Does PCI Public Reporting Affect Angiography for Out-of-Hospital Cardiac Arrest?

Rates of coronary angiography for patients with out-of-hospital cardiac arrest (OHCA) are similar across states with and without public reporting of PCI outcomes, according to a new study, but investigators say the findings might not tell the whole story when it comes to avoiding treatment of high-risk patient populations.  

Overall, there appeared to be modest “spillover” effect of public reporting on the use of coronary angiography in OHCA, although the difference between states was not statistically significant. The unadjusted rate of coronary angiography in New York and Massachusetts, where PCI outcomes are publicly reported, was 20.8%; it was 22.8% in Delaware, Connecticut, Maine, Vermont, Maryland, and Rhode Island, six neighboring states that do not mandate public reporting (P = 0.35).

“I would be careful just looking at the bottom line, to [just] look at the P value and say there is no evidence of risk-avoidance behavior at all,” senior investigator Jay Giri, MD (Hospital of the University of Pennsylvania, Philadelphia), told TCTMD. “I think [our study] raises concerns that patients were less likely to undergo cardiac catheterization in those two states. When you put it in context with prior work, it does seem that the degree of risk avoidance is less strong than for not performing PCI.”

In 2015, for example, Stephen Waldo, MD (University of Colorado School of Medicine, Aurora), and colleagues used data from the Nationwide Inpatient Sample (NIS) database and showed that acute MI patients were significantly less likely to undergo PCI in New York and Massachusetts than in neighboring states without public reporting between 2005 and 2011. Acute MI patients with concomitant cardiogenic shock or cardiac arrest were 42% less likely to undergo PCI in New York or Massachusetts than in states without public reporting. 

Speaking with TCTMD, Deepak Bhatt, MD (Brigham and Women’s Hospital, Boston, MA), who has studied the unintended consequences of public reporting, said these latest results are provocative but should be interpreted with caution given that the difference in the angiography rates was not statistically significant.

Nonetheless, “I do think there is a signal here of a lower use of a nonreported procedure,” he said. “I think it shows the potential risks of public reporting for procedures like PCI.”

Risk Aversion Shifting Behavior

Since the introduction of public-reporting initiatives for PCI, all of which were designed to promote transparency and improve patient care, there has been evidence that it affects physician behavior by increasing risk aversion, particularly in high-risk patients who stand to benefit the most for an intervention, said Bhatt. The present study extends the concern that public reporting may lead to other risk-averse behaviors, such as avoiding angiography.

“It makes sense in that if there is general risk aversion that is created with public reporting of a procedure there would be a spillover effect,” said Bhatt. “If you’re worried about public reporting of PCI, and you’re worried about really sick patients, then why even do the diagnostic coronary angiogram if you’re not planning on doing PCI?”

In contrast, Rishi Wadhera, MD (Brigham and Women’s Hospital), also commenting on the paper for TCTMD, stuck to a more statistical interpretation of the findings, stating that lack of a significant difference between states with and without public reporting suggests the practice “may not push interventional cardiologists to avoid diagnostic coronary angiograms in patients with out-of-hospital cardiac arrest.” What is unknown, however, and is critically important, “is whether these patients are less likely to undergo PCI in public-reporting states,” he told TCTMD.

Like Giri and Bhatt, Wadhera stated there is plenty of solid evidence showing public reporting does encourage risk-averse behavior among physicians, and that this impedes access to care for critically ill patients, such as those with acute MI and cardiogenic shock, for example. 

Higher-Risk Patients Need Most-Aggressive Therapy 

In this new study, published April 8, 2019, in Circulation: Cardiovascular Interventions, the researchers used the NIS database to analyze 50,125 admission records with OHCA to determine if there was additional risk avoidance with respect to the use of coronary angiography, a nonreported variable in states that require public reporting.

Overall, there was a nonsignificant trend toward a lower rate of cardiac catheterization among states with publicly reported PCI outcomes (OR 0.84; 95% CI 0.66-1.06). In a prespecified subgroup analysis—which Giri said should also be interpreted with caution—there was a significant 29% and 26% lower rate of coronary angiography in older patients (≥ 65 years) and Medicare patients, respectively, treated in New York and Massachusetts compared with the other states.

“Despite the fact that older patients are more likely to have out-of-hospital cardiac arrest based on a coronary cause, they were the least likely to go for coronary angiography,” Giri said. “So it makes you a little concerned about the risk-treatment paradox. This is a higher-risk group who might actually have more to gain from a diagnostic cath. If you can imagine, in a doctor’s mind, things might just start to add up: this is an out-of-hospital cardiac arrest patient, it’s already bad news, and am I really going to take this patient for a cath?”

At present, the management of OHCA patients is somewhat controversial given the dearth of randomized, controlled clinical trials. In March at the American College of Cardiology 2019 Scientific Session, investigators presented data from the COACT study showing that immediate angiography did not improve survival when compared with a delayed strategy in resuscitated OHCA patients without signs of ST-segment elevation. There are also two ongoing clinical trials—DISCO and PEARL—looking at early and delayed coronary angiography in resuscitated cardiac-arrest patients with no signs of ST-segment elevation.

For patients with STEMI complicated by cardiac arrest, there is strong evidence for immediate coronary angiography and PCI, according to the researchers. However, some patients who survive cardiac arrest don’t have typical ST-segment elevations on ECG, said Giri. In the present analysis, there was no difference in the utilization of coronary angiography among OHCA patients with or without evidence of STEMI.

In an editorial, Aakriti Gupta, MD (NewYork-Presbyterian/Columbia University Irving Medical Center), Hitinder Gurm, MD (University of Michigan, Ann Arbor), and Ajay Kirtane, MD (NewYork-Presbyterian/Columbia University Irving Medical Center), note that just 9% and 28% of the study sample included OHCA patients with STEMI or ventricular fibrillation, respectively. Because these are the patients who would stand to benefit most from PCI, it would be worthwhile to determine angiography rates across states with and without public reporting for these conditions, they state.

Regarding the overall findings, Gupta, Gurm, and Kirtane also expressed surprise that just one in five OHCA patients underwent coronary angiography regardless of public reporting initiatives. This rate is significantly lower than the 50% observed in the Cardiac Arrest Registry to Enhance Survival between 2011 and 2013, but this is possibly explained by the absence of randomized trials in the arena of cardiac arrest, they add.  

Disease-Based Reporting, Not Procedure-Related

In the modern era of public reporting, the performance metrics of interventional cardiologists are watched closely and there is pressure to maintain a good clinical scorecard, said Giri. Yet, “it’s extraordinarily unlikely that doing a PCI or a diagnostic cath is going to be the cause leading to a bad outcome on [OHCA] hospitalizations,” said Giri. “It’s almost always driven by the fact they came in with cardiac arrest. They’re the sickest of the sick patients.”

To TCTMD, Bhatt said there are different ways to report PCI-related outcomes that do not have an adverse impact on physician behavior. For example, hospitals within the Veterans Affairs (VA) health system, where he worked for a number of years, report PCI outcomes and other procedural data to peers but not to the public at large.

“As a cardiologist in that system, I wasn’t immune when complications arose in my cath lab,” he said. “The system, I felt, did work and was quite objective. It wasn’t just ‘wink, wink, nod, nod.’ People did take quality improvement seriously.”

Giri, who is the director of the cardiac catheterization lab at the Corporal Michael J. Crescenz VA Medical Center, also praised the VA quality improvement approach, which he said includes both quantitative and qualitative performance assessments. In addition to monthly and annual feedback, the VA initiative includes mandatory reporting of all complications (death, stroke, and need for emergent surgery) to a committee of 12 interventional cardiologists who review the case.

“There’s a lot of trust in this process,” said Giri. “I know I’m not going to get railroaded. The interventional cardiologists are in the trenches just like me. I’m not going to avoid risk—they’ll see that it’s a risky patient. They’ll read the chart and then they’re going to talk with me.”

If the practice of reporting data to the public is continued, Bhatt suggested “disease-based” metrics focusing on the outcomes of various pathophysiological and clinical states, including unstable angina, NSTEMI, STEMI, and acute MI complicated by cardiogenic shock or cardiac arrest, for the healthcare system as opposed to for an individual operator or hospital. “I think that would take away some of the perverse incentives that do arise,” he said.

Wadhera lamented that while risk aversion with critically ill patients is a reality of public reporting, policymakers and national societies continue to double down on the initiative in the name of transparency. “We need to be more responsive to evidence and open to rethinking public reporting initiatives to ensure they actually improve care for all patients,” he said.