Volume-Outcome Relationship Questioned in CHIP PCI for Stable Angina

The preliminary data, however, suggest that CHIP may be defined too broadly, a researcher says, urging cautious interpretation.

Volume-Outcome Relationship Questioned in CHIP PCI for Stable Angina

Outcomes of interventions in the “complex higher-risk and indicated patient” (CHIP) population with stable angina are not dependent on operator volume, new data suggest, but the lead researcher said the result could be due to a CHIP definition that casts too wide of a net.

Numbers from the British Cardiovascular Intervention Society (BCIS) database showed that the proportion of PCI procedures considered CHIP cases—based on a definition adapted from a 2016 white paper—increased over time, from 28.1% in 2007 to 36.2% in 2014 (P < 0.001).

Lead author Tim Kinnaird, MD (University Hospital of Wales, Cardiff), estimated that most operators, if asked about the percentage of their cases that are complex, would put the figure somewhere between 10% and 20%.

“I think we were surprised that using those definitions that it was as high as [about] 35% in the last few years. Maybe the definition of CHIP that had been proposed was a little generous,” Kinnaird told TCTMD. Including too many patients under the CHIP umbrella, he added, might explain why there was no significant association between higher operator volume and improved 12-month patient survival, the opposite of what the team expected at the outset.

“One of the things this paper highlights is: I don’t think we really know what the definitions of CHIP are,” Kinnaird said. “ We set out with a relatively clear question but it got more muddy as the study went on, and I think some of the findings led us to question the premise of CHIP perhaps.”

High-Volume Operators Treat More-Complex Patients

Kinnaird noted that prior studies looking into whether a volume-outcome relationship exists for overall PCI have provided divergent results. Some have shown that higher-volume operators provide better patient outcomes, whereas others—including a 2018 study from Kinnaird’s group—have failed to support such a relationship. But there really aren’t any studies exploring the issue in the CHIP population, Kinnaird said.

I think there are a number of questions, actually, that this work has led to rather than putting to bed the concept that volumes are not associated with better outcomes. Tim Kinnaird

In this study, published recently online ahead of print in the American Heart Journal, the investigators identified all cases of CHIP PCI for stable angina performed in England and Wales between 2007 and 2014. Data on operator volumes were available for the last 3 years of the study period.

CHIP cases were defined either by patient characteristics (age 80 or older, LVEF below 30%, prior CABG, or chronic renal failure) or procedural variables (left main PCI, revascularization of a chronic total occlusion [CTO], or use of LV support, rotational atherectomy, or laser atherectomy).

Between 2012 and 2014, there were 30,268 CHIP cases performed by operators with volumes ranging from one to 580 (median 29). Over that span, 409 operators performed 45 or fewer CHIP procedures, 120 performed 46 to 85 cases, 70 performed 86 to 138, and 35 performed 139 to 580.

Higher-volume operators tended to treat patients with a heavier comorbidity burden and more complex coronary disease, undertaking procedures with greater complexity. That, according to Kinnaird, explains why there were significant associations between increasing operator volume and higher rates of in-hospital major bleeding, access-site complications, and coronary perforation. Higher-volume operators did, however, achieve lower rates of periprocedural MI, which drove a reduction in in-hospital MACCE. Also, as volumes went up, so did the likelihood of same-day discharge (P < 0.001).

But operator volume was not significantly associated with mortality either in the hospital (P = 0.394) or at 1 year (P = 0.638).

‘Volumes Do Matter’

Commenting for TCTMD, Arasi Maran, MD (Medical University of South Carolina, Charleston), a high-volume operator specializing in CTO PCI, said the message of the study should not be that volumes do not matter when it comes to CHIP procedures.

That higher-volume operators had more complications and achieved similar mortality is a reflection of referral bias, whereby cardiologists handling more cases are also seeing more-complex patients, Maran said. “A low-volume operator will have better outcomes only because they are cherry-picking their cases, whereas a high-volume operator is going to take all-comers.” She added that high-volume operators may have more complications, but they’re also better equipped to handle them than their peers performing fewer procedures.

It would be “absurd” to conclude that low-volume operators can achieve better outcomes when compared with high-volume operators, Maran said. “Volumes do matter. Practice makes perfect. It’s age-old wisdom passed on for generations, and this paper I feel takes away from that message,” she said, adding that if her mother required a complex, high-risk PCI, she wouldn’t want her to go to a low-volume operator.

Maran agreed with Kinnaird that the CHIP definition used in this study likely includes patients who are not truly complex, saying, “We need a much more stringent definition of what CHIP actually means.” To that end, a Society for Cardiovascular Angiography and Interventions writing group—of which Maran is a member—is working on a paper that will provide an updated CHIP definition; it’s expected later this year.

Practice makes perfect. It’s age-old wisdom passed on for generations, and this paper I feel takes away from that message. Arasi Maran

Kinnaird suggested that older age alone might not be a good indicator of CHIP status. Age often brings with it high-risk features like calcification, left main disease, and a history of CABG, “but if you’re an older patient without any of those, does that make you a complicated patient? I’m not sure that it does,” he said. Prior CABG, too, might not necessarily be a marker of complexity on its own, he added. “So I think we would want to perhaps consider taking out some of those from the criteria.”

In the meantime, research continues in this area, Kinnaird said, pointing to issues related to patient populations and endpoints as well as exploration of the utility of a CHIP score that takes into account the number of complexity factors in an individual patient.

“I think there are a number of questions, actually, that this work has led to rather than putting to bed the concept that volumes are not associated with better outcomes,” he said, calling these data preliminary. “I think we need to do a lot more work.”

  • Kinnaird and Maran report no relevant conflicts of interest.

We Recommend