underuse could be captured by looking at patients who are rated as appropriate for revascularization—whether PCI or CABG—but end up receiving only medical therapy. Such patients are at risk of having poorer outcomes, he noted.

Since their launch, the coronary revascularization AUC have been gradually gaining acceptance among clinicians, Patel said. “[But] I don’t declare victory by any stretch. We have a lot of work to do.” He pointed out that, over this first decade of experience with the AUC concept, the process of their creation has also improved.

“We were hoping to do these to have an effect on practice and not just create more workload on people,” Patel noted, adding, “I believe we’re having more conversations around revascularization. I believe people are at least thinking about it more. Whether or not that all translates into improved outcomes, we don’t know yet.”