Elective PCI: 12% of Cases Found Inappropriate

NEW ORLEANS – About 12% of the more than 140,000 Americans who underwent elective coronary artery stenting during 2009-2010 had an inappropriate procedure, based on an analysis of data from a national registry maintained by the American College of Cardiology.

In contrast, the rate of inappropriate procedures was 1% in the larger group of more than 355,000 patients who had an acute need for percutaneous coronary interventions (PCI) during the period studied, Dr. Paul S. Chan said at the annual meeting of the ACC.

The analysis also showed a striking hospital-to-hospital variation in the rate of elective PCI cases flagged as inappropriate. About 25% of the hospitals doing elective cases had a rate below 6%; another quarter had a rate of 17% or higher. Yet some hospitals had inappropriate rates that exceeded 30%.

Starting in May 2011, the ACC will start reporting data from this analysis to each of the more than 1,000 participating U.S. hospitals. By carefully reviewing cases that have been flagged as inappropriate, it is hoped that hospitals will learn from their mistakes and drive down the inappropriate rate, especially for elective PCIs, said Dr. Chan of Saint Luke’s Hospital in Kansas City, Mo.

The ACC collects the PCI data through the CathPCI portion of its National Cardiovascular Data Registry (NCDR). Dr. Chan and his associates rated each procedure they reviewed as appropriate, inappropriate, or uncertain based on comprehensive criteria established by an ACC expert panel (J. Am. Coll. Card. 2009;53:530-53).

"About 25% of the interventional cases in my practice would be categorized as inappropriate," commented Dr. Edward J. McNulty, an interventional cardiologist at Kaiser Permanente in San Francisco. "I get most of my patients from surgeons or other interventional cardiologists who feel the patients are too high risk. I do a lot of left main and multivessel PCI. There is no way that I can explain on the NCDR form why I consider these patients appropriate. The NCDR data don’t allow you to appreciate nuances. The criteria tend to penalize physicians who deviate from the average [by performing] complex cases."

Additionally, the way patients’ drug use gets recorded on the day of hospitalization is problematic, as patients may stop a chronically used drug on the day before entering the hospital and be erroneously recorded as not being on a therapy. Another issue is misclassification of symptoms, such as a patient who perceives ischemic chest pain as shortness of breath.

"I believe that inappropriate PCIs occur, and these results can certainly show signals. But within the ‘inappropriate’ procedures are some cases with mitigating circumstances," Dr. McNulty said in an interview.

The ACC expert panel published appropriateness criteria for 198 different clinical scenarios based on six separate clinical elements in February 2009. They applied the criteria to 500,154 U.S. cases treated with PCI during July 2009 through the end of September 2010. Inappropriate cases were defined as situations where expected negative consequences exceeded expected benefits.

PCI performed for an acute problem (such as high-risk unstable angina or MI) occurred in 71% of the cases. In this group, 99% of the cases were rated as appropriate, 1% as inappropriate, and fewer than 1% as uncertain. The remaining 29% of PCI procedures occurred in elective cases, of which 50% were rated as appropriate, 12% as inappropriate, and 38% as uncertain, Dr. Chan reported.

The three most common reasons for rating a case as inappropriate included patients with no ischemia, patients with mild ischemia, and asymptomatic patients, he said in an interview.

To address this issue, the ACC should develop a "real-time decision aid" to encourage interventionalists to "take a step back" during a catheterization to review a patient’s history and make a more informed decision on the need for PCI. The interventional cardiologist should review the degree of symptoms a patient has had and the evidence of ischemia, and take those findings into account when deciding whether the patient needs PCI.

Dr. Chan and Dr. McNulty said they had no disclosures.