A new decision-making tool can help doctors determine if imaging is the most appropriate test for patients with suspected coronary artery disease.
Half of the nearly 200 million people who suffer from CAD initially check in with chest pain. And doctors typically turn to either functional stress testing or CT coronary angiography (CTCA) for their initial analyses.
Despite thousands of clinical trials, including the well-known PROMISE and SCOT-HEART investigations, the choice between the two is critically important yet remains “arbitrary,” researchers explained in a recently published European Heart Journal study.
"There are strengths and limitations for each of these diagnostic tests," Rohan Khera, MD, MS, a Yale School of Medicine cardiologist, said on Thursday. "If you are able to establish the diagnosis correctly, you would be more likely to pursue optimal medical and procedural therapy, which may then influence the outcomes of patients."
Motivated by this problem, Khera and co-investigators developed a new machine learning-backed tool to guide clinicians known as ASSIST.
They trained the platform on data from upwards of 9,500 patients enrolled in the PROMISE or PROspective Multicenter Imaging Study for Evaluation of Chest Pain trial. Khera et al. then embedded data experiments within the clinical trial to test ASSIST, doing the same for 2,135 patients in the SCOT-HEART or Scottish Computed Tomography of the HEART Trial.
Their decision-assist tool proved effective for participants who underwent stress test-first or CCTA-first testing, the authors noted. They reported a two-fold lower risk of adverse cardiac events when the clinician’s test matched that recommended by the AI tool.
Using their platform can help doctors understand why they make the choice between either test and, hopefully, enhance patient outcomes.
"While we used advanced methods to derive ASSIST, its application is practical for the clinical setting,” first author Evangelos Oikonomou, MD, DPhil, a resident physician in Internal Medicine at Yale, added. “It relies on routinely captured patient characteristics and can be used by clinicians with a simple online calculator or can be incorporated in the electronic health record."
Khera is set to present ASSIST, which is part of an overarching concept known as Evidence2Health, at the Yale Innovation Summit on May 19.
Matt joined Chicago’s TriMed team in 2018 covering all areas of health imaging after two years reporting on the hospital field. He holds a bachelor’s in English from UIC, and enjoys a good cup of coffee and an interesting documentary.