Board Certification Not a Strong Predictor of Operator's PCI Outcomes: NCDR Analysis

September 23, 2015

NEW HAVEN, CT — Patients treated by cardiologists not certified by the American Board of Internal Medicine (ABIM) in interventional cardiology (ICARD) have a 10% higher risk of dying in the hospital and are more likely to be referred for emergency CABG surgery than patients treated by board-certified interventionalists, according to the results of a new analysis[1].

However, investigators did not observe an association between board certification in interventional cardiology and vascular and bleeding complications.

The bottom line, according to Dr Paul Fiorilli (Hospital of the University Pennsylvania, Philadelphia) and colleagues, is there was no consistent association between ICARD certification and patient outcomes. They say the clinical significance of the mortality and CABG findings is open to interpretation, noting the absolute difference in events was very small. For example, the adjusted absolute increase in mortality was 0.08%, which corresponds to one additional death for every 1250 patients. Regarding emergency surgery, there was one extra CABG performed for every 3333 patients treated by non–ICARD-certified doctors.

Speaking with heartwire from Medscape, senior investigator Dr Jeptha Curtis (Yale University School of Medicine, New Haven, CT) said certification is an indicator that operators have met a number of metrics, not just the certifying exam, but that it's not a reliable predictor of an operator's clinical outcomes.

"I think certification does carry value," he said. "However, our take-home point is that it doesn’t do a very good job of discriminating performance. From a consumer perspective, knowing certification status isn’t providing actionable information about the quality of individual operators. Certification is laudable, but by itself it doesn't convey much information about expected outcomes from a specific operator."

Data From the NCDR CathPCI Registry

In their report, published September 18, 2015 in Circulation, the researchers note that interventional cardiologists without ICARD certification perform a "substantial proportion of PCI procedures." In their analysis of the National Cardiovascular Data Registry (NCDR) CathPCI Registry, they identified 1509 physicians not ICARD certified who performed 111,555 PCI procedures in 2010. In contrast, the remaining 3666 interventional cardiologists were board-certified and performed 399,153 PCIs.

Prior to the investigation, Curtis said he was open-minded about the relationship between ICARD certification and clinical outcomes.

"At the institution where I do most of my procedures, there is a mix of operators who are and are not certified," he said. "On the basis of that experience, I didn't really expect to see a very large effect. On the other hand, we did a similar study looking at outcomes after ICD implantation with electrophysiologists and nonelectrophysiologists and saw a clinically meaningful association between certification status and outcomes."

In the CathPCI Registry, non–ICARD-certified physicians were more likely to treat patients with a prior PCI and patients undergoing an elective PCI, whereas the ICARD-certified interventionalists were more likely to treat patients with heart failure, ST-segment-elevation MI (STEMI), and acute coronary syndrome (ACS). ICARD-certified physicians performed an average of 112 PCIs in 2010 vs 76 PCIs performed by the noncertified interventionalists.

Overall, the crude absolute event rates between the ICARD- and non–ICARD-certified physicians were nearly identical.

Crude clinical outcomes

Outcome Non–ICARD-certified PCIs, n=111,555 (%) ICARD-certified PCIs, n=399,153 (%)
In-hospital mortality 1.4 1.4
Bleeding 1.7 1.8
Vascular complications 0.4 0.5
Emergency CABG 0.2 0.2
Composite (in-hospital death, bleeding, vascular complications, and emergency CABG) 3.4 3.4

In a multivariate-adjusted risk model, which accounted for patient characteristics and PCI volume, the rate of all-cause in-hospital mortality was 10% higher among noncertified interventionalists (odds ratio [OR] 1.10; 95% CI 1.02–1.19) compared with the ICARD-certified physicians. Similarly, the need for emergency CABG was 32% higher among the non–ICARD-certified physicians compared with those who were certified (OR 1.32; 95% CI 1.12–1.56).

Rates of bleeding, vascular complications, and a composite end point that included death, bleeding, vascular complications, and emergency revascularization were no different between the two groups.

The researchers point out that PCI is a much safer and reliable procedure than it was in its early development, and patient outcomes are increasingly being attributed to the totality of care during and after the PCI procedure. These factors might "level the playing field," they add, "making it possible for interventional cardiologists with different training, knowledge, and technical expertise to achieve comparable results."

To Certify or Not to Certify?

In an editorial[2], Dr Spencer King (Emory University School of Medicine, Atlanta, GA) notes the analysis does not provide information on the "most discriminating features" of the intervention, such as the anatomy, disease distribution, or the difficulty of the intervention, such as bifurcation lesions or chronic total occlusions. Among clinical outcomes, MI is notably absent, writes King, pointing out that in-hospital mortality as the primary quality metric for PCI "has always been problematic."

Still, the finding of one excess death per 1250 patients treated by non–ICARD-certified interventionalists is "not an insignificant clinical outcome." Based on the population of Atlanta, GA, that would translate into 10 excess deaths per year.

Like the researchers, King notes the physicians are a heterogeneous group and that 83.5% of the interventionalists had passed the board exams at some point in time. That leaves 16.5% who never passed the board exams, and of these, 2.3% trained during or after 1999, the year American Board of Internal Medicine introduced board certification for interventional cardiology.

"Although the number of physicians in this group is small (149 of 5175), the multivariable outcomes of death and emergency surgery were highest in this group," writes King. "Are these physicians who completed their training after board certification was established without boards because they did not take the exam or because they failed to pass the exam?"

King notes that for physicians who trained more than 16 years ago, there is no option to become board certified, as they have been ineligible to take the exam for over 10 years.

To heart wire , Curtis said there are different reasons a physician might not be certified. For physicians who completed training prior to the introduction of certification, operators may have been grandfathered in and chose not to pursue certification. For more recent trainees, the absence of certification likely reflects the failure to complete one or more requirements, including the certification exam.

"I think certification is an indicator of commitment and a specific type of training, but we have to really figure out what it can and should be used for," said Curtis. "For example, it is often used for hospital credentialing. Our findings suggest this may not be appropriate and could result in high-quality operators being excluded from the cath lab."

The study was supported by the American College of Cardiology's National Cardiovascular Data Registry and a grant from the National Heart, Lung, and Blood Institute. Fiorilli has no  relevant financial relationships. Curtis receives salary support from the ACC NCDR to provide analytic services and with the Centers for Medicare & Medicaid Services to support the development of quality metrics; he holds equity interest in Medtronic. Disclosures for the coauthors are listed in the paper. King has no relevant financial relationships.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.