Individual feedback key to reducing excessive ECGs
An educational intervention designed to reduce ordering of transthoracic echocardiograms (TTEs) informs physicians about their own test-ordering patterns and how their performance compares to their peers, according to recent research. Such knowledge can reduce inappropriate use of the tests.
The study was published online Aug. 21 in the Journal of the American College of Cardiology.
“This may prove to be an effective strategy to improve the use of imaging,” wrote the authors, led by R. Sacha Batia, the Women’s College Hospital Institute for Health Systems Solutions and Virtual Care at Women’s College Hospital in Toronto.
The study cohort included 179 physicians working in ambulatory care settings at 8 hospitals (7 in Ontario, Canada and 1 in Massachusetts). Altogether the researchers analyzed 14,697 TTEs that had been ordered between Dec. 1, 2014, and April 17, 2016. Most of the participating hospitals were large academic medical centers; one smaller ambulatory center and one rural hospital were included.
In contrast to this large, randomized study, most previous investigations of appropriate versus inappropriate ordering of cardiology imaging tests have been small, single-center, nonrandomized studies.
Physicians in the intervention group (1) listened to a 20-minute video lecture about correct ordering of echocardiograms, (2) were offered a downloadable decisions support app from the American Society of Echocardiography, and (3) received monthly feedback reports about the number of inappropriate, maybe-appropriate and rarely-appreciate test they had ordered, alongside data on the performance of their peers at their hospital.
“The mean proportion of rarely appropriate TTEs was significantly lower in the intervention versus the control group (8.6 percent vs. 11.1 percent),” the authors found. They concluded, “This type of intervention could be considered as a strategy to curb the use of low-value imaging services.”
In an accompanying editorial, Randolph P. Martin, with Emory Medical School in Atlanta, pointed out certain limitations of the study, including the fact that there appeared to be no lasting effect of the educational intervention after the study ended. Also, only a small percentage of the physicians in the intervention group checked results monthly.
To ensure that tests are consistently ordered only when appropriate, Martin suggested three steps. First, he said that offering ordering physicians feedback and comparisons of their own performance with peers must be an ongoing activity.
Second, he said that such efforts ought to focus on reducing the rarely appropriate TTEs that are ordered for the situations that account for the most unnecessary tests.
“The top three culprits consisted of 1) routine surveillance of prosthetic heart valves that have been in place for less than 3 years, if there was no known or suspected dysfunction, 2) routine surveillance of ventricular function in patients with known coronary heart disease and no change in clinical status for cardiac examination, and 3) routine surveillance carried out within a one-year period in patient with moderate to severe valvular stenosis without a change in the clinical status or cardiac examination,” he wrote.
Third, he called for artificial intelligence and machine learning to be brought to bear in the ordering systems of electronic medical records. The computers could alert clinicians to guidelines, he said, and intervene so that only appropriate tests are ordered.