Home > Cardiology > Mobile device identifies patients at risk of sudden death from QT prolongation

Mobile device identifies patients at risk of sudden death from QT prolongation

Journal
Circulation
Reuters Health - 09/02/2021 - An artificial intelligence (AI)-enabled mobile device shows accuracy similar to standard 12-lead electrocardiography (ECG) for detecting heart-rate corrected QT interval (QTc) prolongation secondary to drugs, genetics or systemic diseases such as COVID-19, researchers say.

The findings could provide a cost-effective way to screen for long QT syndrome (LQTS) in settings where a 12-lead ECG is not accessible or cost-effective, they suggest.

"Right now, physicians go into the room knowing the patient's height, weight, and blood pressure,' Dr. Michael Ackerman of Mayo Clinic in Rochester, Minnesota told Reuters Health by email. "I would argue that knowing their patient's QTc as another vital sign has far greater implications overall for that patient encounter than having the patient's weight documented."

The current study and the team's QT-related materials "are now being packaged by AliveCor to be submitted for FDA clearance by the end of the year," he said. "We anticipate the indication will be used for detection of a prolonged QTc in any clinical situation where a conventional 12-lead ECG is not practical."

"In March 2020, the device had received Emergency Use Authorization to provide QTc monitoring in the midst of the COVID-19 pandemic when there was widespread use of the QT-prolonging medications hydroxychloroquine and azithromycin," he added. "So, the FDA is aware of these advances in being able to have the AliveCor KardiaMobile 6L device also function as a very accurate 'QTc meter.'"

As reported in Circulation, Dr. Ackerman and colleagues used more then 1.6 million 12-lead ECGs from over half a million patients to train (251,000 patients), test (108,000) and validate (180,000) an AI-based deep neural network to recognize and assess the QTc.

They then compared results from the prototype mobile device - which is similar to the commercially available AliveCor device - with 12-lead ECGs from 686 genetic heart disease patients, half of whom had congenital LQTS.

In the validation sample, there was strong agreement between the AI-predicted QTc values and assessments by a cardiologist (-1.76 ms).

Similarly, within the genetic heart disease dataset, there was only a minimal difference between the mobile device's predicted QTc values and 12-lead ECGs assessed by a QT expert (-0.45) and a commercial ECG laboratory (+10.52 ms).

Overall, the device's ability to detect a QTc value of 500 ms or more yielded an area under the curve of 0.97, sensitivity of 80%, and specificity of 94.4%.

Dr. Ackerman said, "Many effective medications, from anti-cancer to psychiatric, have a QT prolonging side effect, and this technology will help physicians to keep their patients safe. The QTc has always been a vital sign, in my mind. Now, we have the tools to readily obtain this vital sign and act upon it."

Dr. Luigi Di Biase, Section Head Electrophysiology and Director of Arrhythmia Services at the Montefiore Hospital System and a professor at Albert Einstein College of Medicine in New York City, called the device's area under the curve, sensitivity and specificity "encouraging."

That said, he added, "Patients should be informed not to panic about an abnormal measurement. They need to make sure they obtain further ECG checks with an electrophysiologist or cardiologist."

"What concerns me would be the number of false negatives," he concluded.

Dr. Ackerman and Mayo clinic are involved in an equity/royalty relationship with AliveCor and one coauthor is founder of the company.

SOURCE: https://bit.ly/3jwg8b5 Circulation, online February 1, 2021.

By Marilynn Larkin



Posted on