Clinical Decision Support software can
reduce scanning by about six
percent, says a new study

Decision support software could reduce scans by 6 percent: MIT researchers

April 16, 2019
by John R. Fischer, Staff Reporter
Clinical decision support (CDS) software may reduce particular scans by about six percent.

That’s the assertion made in a new, randomized study by researchers at the Massachusetts Institute of Technology (MIT) seeking to address concerns about the overuse of powerful and expensive diagnostic imaging exams.

“Like most of healthcare, we worry about overuse due to third-party payment,” Joseph Doyle, the Erwin H. Schell Professor of Management and Applied Economics at the MIT Sloan School of Management, and one of the authors of the study, told HCB News. “The consumer doesn't pay for the services, and that disconnect can result in overuse. The negative consequences of overuse of high-cost imaging are the direct costs due to billing for the images and radiation exposure, as well as indirect costs from follow-up tests from false positives.”

Incorporating guidelines of the American College of Radiology, CDS informs providers if a test they ordered for a patient is appropriate. These recommendations will be required to be part of imaging orders starting next year, as part of new regulations passed by the Centers for Medicare and Medicaid Services to ensure delivery of reimbursements.

Teaming up with Aurora Health Care, the largest healthcare system in Wisconsin, the researchers distributed ACR Select, a CDS software put out by ACR affiliate National Decision Support Company, to half of 3,511 healthcare providers and compared their use of it for one year to the other half, which relied on the same method it used prior to the trial for determining imaging orders.

Those equipped with CDS saw about a six percent decrease in targeted scans compared to the control group, with the overall number of images ordered remaining unchanged. Four-fifths of overall reductions were in CT scans, which are the most typical, high-cost imaging exams and create the greatest concern for overuse.

The effects continued over time, indicating that CDS alerts can potentially enhance ordering, as well as calm concerns about alert fatigue, a state in which providers are exposed to a large number of alarms and become desensitized to them over time.

Doyle says the study is the first large-scale trial in which physicians and providers were randomly chosen to use CDS for imaging guidance, and that it provides the opportunity to assess how well physicians respond to and utilize information. He adds, however, that reducing scanning requires certain capabilities to be integrated in the CDS software used, and respect from providers.

“We know from the literature and our experience that the system needs an indication (e.g., headache) in order to score the appropriateness of an imaging order; if the order-entry system does not require an indication selection, then the software won't be effective,” he said. “Next, it is important that the ordering providers respect the recommendations. For example, it is important to include a link to the ACR guidelines and the reasoning behind the recommendations. Third, it does seem important to have the system integrated into the order entry system so it is just part of the natural flow of ordering.”

Support was provided by the philanthropy firm Arnold Ventures, and conducted in conjunction with J-PAL North America.

The findings were published in the journal, PLOS ONE.