Public Reporting of PCI Outcomes: High Costs and Physician Burden

Debra L Beck

May 23, 2019

The median reported annual cost for Massachusetts catheterization laboratories to meet their public reporting requirements was between $100,000 and $200,000 yearly, ranging from $0 to $50,000 to more than $300,000 yearly.

Half of interventional cardiologists surveyed (seven of 14) reported they spent between 0 and 5 hours a week on efforts related to reporting, with three respondents saying they spent 10 hours or more.

"As an interventional cardiologist, practicing in a public reporting state and involved in the central adjudication committee for our reporting program, I've appreciated that the time and resource investment is nontrivial and we really wanted to better understand the cost side of the public reporting equation," said senior author Robert W. Yeh, MD, MSc, Beth Israel Deaconess Hospital, Boston.

The health policy team, led by Rishi K. Wadhera, MD, MPP, Brigham and Women's Hospital, Boston, presented their survey data at the American College of Cardiology (ACC) 2019 Scientific Sessions in March and published it in print on May 20 in the Journal of the American College of Cardiology.

They set out to survey all 24 cath labs in Massachusetts that participate in the public reporting program and achieved a 63% response rate. Respondents were more likely to be large hospitals, teaching hospitals, and those with higher mean annual PCI volumes.

NCDR Is Actually the Larger Burden

According to Edward L. Hannan, PhD, MS, providing the ACC's National Cardiovascular Data Registry (NCDR) with their required data is a much bigger burden than state-mandated reporting. PCI outcomes are reported through the ACC's CathPCI data instrument.

Hannan is from the University at Albany, New York, and has been studying the impact of public reporting on cardiovascular care in New York State since the early 1990s. In New York, there is no mandate to collect NCDR data and their state-mandated public reporting form is two sides of a single page.

"I think the big caveat — and the authors themselves mention that they did not separate out time and costs — is that the CathPCI form is gigantic," he said in an interview. "So, the real question is once hospitals have volunteered to be part of NCDR, what is the marginal amount of work required beyond what is needed for NCDR reporting?"

He suggested that the findings of Wadhera et al are somewhat misleading, in that they imply the time and cost burden is directly related to state public reporting requirements, when most of what had to be recorded was done to complete the NCDR forms.

Ravi Hira, MD, University of Washington, Seattle, said the data needed for the NCDR and their public reporting scheme is one and the same. "The data are being collected anyway for NCDR, so it's not an extra burden, at least from the perspective of what we do in Washington State. Our vendor just goes in and pulls out the data we need on the state level."

Hira is the medical director of the Cardiac Care Outcomes Assessment Program (COAP) in Washington. For COAP, there is a separate, per case data submission fee for participating hospitals but no additional abstraction cost or time needed beyond that required for the national registry submissions.

Yeh agreed that there is redundancy in the data requirements — and that the NCDR form has grown exponentially — but the requirements are "sort of indistinguishable because it's mandated that we engage in the NCDR in order to fulfill our public reporting requirements, so it's all kind of one in the same at this time."

The public aspect of state-mandated reporting, however, imposes a need for greater accuracy, he suggested, than that required for NCDR data collection. "We want to make sure our data are very accurate because the stakes for the institution are high, compared to, say, a place where all reporting is nonpublic," Yeh added

He also noted that this new survey was completed before the latest and largest iteration of the CathPCI form was available. The most recent version — CathPCI Version 5 — includes, among other things, new requirements for reporting of patient-reported outcome measures (PROMS), a frailty assessment, details of out-of-hospital cardiac arrest, and assessment of fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR).

"So, if there is a sense that all this money is being spent with no benefit in terms of outcomes, then the same question applies to the national registries that we are part of, like CathPCI and STS," said Hira. "However, I do believe there is a lot of benefit, not just in terms of transparency, but also for sharing best practices and improving quality of care locally by participating in a state public reporting registry."

"I think these costs that we reported, at least for Massachusetts, do represent the costs incurred by requiring participation in the public reporting process," said Yeh.

At present, the public PCI reporting program in Massachusetts is on hold, but institutions continue to collect the required information, Yeh noted, if only because it is unclear whether they are legally still required to do so.

Risk Aversion vs Transparency?

Less transparency in healthcare is a hard sell these days, but opponents of public reporting have suggested that publicly reporting individual physician outcomes, or even institutional outcomes, can lead to physicians not performing interventions on high-risk patients for fear of lowering their scores.

"We don't provide individual provider outcomes in Massachusetts, and I think that is the right approach. In New York they do, and I think that is a particularly dangerous recipe to induce risk aversion," said Yeh.

As reported in 2018, in an online and anonymous survey of interventional cardiologists in Massachusetts and New York, 65% of participants reported avoiding high-risk PCI intervention on at least two occasions.

The study was limited by a low response rate (33%), but showed that interventionalists in these two states where public reporting is mandated — only the latter of which reports physician-specific outcomes — were equally likely to report avoiding high-risk PCIs.

This phenomenon — and more so, the willingness to admit it — is "mind boggling" to Hannan.

"What they're saying is, in effect, if you continue to publicly report data on our outcomes, we are going to not do the right thing and people are going to die because we're afraid we're going to look bad," said Hannan.

"If this is true, then it's certainly an unintended consequence, but it's amazing to me that they're not ashamed to admit this, even anonymously," he added.

"I think the concern is that, sometimes, physicians don't always act rationally…and they may not believe that the risk adjustment adequately captures the risk of the patient in front of them," said Yeh.

Hannan's group published a study in 2017 that asked whether hospitals or operators are disadvantaged with regard to risk-adjusted mortality rates and outlier status as a result of treating shock patients. More specifically, they asked how the equation would change if cardiogenic shock cases were no longer reported publicly.

"We saw that risk-adjusted hospital and physician mortality rates are very highly correlated, regardless of whether shock patients are included in public reporting," Hannan said. "There was a very strong correspondence between reporting shock outcomes and not reporting shock outcomes in terms of who got flagged as an outlier."

Highlighting the lack of clarity on how reporting affects physician behavior, another recent study showed that when patients with ongoing cardiogenic shock were excluded from New York PCI public reports, as was done in 2006, there was a significant increase in the use of PCI for cardiogenic shock and a concomitant decrease in in-hospital mortality that exceeded simultaneously observed trends in comparator states.

"What they didn't say in that paper, however, was that New York's mortality rate for all shock patients was significantly lower compared with the comparison states both before and after exclusion of cardiogenic shock patients," said Hannan.

Hira thinks there is a middle ground between public reporting of hard outcomes and losing the transparency that public reporting represents. He summarized his experiences and thoughts on public reporting in a letter in JAMA Cardiology last year.

"The way we've chosen to do this in Washington is, first, our system is completely voluntary, but also we track a lot of process measures, not just mortality. If you just track mortality, it depends on who you treat, but when you look at other metrics that are less dependent on patient factors, we feel there is more room for process improvement and quality improvement, and less of a worry that risk aversion will occur," said Hira to | Medscape Cardiology.

In Washington, PCI outcomes are publicly reported, but on a hospital level, not physician level. Participation in the program is voluntary, and Hira noted that, currently, they have 100% participation.

"I think our program is pretty unique in that providers really see the value in being able to improve practices, and we have 100% participation without it being mandatory," said Hira.

Mortality and other outcomes are risk adjusted in Washington, as they are in all state-mandated reporting schemes, irrespective of whether the outcomes are hospital-specific or physician-specific.

One way to avoid risk aversion is to move from procedure-based outcomes to reporting disease-specific outcomes, suggested Yeh. Indeed, New York started disease-based reporting for acute myocardial infarction last year.

"The main problem with many of these measures is that the denominator is the patients who get a procedure and all it takes is for physicians to just not offer the procedure to manipulate these metrics and change the denominator," Yeh said. "But really, the denominator of interest if you're measuring quality of care is not who they chose to do a procedure on, but rather who came to be treated for a given problem, whether or not they got a procedure."

He added: "I've never called for our program to be disbanded or anything like that. I think it has value, but it's important to just recognize the cost and have a continuing discussion about whether or not public reporting is working, whether it's providing value, how it might be done better.

Yeh, Hannan, and Hira reported no conflict of interest.

J Am Coll Cardiol. 2019;73:2604-2608. Abstract


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.
Post as: