This article is more than 5 years old. Information may no longer be current.
A Question of Appropriateness
Even in its revised form, the appropriate use criteria for revascularization remain controversial.
Originally introduced with the goal of elucidating the risks and benefits of CV procedures and serving as a guiding document for clinical practice, the appropriate use criteria for revascularization instead incited confusion, controversy and divisiveness among cardiologists.
Many interventional cardiologists felt that the use of the classifications “appropriate,” “uncertain” and “inappropriate” unfairly stigmatized some PCI procedures as being unethical or ill-advised. This not only led to ambiguity in daily practice, but also in refusals by insurers to reimburse certain procedures.
“The purpose for creating the appropriate use criteria was really to help doctors in clinical practice apply evidence-based data to their everyday practices,” said Christopher J. White, MD, director of the John Ochsner Heart and Vascular Institute in New Orleans. “It was intended to be a link between what we were writing in the guidelines and what we were doing on a daily basis, which is a laudable goal. The goal of the guidelines was to integrate evidence-based support into our practices, not to label doctors as criminals who are performing inappropriate procedures.”
In response to strong backlash from the interventional cardiology community, the American College of Cardiology has since revised the terminology with the categories “appropriate,” “may be appropriate” and “rarely appropriate.”
The new guidelines, which use a nine-point scale to address gradations of appropriateness, have been a welcome change to some clinicians, but to others, it still fails to clearly define the meaning of these categories.
“It’s not really defined what they mean by ‘rarely inappropriate’ or ‘appropriate,’” said Lloyd W. Klein, MD, professor of medicine at Rush Medical College in Chicago and Cardiology Today’s Intervention Editorial Board member. “For example, if you have a lot of data on a certain procedure, but the data are conflicting, that’s completely different than if there are no data on the subject. Those are two different kinds of ‘may be appropriate,’ and yet you’re stuck with voting on a scale in which the intermediate scores — 4, 5 and 6 — don’t convey these differences.”
The Delphi Method
The appropriate use criteria (AUC) were created through the Delphi Method, a methodology developed by public policy group the RAND Corporation.
“The Delphi Method is a methodology that allows a large group of people to work toward a consensus,” said White, who is also a board member of Cardiology Today’s Intervention. “So, to formulate the AUC consensus documents, everyone gets around a table and debates, and then they finally agree and establish ranges from low to medium to high. That’s where the terms ‘appropriate,’ ‘uncertain’ and ‘inappropriate’ came from. That terminology was not medical terminology; it came with the Delphi Method.”
According to White, the use of such non-medical classifications for medical procedures was partly to blame for the strong reactions that ensued.
“The ACC didn’t sit down and intend to term a medical procedure as ‘inappropriate’ — that was the scorecard that comes with the package of how to run a Delphi Method,” he said. “So that obviously injected significant emotion into it. Just because a procedure falls into the ‘inappropriate’ category doesn’t mean it would be an inappropriate procedure. That’s why we renamed the categories.”
White said the new categories are intended to reflect that the guidelines are not meant to be black and white.
“The idea was to explain that if something is rarely inappropriate, you could still justify it for a given reason, and we often do, but it’s rare — it doesn’t happen every day. That’s where the renaming takes a lot of the emotion out of the AUC.”
Manesh R. Patel
According to Manesh R. Patel, MD, director of interventional cardiology and catheterization labs at Duke University Health System, Durham, N.C., and one of the authors of the AUC documents, the original terms were part of an attempt to retain the language of the RAND methodology and to match the terminology used in similar manuscripts.
“The term ‘inappropriate’ was intended to describe a pattern of care and to measure against other groups caring for patients to improve quality,” Patel told Cardiology Today’s Intervention. “In our manuscript’s introduction and methods, ‘inappropriate’ was stated not to mean a procedure could not be done, just that likely more information or justification might be necessary for a specific case.”
Shades of Gray
The updated language of the AUC categorizes a procedure based on its score on the 9-point continuum, with the score range grouped as 1 to 3, 4 to 6, and 7 to 9. According to the published document, which appeared in the Journal of the American College of Cardiology in March 2013, the current categories are as follows:
- Appropriate Care, median score 7 to 9: Procedures or treatments that fall into this category are considered “an appropriate option for individual care plans, although not always necessary, depending on physician judgment and patient-specific preferences.”
- May be Appropriate Care, median score 4 to 6: Treatments or procedures in this classification are “at times, an appropriate option for management of patients in this population, due to variable evidence or agreement regarding the benefits/risks ratio, potential benefit based on practice experience in the absence of evidence; and/or variability in the population.” The document adds that the potential efficacy of treatments in this category need to be determined through consultation between the patient and the doctor based on “clinical variables and patient preference.”
- Rarely Appropriate Care, median score 1 to 3: Treatments and conditions in this category are “rarely an appropriate option for management of patients in this population due to a lack of a clear benefit/risk advantage; rarely an effective option for individual care plans.” The document adds that “exceptions should have documentation of the clinical reasons for proceeding with this care option.”
According to Klein, these classifications are an improvement over the original version, but are still imprecise and confusing.
“You really don’t need three levels of inappropriate, or three levels of appropriate,” he said. “It’s either appropriate or it’s not, and that’s part of the issue.”
Additionally, Klein said that the toned-down language of the updated AUC may serve to obfuscate its original purpose.
“The truth is, there are indications that really are inappropriate,” he said. “And so what we’re doing with this change is, we’re getting away from the whole point in the first place, which is to define situations where you shouldn’t be doing the procedure and defining the situations where you really should be doing the procedure. Now, everything is some shade of gray, so then what’s the point?”
Klein added that the nuances of the 9-point scoring system are subtle, almost to the point of meaninglessness.
“I thought the point of developing the AUC was to classify levels of appropriateness keyed to the evidence base, and so when an indication has one but not five supporting studies, to define the level of strength of the evidence. We would be better off with a guideline-based, evidence-based decision tree. Without that, we have nine categories of appropriateness that are not directly defined by the strength of evidence, so we’ve dumbed down the categorizations so much that they don’t really mean anything.”
Problems with Payers
According to White, the revised AUC will hopefully mitigate some of the negative repercussions wrought by the original version in terms of insurance payers.
“The AUC gets used by payers. For example, you might have a Blue Cross guy say, ‘I’m not going to pay for that procedure, it’s uncertain,’” he said. “And many doctors were really uncomfortable with payers who hijacked the criteria and used them to make it more difficult for us to practice medicine and take care of patients. That’s why doctors really revolted against the AUC.”
According to Klein, the use of the AUC by insurers as a concrete rule for guiding payment fails to take into account the issue of patient preference, which in some cases is strong enough to override the physician’s advice.
“When you have a concrete AUC that says something isn’t appropriate, and the patient decides not to go that way, the clear implication — and definitely the implication that the insurance industry wants to take — is that this is the doctor’s fault for not informing the patient better. And I’m not saying that never happens, but it’s often inaccurate for the insurers to assume that.”
Lloyd W. Klein
Klein said the AUC should not be used by payers to refuse payment without taking into consideration the uniqueness of each patient’s situation.
“Sometimes the patient has a strong feeling, and as the physician, it’s my job to advise them, but if they really want to do something, unless I think it’s criminal, it’s my judgment as to whether to go that way. And that has to be allowed,” he said. “To make a procedure not payable strictly because of the guidelines is really the height of absurdity.”
White said he thinks the new guidelines will make it more difficult for insurers to make this kind of blanket judgment of complicated situations.
“The changes have gone a long way toward lowering the tone and taking the emotions out of the whole situation,” he said. “And when you say ‘rarely appropriate’ instead of ‘inappropriate,’ it makes it harder for insurance companies to say no.”
Evidence-Based vs. Patient-Centered Care
White agreed with Klein that with each patient comes a unique, dynamic set of circumstances that cannot be reduced to a list of rules.
“There are times when the doctor and patient have discussed options, and the patient simply has a strong preference,” White said. “For example, say I have a patient who has an angiogram and really needs a bypass; but the patient says, ‘My grandfather and father and two uncles all died from a bypass — isn’t there any other way you can handle this?’ And I will tell him that I can put some stents in, but that the evidence suggests that it might not be the best thing for him. But the patient refuses to undergo a bypass. For that patient, the AUC says he needs a bypass, but the patient and doctor decide they can come to some other compromise. The AUC don’t really account for individual variations.”
Moreover, White said evidence-based medicine is important, but patient centeredness is another important mandate of good medicine.
Christopher J. White
“We really want to focus on what suits our patients, and their preferences have a lot to do with how we treat them,” he said. “We can’t just show them a randomized trial and say, ‘You have to do it this way.’ They have a lot to say about what risks they’re willing to take. The AUC are guides, not rules. That’s why the strong language didn’t really sit well with how clinical practice is conducted on a daily basis.”
According to Patel, the revised guidelines more accurately reflect patterns of care in clinical practice.
“The terms ‘uncertain’ and ‘inappropriate,’ both from the RAND methodology, were understandably not specific enough to convey the message around patterns of care,” he said. “Any time criteria are used to measure care, there is likely to be some controversy, but at this time, the AUC are used in many parts of clinical medicine, including cardiology, orthopedics and dermatology.”
White said he considers the new guidelines to be an improvement, and that the AUC are valuable overall.
“The AUC are a good thing,” he said. “When used properly, they have changed my practice. They have taught me better lesion selection; for example, on moderate lesions we use things like fractional flow reserve more frequently because the AUC taught us that. I think the good of the AUC far outweighs the bad; nobody should be throwing this baby out with the bathwater. We’re better off for the revision; it makes the AUC stronger and better. These are tools that help doctors who want to do the right thing.”
Klein maintains that the revision has simply not changed the AUC enough.
“I’m very worried that the new criteria have not, in a serious way, addressed these issues I’ve been talking about,” he said. “That really troubles me. The first time around, 5 years ago, I thought, ‘OK, this is the first time through, it’s not going to be perfect.’ But nothing has happened in the intervening 5 years, and that bothers me. The new criteria only changes a little bit of what was wrong. That’s discouraging.” – Jennifer Byrne