Underuse of Echocardiography, Particularly in AMI, Linked to More In-hospital Deaths
Despite mounting use of echocardiography in the United States—leading to concerns the modality may be overused—numbers from the Nationwide Inpatient Sample (NIS) suggest that echo is actually underused in a number of guideline-recommended settings, including acute MI. Moreover, use of echo for specific high-risk conditions is associated with significantly reduced mortality, according to Alexander Papolos, MD, of Mount Sinai Medical Center (New York, NY), and colleagues.
The findings imply that even greater use of echocardiography might help improve survival following hospitalization for a range of critical cardiovascular conditions.
Speaking with TCTMD, senior author Partho Sengupta, MD, also of Mount Sinai, acknowledged he was initially “surprised” by the results, given that most of the focus on echocardiography in recent years has been on its growing volumes and, with that, concerns that of inappropriate use.
Echo Use in 5 Critical Cardiac Conditions
For their analysis, published in the February 9 issue of the Journal of the American College of Cardiology, Papolos et al identified 5 cardiovascular diagnoses for which echo is most commonly recommended: acute MI, cardiac arrhythmia, acute stroke, congestive heart failure, and sepsis. Using the NIS database, which represents 20% of US community hospitals, the investigators then looked at inpatient echo use for these conditions between 2001 and 2011 and linked it with all-cause in-hospital mortality.
Over the decade, use of echo increased steadily by approximate 3% per year for a total of almost 7.7 million echo exams. Echo was used in 22% of admissions for valvular disease and in 12% for stroke, but in just 6% of both AMI and CAD admissions. Coronary catheterization, not surprisingly, was used in a much higher proportion of AMI and CAD patients at 64% and 62%, respectively.
Looking specifically at the year 2010, echo use was associated with significantly lower in-hospital mortality for all 5 conditions of interest. Of note, these conditions accounted for a full 3.7 million hospital admissions, yet echo was only used in 8% of cases, despite appropriate use criteria backing echocardiography in these conditions.
Surprised by the numbers, Papolos et al elected to repeat the analysis at their own institution. They found that echo use had increased at their hospital at an average annual rate of 4.75% between 2003 and 2014. Moreover, 25% of all AMI patients were discharged without an echo exam and 17% without any type of cardiac imaging. Mortality was “marginally” higher among patients who did not undergo echocardiography, although the difference did not reach statistical significance.
Claims of Overuse ‘Have Never Been Substantiated’
In an accompanying editorial Christine Jellis, MD, and Brian Griffin, MD, of the Cleveland Clinic in Ohio, detail the expansion of multiple modalities of echocardiography—particularly transthoracic echo and more portable and hand-held machines—over the last decade.
“The relatively low cost, absence of radiation, and ability to perform studies at the bedside has placed echocardiography at a significant advantage over other techniques including cardiac CT, MRI, and nuclear medicine,” they observe. Given echo’s rapid rise, “there has been a perception that it is overused and potentially employed inappropriately in many instances,” Jellis and Griffin write, noting that “these claims have never been substantiated.”
On the contrary, the new data, while retrospective and observational, “should be a wake-up call” for physicians, Sengupta said, particularly for interventionalists who may elect to skip the post-PCI echocardiogram in their acute MI patients, assuming echo will be performed on follow-up.
The problem with that strategy, Sengupta continued, is that LV remodeling due to low ejection fraction following MI begins very early. “If you perform the echo and you see that the patient has LV dysfunction, you would start them on an ACE inhibitor or ARB, and that’s very important in the process of remodeling,” he said. “But if you don’t know the ejection fraction because you didn’t do the echo, you might not start them immediately on these drugs, and that’s when problems happen.
Sengupta acknowledged that the paper does not definitively identify a mechanism linking echo use to mortality. It may be that performance of guideline-supported echocardiography during the index hospitalization is a marker for overall better care that in turn leads to improved patient survival.
But his hunch is that in-hospital mortality, in most cases, is related to the reason the patient was first admitted. “Let’s assume that’s the case,” he argued. “Now that morbidity, that process—LV function, arrhythmia, or some catastrophic problem—had we done an echo and discovered the structural or functional anomaly, that would have led to a difference in the management of that patient. This is been previously shown: when you perform an echo, you impact on the care of that patient.”
Both the authors and the editorialists call for randomized trials to substantiate the findings of these observational data but agree that the findings point to underutilization. “This paper reminds us that underutilization of safe, effective technologies such as echocardiography may also have a broad economic impact,” the editorialists conclude. “Healthcare strategies that may limit their utilization should be subjected to clinical trials and the cleansing light of actual data.”
1. Papolos A, Narula J, Bavishi C, et al. U.S. hospital use of echocardiography: insights from the Nationwide Inpatient Sample. J Am Coll Cardiol. 2016;67:502-511.
2. Jellis CL, Griffin BP. Are we doing too many inpatient echocardiograms? The answer from Big Data may surprise you [editorial]! J Am Coll Cardiol. 2016;67:512-514.
- Post-PCI Stress Testing Trends Downward
- Updated Consensus Statement Outlines Appropriate Cardiac CT Use
- Curbing Overuse of Cardiac Imaging Requires Audits and Feedback, Analysis Shows
- Sengupta reports serving as an advisor to Saffron Technology Hearts Labs and as a consultant to Edwards Lifesciences.
- Jellis and Griffin report no relevant conflicts of interest.