Most Acute PCIs Justified, But Nonacute Procedures Called into Question

Download this article's Factoid (PDF & PPT for Gold Subscribers)

A nationwide sample of percutaneous coronary intervention (PCI) in contemporary US practice has found that the choice to perform the procedure in acute cases was nearly always the right one. But the picture is less clear among patients who underwent PCI for nonacute indications, only half of whom received the treatment appropriately, according to a paper published in the July 6, 2011, issue of the Journal of the American Medical Association.

The results were previously presented in April 2011 at the American College of Cardiology (ACC) Scientific Session/i2 Summit in New Orleans, LA.

In 2009, 6 professional organizations—including the ACC and the Society for Cardiovascular Angiography and Interventions (SCAI)—jointly developed appropriate use criteria for coronary revascularization (Patel MR. J Am Coll Cardiol. 2009;53:530-553). To study how well actual practice meshed with these criteria, Paul S. Chan, MD, MSc, of Saint Luke’s Mid America Heart and Vascular Institute (Kansas City, MO), and colleagues prospectively assessed data from the National Cardiovascular Data Registry (NCDR) CathPCI Registry on patients undergoing PCI at 1,091 US hospitals between July 2009 and September 2010.

Using information collected at participating hospitals, the researchers categorized the PCI procedures according to their clinical indications. Each indication fell into 1 of  3 categories based on the 2009 appropriate use criteria:

  • Appropriate: PCI would likely improve patient health status (symptoms, function, or quality of life) or survival
  • Uncertain: More research, more patient information, or both were needed to further classify the particular clinical indication. For example, PCI was considered uncertain in STEMI patients who had successful thrombolysis and were stable
  • Inappropriate: PCI was unlikely to improve patient health status or survival

In all, the study cohort included 500,154 procedures, of which 71.1% were for acute indications including MI and high-risk unstable angina. The remaining 28.9% of cases involved nonacute indications.

Most acute PCIs were deemed appropriate (98.6%). Only 1.1% of cases were classified as inappropriate; all involved asymptomatic, stable patients who received PCI more than 12 hours after symptom onset following STEMI but did not have hemodynamic or electrical instability. Minimal variation was seen among hospitals for inappropriate use in acute cases (median 0.7%; interquartile range [IQR] 0-1.5%).

Many Nonacute PCIs Don’t Measure Up

Half of nonacute PCIs were classified as appropriate (50.4%). But over a third (38.0%) were deemed uncertain and the rest (11.9%) inappropriate. Reasons for the ‘inappropriate’ designation included:

  • No angina (53.8%)
  • Low-risk ischemia on noninvasive stress testing (71.6%)
  • Suboptimal (≤ 1 medication) antianginal therapy (95.8%)

In contrast to acute PCI, the proportion of inappropriate use of nonacute PCI showed substantial variation at the hospital level (median 10.8%; IQR 6.0-16.7%) and accounted for up to 55% of cases at the highest end of the spectrum. Yet the researchers found that a hospital’s annual volume of nonacute cases had little influence on the likelihood of inappropriate use, nor did the prevalence of inappropriate use differ between private and public hospitals.

“Although some of the inappropriate procedures may be explained by extenuating circumstances (eg, high-risk coronary anatomical findings not captured in the appropriate use criteria), these factors are expected to be uncommon. . . . It is also possible that patient preferences may influence physician decisions about coronary revascularization,” the investigators comment, but add that clinician own misperceptions of risk/benefit are largely responsible for the current findings.

Dr. Chan and colleagues therefore advocate for better physician education to improve patient selection in the nonacute setting. “One strategy for improvement might be the development of additional decision tools that can provide physicians performing the diagnostic coronary angiogram with real-time guidance about the appropriateness of proceeding to PCI,” they suggest.

Uncertain cases, meanwhile, “represent gaps in knowledge and underscore the need for future outcomes-based studies to clarify the benefits of PCI,” the researchers add.

Looking Back to Move Forward

Christopher J. White, MD, of the Ochsner Clinic Foundation (New Orleans, LA), and SCAI president, expressed concern that the study results could easily be misinterpreted.

In fact, the results are mixed, he told TCTMD in a telephone interview. “There’s good and there’s bad, and there’s work to be done,” he said.

For example, many inappropriate or uncertain classifications may be due to coding errors or the fact that hospitals collect data in a way that translates poorly to the NCDR database, Dr. White proposed. “The reason why there is such a huge uncertain group in the electives stems from a lot of the coding issues. They didn’t [fall on] 1 side of the line or the other, so they just had to score them as uncertain,” he said. “But clearly once we learn what the magic words are, the coding words, how to get people to fit in the right boxes, a lot of those uncertains are going to move [to 1 side or the other].”

Dr. White stressed that appropriate use criteria are not used upfront by clinicians as a way to guide patient management. Rather, they serve as a score card for benchmarking the quality of care on a regular basis. As important as it is to remember that the score card can have gray areas, “this really is a good tool to guide our therapies,” he said, adding, “So if we look at our score cards in certain categories and [find] we’re falling below or not doing as well as we think we are, we will then drill down into that and say, ‘What’s the problem here?’”

However, he cautioned that the JAMA paper does not point to specific solutions for clinical practice. “This paper is simply announcing to the community at large . . . that we can now gauge the quality metric of appropriate use criteria. We could change the criteria tomorrow if practice changes, and then the scores would change. It’s not going to be something written in stone or inflexible over time, but it’s a very valuable and objective tool for continuous quality improvement.”

Indeed, “the results of this study should be interpreted knowing that the [appropriate use criteria] are suggested approaches to care,” ACC president David R. Holmes Jr, MD, of the Mayo Clinic (Rochester, MN), commented in a press release.

The criteria “are not mandated but represent the knowledge and experience base present at the time when the criteria are written,” he noted. “The most important appropriate use approach is full communication of potential risks and benefits of a specific procedure to the patient and family. Clinical judgment and full patient understanding should always guide care. There may be times when what is best for the individual patient is at variance with either [appropriate use criteria] or guidance documents.”

In the paper, Dr. Chan and colleagues point out that their analyses were conducted prior to hospitals knowing how well they measured up to appropriate use. “[F]uture studies of procedural appropriateness will need to account for potential ‘gaming’ of key variables used in appropriateness assessments, such as symptom severity,” they write, arguing for objective and validated patient-centered health status questionnaires as well as routine data audits to ensure integrity.


Chan PS, Patel MR, Klein LW, et al. Appropriateness of percutaneous coronary intervention. JAMA. 2011;306:53-61.



Related Stories:

  • The NCDR CathPCI Registry is an initiative of the ACC Foundation and the SCAI. The study was supported by the NCDR.
  • Drs. Chan and White report no relevant conflicts of interest.