By Jeffrey Zimmet, MD, PhD
Associate Professor of Medicine, University of California, San Francisco, Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center
Dr. Zimmet reports no financial relationships relevant to this field of study.
SOURCE: Waldo SW, et al. Association between public reporting of outcomes with procedural management and mortality for patients with acute myocardial infarction. J Am Coll Cardiol 2015;65:1119-1126.
Public reporting of clinical outcomes associated with cardiovascular care has been touted as a means to improve health care delivery and patient outcomes. Three states (New York, Massachusetts, and Pennsylvania) have instituted reporting of outcomes associated with percutaneous coronary intervention (PCI), and additional states are considering this practice. Much has been discussed regarding the potential costs of such measures, but little data thus far have supported putative harms of public reporting.
In the recent issue of the Journal of the American College of Cardiology (JACC), Waldo and colleagues add substantially to this debate. They mined the Nationwide Inpatient Sample for patients with a discharge diagnosis of acute myocardial infarction (AMI) in Massachusetts and New York between 2005 and 2011, restricting the data set to hospitals with on-site PCI. These data were compared with those from geographically similar states without public reporting. Demographics were available for each patient, as were high-risk characteristics such as cardiac arrest and cardiogenic shock. Revascularization procedural management and in-hospital mortality were compared between reporting and non-reporting states.
After multivariate adjustment, patients presenting with AMI in public reporting states were significantly less likely to undergo percutaneous coronary revascularization than those in non-reporting states (odds ratio [OR], 0.81; 95% confidence interval [CI], 0.67-0.96). This difference was enhanced among patient subsets with higher-risk profiles, including those over age 65 (OR, 0.75; 95% CI, 0.62-0.91), those presenting with ST elevation (OR, 0.63; 95% CI, 0.56-0.71), and those with cardiac arrest or cardiogenic shock (OR, 0.58; 95% CI, 0.47-0.70). Rates of surgical revascularization by coronary artery bypass graft surgery (CABG) was similar between reporting and non-reporting states. There was no difference between the two public reporting states with regard to likelihood of undergoing PCI.
In-hospital mortality among patients in the study group presenting with AMI was 6% overall; breaking this down by reporting status, the mortality rate was 6% in reporting states vs 5% in non-reporting states. The adjusted risk for in-hospital mortality was consequently higher for patients in reporting states than for those in non-reporting states (OR, 1.21; 95% CI, 1.06-1.37). The mortality difference was most prominent in those presenting with non-ST segment elevation myocardial infarction (NSTEMI).
When broken down by whether or not patients underwent percutaneous revascularization, the adjusted risk for mortality was actually lower in public reporting states among patients who underwent PCI (OR, 0.71; 95% CI, 0.62-0.83) and higher in those who did not (OR, 1.30; 95% CI, 1.13-1.50; P for interaction < 0.001). The authors interpreted this in the following way: Public reporting changes operator behavior such that critically ill patients are offered percutaneous intervention less often. This results in a lower mortality for those patients who actually undergo PCI, a higher mortality among those managed without intervention, and a higher mortality overall.
At the most basic level, public reporting for procedures is designed to alter physician and hospital system behavior in ways that improve patient outcomes. When considering procedures such as PCI, it is clear that the patients at greatest risk for mortality are those presenting with AMI as opposed to elective PCI patients. When the mortality numbers are relatively low to start with (in-hospital mortality was reported at 6% for AMI patients in the current study), every unfavorable outcome has the potential to adversely affect one’s statistics. In this light, the results of the current study offer no surprises.
In states with public reporting, as compared with those that do not engage in reporting, patients with AMI are less likely to undergo potentially lifesaving percutaneous revascularization. This effect is most striking in those with higher-risk profiles, and is greatest among those who are most likely to die. Unfortunately, these are the same patients who are most likely to benefit — patients who have experienced cardiac arrest or cardiogenic shock. The observed effect on mortality is precisely what one might predict a priori: that reporting results in a reduction in PCI-related mortality, primarily by shifting those deaths to the pool of patients who do not undergo procedures. That patients who are most in need of revascularization might be denied these procedures is clearly not what was intended with reporting programs. States that are considering adding public reporting for PCI should read this study carefully and give careful consideration to the potential for unintended consequences.