Rate of Diagnostic Angiography Dips After AUC Publication, Canadian Study Finds

After the publication of appropriate use criteria (AUC) for diagnostic angiography in 2012, use of the test decreased among patients treated in the Canadian province of Ontario. The proportion of cases categorized as appropriate, uncertain, or inappropriate didn’t change significantly, however, nor did the proportion of angiograms that resulted in PCI.

With the caveat that the study “was done from a population level and therefore is ecologic,” rather than an attempt to ascertain cause and effect, senior author Harindra C. Wijeysundera, MD, PhD, of Sunnybrook Health Sciences Centre (Toronto, Canada), told TCTMD in an email that it “adds to the body of emerging data on AUC.”

The findings were recently published online in the American Heart Journal.

To chart how the AUC might have affected temporal trends in angiography and intervention, Wijeysundera along with lead author Yaron Arbel, MD (Sunnybrook Health Sciences Center), and colleagues looked at 114,551 angiograms for stable coronary disease performed in Ontario between October 1, 2008, and October 31, 2013.

The average monthly age-sex standardized rate of angiography was higher before the AUC were published online in May 2012 and lower thereafter. Analysis of 29,358 PCIs that occurred in conjunction with these tests showed that the monthly PCI rates per 100 angiograms remained unchanged.

After excluding patients with previous PCI or CABG, the overall rates of appropriate, inappropriate, and uncertain angiograms during the study were 41%, 17%, and 43%, respectively. “From 2008 to 2013, there was a modest increase in the proportion of appropriate angiograms, from 40% to 42%, with no change in the proportion of inappropriate cases,” the researchers report. “There was a small decrease in the proportion of uncertain angiograms.”

There were some differences between patients treated before versus after the AUC. Most notably, the later cohort tended to have higher Canadian Cardiovascular Society class angina, were more likely to have hyperlipidemia, and less likely to have diabetes or hypertension.

The Potential to Improve Care

The results hint that while the rate of diagnostic angiograms performed was lower, a greater proportion of these actually led to PCI, Wijeysundera said. “It suggests that potentially practitioners are targeting patients who are more likely to benefit. However, our sensitivity analysis on the proportions of appropriate angiograms did not confirm this—the rate of appropriate angiograms did not change much over the time period of our study.” He pointed out, though, that the latter analysis only looked at a slice of the population and said future work will attempt to measure it in its entirety.

AUC are “a valuable tool” that must continue to be refined, Wijeysundera said. It is important, he added, to evaluate how they influence physician behavior, so that efforts can be made to ensure the AUC—when used properly—can translate into higher quality care.

Kishore J. Harjai, MD (Geisinger Wyoming Valley Medical Center, Wilkes-Barre, PA), commenting on the study, expressed some skepticism. It’s crucial to ask, he said, “How many Canadian cardiologists actually using the AUC, and how often? We’ve had some difficulty with penetration of AUC [among] American physicians, so I’m a little surprised to see that Canadian physicians would be adopting AUC so quickly. It’s possible, but I’m not convinced.”

Hard to Trace the Source of Shift

Harjai also pointed out that the curve for diagnostic angiography published in the paper shows a decrease occurring well before 2012, with no abrupt change around when the AUC came out.

This pattern “makes me suspect that other factors might be at play in decreasing this utilization,” he told TCTMD. “All one can say from this is that there is a decline . . . but the factors which led to the decline are unclear, and I seriously doubt that it is because of appropriate use criteria.”

Concluding the relationship is causative would be “extremely controversial,” Harjai stressed. Overall healthcare utilization, the supply of physicians, upstream availability of nuclear stress testing, and the risk level of the population as a whole—including patients who did not undergo diagnostic angiography—all could be influential, he suggested.

“There is certainly a background decrease in the rates of angiography,” Wijeysundera acknowledged. It could stem from multiple causes, he said, “including [more use of] medical therapy or a lower potential prevalence of CAD in the population from better primary prevention, for example.

“The analysis we did was trying to determine if the rate of this background decrease was impacted by publication of the AUC,” he continued. “The analysis suggested that it was, implying that in addition to the decrease in angiography [arising from other factors], there may have been a higher threshold among practitioners to perform angiography as a result of the AUC.”

• Arbel Y, Qiu F, Bennell MC, et al. Association between publication of appropriate use criteria and the temporal trends in diagnostic angiography in stable coronary artery disease: a population based study. Am Heart J. 2016;Epub ahead of print.

Related Stories:

• Center Experiences Declining PCI Volume, Reimbursement After Implementing AUC
• Inappropriate PCIs Slashed by Half Since First AUC Released, But Key Questions Remain
• AUC ‘Here to Stay’ but Should Evolve, Expert Says