This article is more than 5 years old. Information may no longer be current.
Should the AUC be used to guide reimbursement?
Click Here to Manage Email Alerts
The New York State Medicaid Program recently announced that it would recoup payments from physicians and hospitals for PCI procedures considered to be “inappropriate,” according to the current appropriate use criteria for coronary revascularization.
This news was greeted with concern and frustration by New York interventionalists concerned that the appropriate use criteria (AUC) — which was written by eight professional medical societies, including the American College of Cardiology and the Society for Cardiovascular Angiography and Interventions — were being misused and that their patients might lose access to necessary care. What followed was a coordinated advocacy effort by SCAI and the ACC, with the goal of educating New York policymakers about the AUC. I was among the small group of physician advocates and staff who traveled to Albany to address this concern. Here is a summary of the key points we shared and my observations from having participated in the process:
Dmitriy N. Feldman
- Policymakers recognize the importance of the societal work that went into creating the AUC and see cardiologists as being at the forefront of developing quality improvement tools.
- The New York State policymakers believe that the AUC were created by a consensus of multispecialty experts and that inappropriate procedures are usually “unnecessary” and, therefore, should not be paid for with taxpayers’ dollars.
- The physicians stressed that in clinical practice, an individualized approach is necessary and that more than six AUC criteria go into the complex decision-making on whether to perform a procedure.
- We stressed that the AUC were never intended as a means to determine reimbursement and that such policy can limit access to procedures for underserved patients.
- As a result of the meeting, the payments will not be recouped immediately (before a medical review). Physicians/institutions will have an opportunity to explain and submit supporting data on why the procedure should be payable.
AUC have limitations
When our professional societies began developing the AUC, the intent was to assist physician decision-making, promote patient education regarding expected benefits from revascularization and to allow assessment of utilization patterns for revascularization procedures. I applaud the AUC Task Force for boiling down complex clinical scenarios to improve quality of care; however, we must be sure that clinicians, researchers and policymakers truly understand the inherent limitations of the AUC.
The current AUC are based on six clinical domains: patient stability varying from stable angina to STEMI; severity of symptoms and functional class; risk assessment based on non-invasive testing; amount of medical therapy; and coronary disease burden (one vessel to multivessel disease). There are many factors not considered in the AUC that are important in a clinician’s decision. For example, the AUC do not consider age, left ventricular function, lung or kidney comorbidities and, importantly, patient preference.
As estimated by the writing committee, if all clinical factors were used to develop the AUC, there would be well over 4,000 possible combinations and clinical scenarios. This would render the AUC cumbersome, thus there are many clinical scenarios that simply fall outside the scope of the AUC. Unfortunately, the limited number of AUC scenarios has led some policymakers to believe they can condense clinical decision-making into a series of right-or-wrong decisions, each informed by a “cookbook” approach. Clinicians understand the value of guidelines and the AUC, but also understand that care must be individualized for each patient.
‘Inappropriate PCI’ in New York
Two important questions have arisen concerning PCI in New York: Who are the patients whose PCI procedures are being deemed “inappropriate” or “rarely appropriate” by Medicaid reviewers, and how many procedures are suspect? An analysis of 2009-10 New York State data revealed that when the AUC were applied to data from the state database, 28% lacked sufficient data to determine appropriateness and, of the rated PCIs, 14.3% were deemed inappropriate. These findings are similar to ACC-National Cardiovascular Data Registry (NCDR) data for inappropriate PCIs, where 11.6% of elective cases were classified as inappropriate.
The majority of patients in New York State whose cases were rated inappropriate had one- or two-vessel CAD without proximal left anterior descending (LAD) artery disease and were receiving no or minimal anti-ischemic medical therapy. When we met with the policymakers in Albany, we explained that medical reviewers need to take into account all the extenuating circumstances (quality/severity of symptoms, specific stress test findings, ability to optimize medical therapy, etc.) when determining whether the procedure can truly be classified as inappropriate.
Let’s consider a hypothetical Medicaid 56-year-old patient with CCS class II angina and low-risk stress test findings who was taking one anti-anginal medication (Toprol XL, 200 mg) and had two-vessel disease on diagnostic angiography (90% mid-LAD artery and 80% proximal left circumflex artery stenoses). According to the AUC criteria, a PCI performed for this patient would be classified as inappropriate or rarely appropriate, primarily because of treatment with one, rather than two, anti-anginal agents. Based on New York’s policy, this case would have resulted in payment to the hospital and physician being rescinded.
What policymakers did not understand before we met with them was that, in daily clinical practice, PCI might have been clinically appropriate for this patient. Would a high dose of one anti-anginal medication not be considered as “optimal therapy” compared with two low-dose medications? What about patients in whom class II angina causes important limitations to their lifestyle and impacts their state of mind? What about patients who cannot tolerate or afford multiple medications? What if this patient is undergoing an evaluation prior to high-risk transplant surgery? And how reliable or specific are low-risk stress test findings in someone with this level of CAD?
How will AUC be used in the future?
As the situation in New York illustrates, there is potential for the AUC to be misused by policymakers who don’t fully understand the intention behind them. In my view, the following are the key issues and questions our profession will have to grapple with as issues arise:
- Should the AUC ever be used to determine the need for reimbursement? This was never the intent of the AUC, but, perhaps, given the need to reduce health care costs, the cardiology community should have realized that policymakers would see the “opportunity” to use AUC in this manner. I hope it is not too late to change the course of the New York State Medicaid Program or at least modify the program to have more clinician input into the evaluation of procedures as inappropriate. We should applaud the AUC Task force for changing the AUC terminology, since the implications of the word inappropriate are much different from the new phrase “rarely appropriate.” However, the original terms were adopted directly from the original RAND methodology.
- Has the precedent been set for the AUC being used punitively? When the AUC were first introduced, we were told that they would be a tool to guide clinical decisions and understand patterns of utilization. Will other specialties and their professional societies go forward with creating their own AUC, having learned from our experience that they may put themselves at risk?
- If the AUC will be used to determine reimbursement, shouldn’t a retrospective review identifying a pattern of potential inappropriate use be performed first? The AUC clearly state that recoupment of payments for PCIs “should not be applied retrospectively to cases completed before issuance of this report or documentation of centers/providers performing an unexpectedly high proportion of inappropriate cases as compared with their peers.” Therefore, the answer clearly is yes.
- What should be the frequency of inappropriate PCIs? As shown in the survey of about 500,000 cases in the NCDR, 11.6% of elective PCIs were classified as inappropriate. No doubt some were inappropriate, but many would likely have been considered appropriate had all of the clinical information been included in the AUC algorithm. The proposed New York Medicaid policy means the standard in New York is 0% of cases are allowed to be classified as inappropriate. Given the limitations of the current AUC, this holds New York’s interventional cardiologists to an unrealistic and unreachable standard. This is why retrospective quality reviews are essential to augment the intent of the AUC.
- How will all of this affect the Medicaid patient population? Will the steps that New York has taken potentially limit access to procedures for underserved patients who may lack the ability to advocate for themselves? Everyone is focused on overuse of PCI, but underuse of needed therapies is also an important issue. We should be reminded that one of the first publications using AUC was focused on the underuse of coronary revascularization procedures.
What do you think? Please share your thoughts on this important issue and the implications of the Medicaid reimbursement policy for inappropriate PCIs.
Finally, I encourage you to get involved. Email SCAI’s Advocacy Committee (wpowell@scai.org) to learn how you can help. We want to know what you think and hear your concerns.
References:
Chan PS. JAMA. 2011;306:53-61.
Hannan EL. J Am Coll Cardiol. 2012;59:1870-1876.
Laouri M. J Am Coll Cardiol. 1997;29:891-897.
Patel MR. J Am Coll Cardiol. 2012;59:857-881.