PCI Public Reporting: Where Some Point to Problems, Others Propose Solutions

Public reporting of PCI outcomes—and its unintended consequences—takes center stage in a collection of papers being published today in JAMA Cardiology.

In one, based on a survey of interventional cardiologists in New York and Massachusetts, investigators conclude that current approaches to public reporting appear to foster risk-averse practice patterns, with early-career physicians expressing the greatest doubts. Fully 65.1% of the survey respondents said they’d avoided doing a PCI at least twice because of worries that a bad outcome would negatively impact their publicly reported data, researchers found.

Another article details how, for now at least, the sharing of PCI outcomes is falling short of its goals, with experts offering ideas on how to improve.

Rishi K. Wadhera, MD, MPhil (Brigham and Women’s Hospital, Boston, MA), lead author of the latter paper, told TCTMD that despite this recordkeeping of PCI results being “part of a broader movement towards transparency in healthcare,” it remains a “polarizing issue” for many in the cardiovascular field.

“There certainly are a lot of advocates who feel that transparency is important, it provides a strong incentive to providers to improve care quality, and [it] empowers patients,” Wadhera said. “But there are also critics who emphasize the methodologic challenges and the unintended consequences of reporting.”

If the healthcare community exits from public reporting, that vacuum will quickly be filled by other entities who will publicly present physician performance data in ways that may or may not benefit patients. William B. Borden

At a time when public reporting initiatives are multiplying across medical specialties in the United States, Wadhera and colleagues Karen E. Joynt Maddox, MD (Washington University School of Medicine, St. Louis, MO), Robert W. Yeh, MD (Beth Israel Deaconess Medical Center, Boston, MA), and Deepak Bhatt, MD (Brigham and Women’s Hospital), sought to synthesize what’s known about this practice in PCI.

“We have two decades of data and evidence on public reporting of PCI outcomes in particular,” Wadhera said. “And I think the three main objectives of public reporting are to improve care delivery and patient outcomes, to make sure that comparisons of provider and institutional care quality are fair and accurate, and to empower patients to make meaningful decisions about where they choose to receive care.”

To date, he said, the evidence indicates these goals haven’t been accomplished.

The Path Forward

An accompanying editorial by William B. Borden, MD (George Washington University, Washington, DC), makes the case for continued public reporting. “Turning away from public reporting is not an option,” Borden says. “Despite the low public knowledge about PCI registries, the clear and strong societal trend is toward the public demanding more information and greater transparency.

“Empowering individuals with knowledge is a good thing. That knowledge must be as accurate as possible and presented with proper context so that patients can make appropriate informed decisions, and physicians and hospitals can adapt care delivery to improve health outcomes and equity, rather than the converse,” he continues. “Inferences about physician and hospital performance can be made from a variety of increasingly available data sources. If the healthcare community exits from public reporting, that vacuum will quickly be filled by other entities who will publicly present physician performance data in ways that may or may not benefit patients.”

Given that the enthusiasm for reporting does not appear to be waning, Wadhera et al describe several paths forward: disease-based rather than procedure-based reporting, more focus on processes and patient-centered outcomes, and potentially the creation of separate databases that aren’t open to public scrutiny, among others.

If public reporting is to continue, “the most appealing solution is probably disease- or condition-based reporting,” Wadhera commented.

It would facilitate transparency about how the entire healthcare team is delivering care,and not just interventionalists alone. Rishi K. Wadhera

This method, which would compare outcomes among patients treated for acute MI, for example, or even separate out those with NSTEMI versus STEMI, “provides a more comprehensive and robust assessment of care quality, irrespective of whether a patient’s treated with medical therapy or a procedure,” Wadhera noted.

Disease-based reporting would also help alleviate risk aversion among interventionalists as well as “provide more meaningful information to institutions about where they’re performing well and where they’re performing poorly in terms of specific conditions, so that they can develop more targeted interventions,” Wadhera continued. Additionally, by providing more balanced comparisons, he said, it would reduce the need for risk adjustment.

“Probably most importantly, it would facilitate transparency about how the entire healthcare team is delivering care, and not just interventionalists alone,” Wadhera stressed.

Currently, he said, the Centers for Medicare & Medicaid Services reports 30-day mortality rates for acute MI at a hospital level, though not at a provider level or stratified by clinical presentation.

Risk Aversion Abounds

The survey results appearing in JAMA Cardiology, meanwhile, add greater depth to data previously reported at the American Heart Association 2017 Scientific Sessions. Lead investigator Daniel M. Blumenthal, MD (Massachusetts General Hospital, Boston), said it seems that “for better or for worse, people are not all that surprised by the findings” of widespread risk aversion.

“Nobody is against policies which improve outcomes,” he added. “Everybody’s 110% in favor of public policies” that as a whole lead to more benefit than harm for patients, Blumenthal observed. But “the interventional cardiology community I think is far from certain that that is the case for public reporting. We’re all arguing our absolute best to try to deliver the best possible care to our patients in the environments where we’re practicing,” whether that’s in a state with public reporting or not.

For the study, he and his colleagues approached 456 interventional cardiologists in the states of Massachusetts and New York, which mandate public reporting of PCI outcomes, via an online survey. In all, 32.7% responded, though the response rate in Massachusetts was nearly double what it was in New York.

Unlike prior surveys that were focused only on New York, the current effort valuably shows that the unintended consequences of public reporting are not limited to a single state, Blumenthal noted to TCTMD.

There are some differences in policies between the two states, he said. New York, for example, doesn’t mandate reporting the outcomes of patients in cardiogenic shock, while Massachusetts allows cardiologists to retrospectively apply for very sick patients to be exempt from public reporting. Yet the survey revealed no significant differences based on where interventionalists were practicing.

For better or for worse, people are not all that surprised by the findings. Daniel M. Blumenthal

Beyond acknowledging their own risk aversion, the survey respondents also said they felt peer pressure: 59% reported sometimes or often being urged by colleagues to avoid performing PCI in a patient at high risk of death. Nearly all (94.7%) thought that other interventionalists in their state sometimes or often avoid PCIs for this reason. Approximately half (51.7%) were concerned “some or a lot that their superiors would not support them for performing an indicated PCI in a critically ill patient who later died of a PCI-related complication,” Blumenthal et al note.

It’s important to keep in mind, though, that interventionalists’ perceptions of their peers may not match reality, he cautioned to TCTMD.

The researchers also found that 81.2% of respondents said they knew some or a lot about risk-adjustment methods used in public reporting, while 73.8% had little or no trust in the ability of these methods to work.

On multivariable adjustment, less-experienced operators were more likely to feel pressured to avoid an indicated PCI and to think other cardiologists were avoiding PCIs, and they were less likely to have faith in risk adjustment.

“This study by Blumenthal and colleagues really bolsters the case that risk aversion is a reality of public reporting,” Wadhera said.

The differences based on experience level could be due to “younger interventional cardiologists [feeling] like they have more at stake, because they’re less established,” he suggested. “Older, more established interventional cardiologists may not feel the same way.”

Blumenthal hesitated to speculate why age mattered, but said it does “highlight the need for some additional efforts to understand which subsets of interventional cardiologists feel more pressured to perhaps be more risk averse in the face of public reporting.

“Ultimately if we want to try to mitigate or ameliorate some of the unintended consequences, we need to really start to dig down into those details,” he concluded.

Note: Ajay J. Kirtane, MD, and Roxana Mehran, MD, who co-authored the study led by Blumenthal, are faculty members of the Cardiovascular Research Foundation, the publisher of TCTMD.