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.
SYNOPSIS: Since the exclusion of cardiogenic shock from public mortality reporting in New York in 2006, rates of intervention in these patients have risen dramatically. However, these rates remain below those in non-reporting states, suggesting continued reluctance to treat the highest-risk patients due to public reporting.
SOURCES: McCabe JM, Waldo SW, Kennedy KF, Yeh RW. Treatment and outcomes of acute myocardial infarction complicated by shock after public reporting policy changes in New York. JAMA Cardiology. doi:10.1001/jamacardio.2016.1806.
Bangalore S, Guo Y, Xu J, et al. Rates of invasive management of cardiogenic shock in New York before and after exclusion from public reporting. JAMA Cardiology. doi:10.1001/jamacardio.2016.0785.
Today’s healthcare environment is marked by efforts to improve quality and increase transparency, especially in regard to invasive procedures. In the early 1990s, New York became the first state to publicly report mortality figures following percutaneous intervention (PCI), and it continues to do so at an individual-operator level. Amid concerns that public reporting could lead PCI operators to avoid the highest-risk patients, the New York State Department of Public Health ultimately chose to exclude patients with cardiogenic shock from its publicly reported analysis of PCI. This began on a trial basis in 2006-2007 and became permanent in 2008. Two recently published reports examined the results of this policy change.
The analysis by McCabe et al identified patients presenting with acute myocardial infarction (MI) and shock from statewide databases between 2002 and the end of 2012. Massachusetts, Michigan, New Jersey, and California data were used as comparators. Across all states, the number of patients receiving PCI for cardiogenic shock increased when comparing the baseline reference period of 2002 through 2005 to the post-policy change period from 2006 through 2012, and in-hospital mortality also decreased. In New York specifically, the authors noted an increase in the use of coronary angiography (63.6% vs. 67.9%; P < 0.001), an increase in the use of PCI (30.5% vs. 39.7%; P < 0.001), and a decrease in the rate of in-hospital death (47.1% vs. 35.5%). After the reporting policy changes, operators in New York were 28% more likely to perform PCI on patients with acute MI and cardiogenic shock than they were previously (adjusted relative risk, 1.28; 95% confidence interval, 1.19-1.37; P < 0.001). By comparison, operators in the comparator states only were 9% more likely to perform PCI in the later era relative to the former. The “difference in differences” also was statistically significant (P < 0.001 for interaction). These data suggest that the policy change had the desired effect of reducing avoidance of these high-risk cases, allowing more patients to be offered PCI. However, the data also show that New York began with lower rates of PCI and higher mortality for cardiogenic shock compared to other states. Even after the policy change, New York continued to lag behind other states in terms of rates of revascularization for cardiogenic shock throughout the study period. The authors interpreted this finding as indicative of “continued risk aversion on the part of PCI operators in a public reporting environment.”
In a separate analysis, Bangalore et al used propensity score matching to identify cardiogenic shock patients with similar baseline characteristics in New York and Michigan. The matched cohort contained 1,063 patients in New York and 1,063 patients in Michigan. Events were compared among three time periods: the era prior to censoring of cardiogenic shock from public data (2002-2005), the trial censoring period (2006-2007), and the period after permanent exclusion of cardiogenic shock from reporting in New York (2008-2011). An analysis of New York patients revealed a graded increase in the proportion of patients undergoing PCI, invasive management, and revascularization over the three time periods. However, comparison to the propensity-matched cohort showed that during the same periods, a greater proportion of patients underwent PCI, invasive management, or revascularization in Michigan. Sensitivity analyses comparing New York to other non-reporting states, including California and New Jersey, led to similar results. The authors noted that the relatively strict definition of cardiogenic shock qualifying for exclusion leads to other changes in operator behavior geared toward meeting exclusion criteria. For example, they reported that while overall rates of right heart catheterization in MI patients has decreased over time, the proportion of patients with cardiogenic shock who underwent right heart catheterization (one method to qualify patients for cardiogenic shock data censoring) increased significantly after the exclusion of cardiogenic shock from public reporting. As in the first paper, these authors posited that persistent gaps in intervention between New York and non-reporting states may be due to “continued reluctance to perform interventions on high-risk cases.”
On the face of it, examination of outcomes data from invasive cardiac procedures appears to be a good idea. Clearly, it is within the public interest to identify variability in care and to address deficiencies in poorly performing operators and institutions. The cardiogenic shock example illustrates some basic fallacies in the approach of using mortality as the primary metric to measure quality in PCI when patient presentation has such a dramatic effect on outcomes. For example, contemporary figures for mortality in elective PCI hover around 0.6%, while PCI in ST elevation myocardial infarction carries a mortality risk approaching 5%, and in cardiogenic shock averages more than 40%. Conflating these very different scenarios in the evaluation of a common procedure carries obvious objections. These studies show how profoundly such reporting affects operator behavior, often to the detriment of patient care. Although the 2006 reporting change in New York effected a nearly immediate and positive change in the invasive treatment of patients with cardiogenic shock, rates of intervention still lag behind those in non-public reporting states. The patients who are most at risk and have the most to gain from appropriate intervention ultimately are affected by these policies.