Public reporting of percutaneous coronary interventions

Despite enhancing transparency, public reporting may be leading to avoidance of life-saving procedures

In the United States, public reporting of percutaneous coronary interventions (PCI) has been implemented in several states to enhance transparency and accountability, with the intent of improving patient outcomes. However, the impact of public reporting remains a controversial issue. A growing body of evidence suggests that public reporting has not improved patient outcomes. In fact, robust evidence shows that it has instead led to risk aversion. Public reporting of PCI stands at the crossroads of competing priorities. By the principle of beneficence — “doing good” — public reporting seeks to enhance transparency, but it may violate the principle of non-maleficence — “do no harm” — through the unintended consequence of avoidance of high risk patients.

In this article, we discuss how PCI has affected patient outcomes and physician and patient behaviour, and suggest potential steps for improvement.

Impact of public reporting on patient outcomes: evidence on the experience of the United States

Public reporting of PCI outcomes in the US was initially implemented in New York State in 1997, and has since spread to three other states: Massachusetts, Texas and Washington. While public reporting in general has grown increasingly common across the world, few other countries publicly report PCI outcomes to the same degree as the US. In Canada, for example, PCI outcomes are reported by hospital, but there is little literature on the experience and impact of PCI public reporting outside the US.1

As experience has increased with public reporting for PCI in the US, a growing body of evidence has emerged that elucidates the impact of public reporting on patient outcomes. Overall, evidence has not shown that reporting has led to any clear improvement in patient outcomes. For example, a study of Medicare beneficiaries in the US found that there was no difference in 30-day mortality rates for patients with acute myocardial infarction (AMI) in reporting states versus non-reporting states.2 However, the analysis did note that patients in the ST-elevation myocardial infarction (STEMI) subgroup did have higher mortality in reporting states compared with non-reporting states.2

A subsequent and more recent analysis found increased in-hospital AMI mortality in reporting states compared with non-reporting states.3 The AMI mortality rate in reporting states was driven by increased mortality in patients who were not offered PCI. In contrast, mortality among patients with AMI who underwent PCI was actually lower in reporting states compared with non-reporting states. Thus, while PCI outcomes were better in reporting states, this likely reflects risk avoidance, as outcomes for patients with AMI who did not receive PCI were worse in reporting states.

Physician behaviour: aversion and upcoding

A primary mechanism as to why public reporting has not led to an improvement in PCI outcomes may be explained by physician behaviour — particularly, aversion to high risk PCI for patients who may need the procedure the most. On the whole, evidence suggests that the volume of coronary angiography and PCI for AMI is lower in reporting states compared with non-reporting states.2,3 If these PCI cases were attributed to patients in which PCI is futile, where the harms clearly outweigh the benefits, then the decrease in PCI volume would be worthwhile and reduce health care costs. However, lower rates of PCI in reporting states appears most pronounced among high risk patients, such as those with AMI and cardiogenic shock2-4 — patients who may benefit most from the procedure.

To support this point, two surveys have revealed that public reporting has altered interventional cardiologists’ decisions to perform PCI.5,6 In a survey of cardiologists in New York, 78% of physicians believed that public reporting resulted in avoidance of PCI in patients who would have benefitted from the procedure.5 In another survey of interventional cardiologists subject to public reporting in New York and Massachusetts, 65% of cardiologists reported having turned down PCI on at least two occasions due to the impact a negative outcome would have had on their publicly reported outcomes.6 Cardiologists are affected not only by patient perceptions but also by peer perceptions, as 59% of physicians reported pressure from their colleagues to avoid performing high risk PCI due to concerns about a poor outcome.6

Furthermore, there is also evidence of upcoding of risk factors and patient comorbidities in patients who undergo PCI in reporting states, reducing the accuracy of public reporting of PCI. Independent adjudication has shown evidence of upcoding in reporting states, and surveys of interventional cardiologists show that many physicians believe upcoding is still quite prevalent.6,7

Patient behaviour: patients do not use public reporting data

Public reporting aims to improve patient outcomes by increasing transparency. However, hospitals are held accountable if patients not only view but also use public reporting data to alter their decisions for seeking health care. Current literature suggests that patients very rarely use public reporting data to influence their health care decisions.8 In New York, where public reporting of PCI is well established, only 3% of patients had viewed public reporting data for PCI.5 While the goal of public reporting would be to guide patients away from low performing hospitals towards high performing hospitals, there is also little to no evidence that public reporting leads to changes in referral patterns.8 Furthermore, in emergent and urgent conditions such as AMI or cardiogenic shock, patients have virtually no flexibility in who performs their procedures, thus limiting the role of PCI public reporting in the most important situations.

Potential steps forward

Excluding high risk patients

In evolving efforts to decrease risk aversion, New York excluded patients with refractory cardiogenic shock from PCI public reporting. This exclusion, implemented in 2006, was associated with both an increase in PCI for patients with AMI and cardiogenic shock as well as a decrease in in-hospital mortality compared with other states.9 These results were consistent with a survey that revealed that interventional cardiologists in New York felt that they were more likely to perform PCI in patients with refractory shock after the exclusion of such patients from public reporting.5

In a similar approach, there have also been efforts to exclude patients with out-of-hospital cardiac arrest from public reporting in an attempt to decrease risk aversion and ensure that patients receive proper care.10 In 2010, New York excluded patients with cardiac arrest and coma. However, data have not shown any resultant changes in PCI or mortality rates.10 Likewise, Massachusetts no longer includes patients with ongoing cardiopulmonary resuscitation at the onset of PCI in public reporting.11

Non-public reporting

An alternative approach to public reporting is non-public reporting.12,13 In this approach, outcomes could instead be shared in a formal system between hospitals within a state (or regionally or nationally) such that physicians could still be held accountable for patient care. This approach would mitigate the public shame component of current reporting initiatives and potentially diminish physician risk aversion. As data have shown that patients rarely use public reporting, this non-public reporting policy would have little impact on patient behaviour. While these non-public reporting programs do exist, it is not clear why they are not as prominent as public reporting programs. Perhaps one reason is that the public reporting programs were born out of transparency, which would not be the case for non-public reporting.

Examples of non-public reporting initiatives include the Department of Veterans Affairs’ Clinical Assessment, Reporting and Tracking tool (www.hsrd.research.va.gov/for_managers/stories/cart-cl.cfm), to improve transparency among Veterans Affairs catheterisation laboratories, and Michigan’s Collaborative Quality Initiatives (CQI), a program in which hospitals and providers in the state can share and compare PCI outcomes and other quality and process measures. Current evidence suggests that CQI is associated with improved PCI outcomes.14 While physicians would continue to face pressure from colleagues under non-public reporting, evidence still suggests that CQI has reduced physician risk aversion, likely due to removal of the public shaming component.15

Disease-based public reporting

Another viable alternative to public reporting is to instead report outcomes based on condition, such as unstable angina, non-STEMI and STEMI, irrespective of whether a patient has received PCI. The care of a patient undergoing PCI is multifactorial and not solely dependent on the quality of PCI but also on the many aspects of pre- and post-procedural care by emergency medicine, critical care and inpatient teams. By reporting disease-based outcomes, risk aversion based on procedures could be prevented. Furthermore, disease-based reporting would improve transparency for all members and teams taking care of the patient, rather than just the proceduralist involved in one aspect of patient care.

Conclusion

As general public reporting initiatives have become increasingly common around the world, efforts may be made to spread public reporting specifically to cardiovascular procedures such as PCI. However, two decades of experience in the US suggest that public reporting of PCI is not associated with improved patient outcomes and is instead associated with possible patient harm through risk aversion. Taken together, in an effort to enhance transparency for patients by improving access to information, public reporting has unintentionally caused harm by leading to avoidance of life-saving procedures for patients who may need them the most.

As other countries move to improve quality and accountability in PCI and cardiovascular care, viable alternatives to public reporting include disease-based reporting, which would capture the full spectrum of patient care and reduce risk aversion, or non-public reporting, which would still hold physicians accountable without the unintended consequence of public shame and risk aversion.