Author + information
- Duane S. Pinto, MD, MPH∗ ( and )
- Yuri B. Pride, MD
- Division of Cardiovascular Medicine, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts
- ↵∗Reprint requests and correspondence:
Dr. Duane S. Pinto, Beth Israel Deaconess Medical Center, 185 Pilgrim Road, W/Baker 4, Boston, Massachusetts 02215.
New York mandated open access to data regarding outcomes associated with percutaneous coronary intervention (PCI) in 1991, and in the subsequent 2 decades, public reporting has been a divisive topic (1). Over that time, Pennsylvania and Massachusetts have also instituted public reporting of PCI outcomes. The Center for Medicare and Medicaid Services reports 30-day hospital mortality and readmission rates for acute myocardial infarction in addition to a myriad of patient-centered healthcare outcomes (2).
Public reporting of PCI outcomes as envisioned by providers, patient advocacy groups, and payers has been implemented with the intent that patients would make educated decisions about where to get their healthcare, and providers would make practice improvements and invest in systems of care. Furthermore, hospitals and providers are more apt to rapidly adopt quality improvement measures when outcomes are publicly reported. Such transparency certainly strengthens the trust between patient and provider and serves to reduce avoidable patient harm. Finally, the development of robust PCI registries may result in future improvements in determining the cost-effectiveness and appropriateness of PCI. However, measurement of outcomes, enthusiastic efforts, and admirable goals should not be equated with success. Implicit in public reporting is that the outcomes measures should be reliable and ultimately useful in improving patient outcomes.
Many involved physicians, administrators, nurses, statisticians, and data auditors are well aware of the inherent limitations of incomplete and inaccurate data collection, opportunities for covariate manipulation, and the unintended consequences of public reporting. Most agree that patients deserve to know the quality of care they are likely to receive, are interested in comparisons with other providers and hospitals, and want to improve patient outcomes. Most also recognize that inspecting a few mortality statistics does not accomplish this goal and is much like trying to read hieroglyphs without the Rosetta Stone—interesting to look at, but the uninitiated cannot make much sense of them.
Moreover, it is evident that hospital systems and PCI providers are keenly aware of the pressure of public reporting. Public reporting has the unintended consequence of shifting the paradigm from “Let's give them a shot” to “It probably won't make a difference.” Whether this shift is beneficial or not remains to be seen, and declining PCI mortality rates spuriously created by risk avoidance and careful data capture among ill patients reinforces this fallacy. Because of the time pressures inherent with PCI among patients with ST-segment elevation myocardial infarction (STEMI), cardiogenic shock, and cardiac arrest, decisions of whether to perform PCI are often based on incomplete clinical information, and interventionalists only have a partial inkling of the probability of survival in an individual. A growing body of literature suggests that in states with public reporting, patients presenting with STEMI and cardiogenic shock—those likely to gain the most from effective, rapid reperfusion via primary PCI—are less likely to receive it than are patients in states without public reporting systems (3,4).
If the intent of public reporting is for consumers to choose care appropriately, it seems counterintuitive that PCI mortality rates would be helpful in this regard because: 1) most PCI mortality occurs in patients presenting with emergent conditions who will have no choice where to get their care; 2) mortality following elective PCI is so rare that random chance may haunt highly skilled interventionalists and benefit those who perform poorly (5); 3) the data is outdated when it finally becomes public so consumers will not be making choices based on current levels of hospital and/or provider performance; and 4) not all deaths within 30 days of PCI are complications of PCI.
To that end, in this issue of the Journal, Aggarwal et al. (6) searched the Cleveland Clinic's institutional PCI registry over a 37-month span and determined the cause of death among all those who died within 30 days. What they found will not surprise interventionalists but should give pause to those who scrutinize 30-day mortality as the endpoint of choice when determining quality of a hospital or provider.
Of the 4,078 patients who underwent PCI, there were 81 (2%) who died within 30 days. Thirty-day mortality among patients who presented with cardiac arrest (41%), cardiogenic shock (32%), and STEMI (7%) was markedly higher, and those patients accounted for the majority of deaths. The investigators used an inclusive definition of PCI-relatedness, which they defined as death from vascular dissection, aneurysm, perforation, bleeding, renal failure, and definite or probable stent thrombosis, and with this definition, less than one-half (42%) died of PCI-related causes. Of the 34 patients who died of PCI-related cases, only 9 (0.2% of all patients undergoing PCI) died of causes other than probable or definite stent thrombosis, meaning the majority of PCI-related deaths were unlikely to be complications of the intervention itself or sequelae of the cardiac catheterization (7).
This important work highlights a major pitfall of public reporting: the misclassification of the cause of death and its relationship to PCI. There is a substantial desire to define a patient's entire hospitalization, illness, and outcome by the fact a PCI was performed, but this is simply not the case. No amount of statistical modeling can account for the variability in baseline patient risk among the small group of shock and STEMI patients or for the unmeasured confounding present in the few deaths occurring after elective PCI. Whereas death may seem to be an unequivocal outcome, Aggarwal et al. (6) highlight that many patients die in spite of PCI rather than from PCI. The problem of misattribution and classification is further exacerbated in places such as Massachusetts where only in-hospital events are collected so deaths among patients transferred to another hospital or to hospice are not attributed correctly if at all.
The study by Aggarwal et al. (6) suggests that reported PCI mortality does not reflect the actual quality of PCI procedure, and death certificates do not accurately codify the cause of death. If the intent is to improve patient outcomes, then the focus should not be on “the PCI” but rather on the complete system of care of which deploying a stent in an artery is only a small facet in some cases. Given that the majority of deaths after PCI occur due to causes unrelated to the PCI, emphasizing PCI mortality statistics to guide quality improvement is certainly a missed opportunity.
Public reporting has certainly brought about positive change. There is no doubt that it has forced us to examine ourselves and our quality assurance programs, adhere more strictly to guideline-based care, and examine ways in which we can reduce avoidable patient harm. Despite these benefits, however, the implementation of public reporting remains a work in progress, one that we must all take seriously and help to improve. The work of Aggarwal et al. (6) points out perhaps the most significant flaw: the outcome that is publicly reported has less to do with the operators or hospital systems than the public likely thinks it does.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Resnic F.S.,
- Welt F.G.
- ↵Center for Medicaid and Medicare Services. Hospital Compare. Available at: http://www.medicare.gov/hospitalcompare/?AspxAutoDetectCookieSupport=1. Accessed May 20, 2013.
- Moscucci M.,
- Eagle K.A.,
- Share D.,
- et al.
- Aggarwal B.,
- Ellis S.G.,
- Lincoff A.M.,
- et al.
- Cutlip D.E.,
- Nakazawa G.,
- Krucoff M.W.,
- et al.