AI-powered ECG screening boosts patient outcomes—when clinicians give it a chance
Artificial intelligence (AI) has the potential to be a true game-changer for patient care—but it can only reach that potential if physicians and other healthcare professionals give it a chance.
New research published in Mayo Clinic Proceedings, for example, confirmed that using an AI-powered solution to screen electrocardiograms (ECGs) can improve the diagnosis and treatment of patients with low left ventricular ejection fraction (EF). Clinicians who most frequently followed the AI tool’s recommendations were much more likely to identify patients with low EF.
The study focused on more than 22,000 patients who underwent an ECG from August 2019 to March 2020. Clinicians from 48 different Mayo Clinic primary care practices participated in the study. They were randomly chosen to either use the AI screening tool or proceed with patient care like normal.
The clinicians who were selected to use the AI tool ordered more than 11,500 ECGs. The mean patient age was 60.5 years old, and 53.9% were women. The AI tool screened each ECG, displaying a “negative” or “positive” result. Negative results meant no further testing was recommended, but a positive report indicated an echocardiogram was recommended. Clinicians were viewed as a “high adopters” of the AI technology if they ordered echocardiograms at least 64.3% of the time it was recommended
Overall, high adopters were twice as likely to successfully diagnose left ventricular EF than low adopters.
“It was surprising to see the significant difference in the rate of diagnosis between high adopters and low adopters,” lead author David Rushlow, MD, a Mayo Clinic physician, said in a prepared statement. “The tool is extremely helpful, but we did not expect to see a full doubling of the diagnosis rate of low EF as compared to low adopters.”
Rushlow et al. also found that clinicians who were the most likely to be a high adopter were the ones with less experience treating complex patients.
“This demonstrates the importance of AI systems that integrate seamlessly into the workflows of clinicians,” Rushlow said. “Given the technical nature of AI in healthcare, it often is initiated and developed in academic specialty practices. To maximize AI's benefits, more collaboration is needed between specialty practices and primary care.”
The AI tool used for this analysis was designed and developed by Mayo Clinic specialists, though Mayo Clinic “will not benefit financially from its use in the care of patients at Mayo Clinic.” Some of the study’s co-authors—but not Rushlow—do have a financial connection to the AI tool.
Michael has more than 18 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.