Author + information
- Steve Miner, MD∗ ( and )
- Lynne Nield, MD
- ↵∗Southlake Regional Health Center, University of Toronto, Newmarket, Ontario, L3Y 2P9 Canada
Sherwood et al. (1) report that groups that perform more high-risk percutaneous coronary interventions have similar risk-adjusted mortality to those who perform fewer. The data are interesting, but a number of limitations preclude the final conclusion that adopting a more aggressive practice pattern will not increase risk-adjusted mortality.
Subgroups do not have uniform risk. Patients in cardiogenic shock have mortality rates that range from 22% to 88% (2). Physicians preferentially treat patients at the lower spectrum of risk and thus will have lower observed mortality than predicted by risk scores. At the same time, motivated practitioners have an incentive to “up-code,” which artificially inflates the estimated risk. The fact that in Sherwood et al. (1), “high-risk” cases tended to have lower-than-expected mortality is consistent with these limitations.
The conclusion also assumes that risk-averse operators are as adept as those who regularly perform high-risk cases. One of the benefits touted for public reporting is that it directs higher-risk cases towards superior operators (3).
A final limitation is the exclusion of patients who receive angioplasty at one site, but are then transferred to a different site. This excludes high-risk patients and procedural complications that might significantly alter the final results.
These limitations were not present at a Canadian regional care center, free of the medico-legal and public reporting concerns of the United States. In this setting, regional efforts to more aggressively treat high-risk myocardial infarction patients led to an increase in risk-adjusted mortality despite evidence for preserved procedural quality (4).
This debate also distracts from the more important issue. The real question is whether risk aversion related to public reporting results in public harm. This analysis must include, not only those who undergo the procedure, but also those in whom the procedure was deferred. In 2012, Joynt et al. attempted to answer this question by exploring the outcomes of all patients with myocardial infarction in states that adopted public reporting, compared with those that did not. They found that in states with public reporting, mortality rates were significantly higher for patients presenting with ST-segment elevation myocardial infarction (p = 0.004) with a trend toward worse outcomes for the larger cohort of all patients with myocardial infarction (p = 0.10). More recently, this same approach was applied to a much larger population, revealing a dramatic 21% increase in mortality for patients presenting with myocardial infarction in states with public reporting (p = 0.013) (5). This was driven primarily by an increase in mortality in patients in whom intervention was deferred. With these results, we must conclude that public reporting of procedural outcomes results in public harm.
We applaud Sherwood et al. (1) for their efforts. At the same time, we wonder whether the time has come to move away from procedure-based risk scores and toward diagnosis-based databases that examine the outcomes of all patients, not just those subgroups selected to undergo specific procedures.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Sherwood M.W.,
- Brennan J.M.,
- Ho K.K.,
- et al.
- Waldo S.W.,
- McCabe J.M.,
- O'Brien C.,
- Kennedy K.F.,
- Joynt K.E.,
- Yeh R.W.