Diagnostics, Artificial Intelligence

Startup says algorithm identifies new risk factor for heart disease but a doctor is skeptical

The algorithm — which measures variations in heart rate — is designed to predict heart problems, but a cardiologist wonders whether it can outperform other risk factors.

An Israeli biotech company recently said that a study has endorsed what it describes as a new risk factor for heart disease as analyzed by the company’s proprietary algorithm.

The predictive power of the factor – variation in heart rate – was borne out by a clinical study led by Dr. Ilan Goldenberg, director of clinical cardiovascular research at the University of Rochester Medical Center.

The algorithm was developed by Lev El Diagnostics of Heart Diseases, a biomedical company based in Jerusalem. Its product, HeartTrends, listens to heartbeats for 20 minutes through a wearable monitor connected to an app. The app sends data through the cloud back to Lev El for analysis of heart-rate variation. It can work on any monitor and does not require a baseline reading first.

The question, however, is whether variation in heart rate is more predictive than existing risk factors, such as age, family history and fitness levels, according to one cardiologist who was not involved in the study.

“I think this is true for any new test, whether it’s to detect heart disease or cancer or cognitive dysfunction,” the cardiologist, Dr. Richard Josephson, said in a phone interview. He is a professor at Case Western Reserve University School of Medicine and director of cardiovascular and pulmonary rehabilitation.

HeartTrends currently is not being used in the U.S., said Dov Rubin, CEO of Lev El Diagnostics. But the algorithm is FDA-cleared. It is currently being used in other countries by life insurers as a way to assess actuarial risk in younger, healthier people, Rubin said, noting it is more convenient and less costly than traditional stress tests, which typically involve measuring heart rate as a person walks on a treadmill or rides a stationary bike.

sponsored content

A Deep-dive Into Specialty Pharma

A specialty drug is a class of prescription medications used to treat complex, chronic or rare medical conditions. Although this classification was originally intended to define the treatment of rare, also termed “orphan” diseases, affecting fewer than 200,000 people in the US, more recently, specialty drugs have emerged as the cornerstone of treatment for chronic and complex diseases such as cancer, autoimmune conditions, diabetes, hepatitis C, and HIV/AIDS.

Rubin expects to gain a U.S foothold through life insurers but sees a bigger role in healthcare for the algorithm, which his company has been developing for nearly eight years.

“The vision is it should be part of your annual health exam,” Rubin said in a phone interview.

The company’s website pitches its test as ideal for healthy adults with at least one risk factor or who are unable to undergo a traditional stress test. It might be similar to the Apple Watch’s apparent ability to detect atrial fibrillation or an irregular heartbeat, Rubin said.

“We would like to follow in that because this is really a non-threatening type of test,” he said. “This is an early warning. This is not something that says we have to go to surgery tomorrow.”

A typical person’s heart rate varies but the variation declines when coronary arteries start to clog, a condition known as myocardial ischemia, Goldenberg said. The variation can start to decline even in people who appear healthy otherwise.

The results of the study, the third on the HeartTrends algorithm, were presented in mid-November at an American Heart Association scientific conference in Philadelphia and are slated for publication in the association’s journal. The study looked at 1,043 people who had no known coronary artery disease. Participants were enrolled in the study between 2015 and 2018 either through the Mayo Clinic or at Sheba Medical Center in Israel. The average age was 59, and 38% were women. Of the total, 6.3% were found to have myocardial ischemia. Goldenberg said his study showed the algorithm was more likely to predict myocardial ischemia than conventional stress tests.

As with any new test, however, the challenge lies in convincing clinicians to use it – and insurers to pay for it.

Josephson acknowledged that heart-rate variability may be a risk marker for coronary artery disease. But, he said, based on what he had seen of the study, it was unclear whether the algorithm truly added to the predictive powers of established risk factors, such as age, fitness and past tobacco use. He also noted that tests of otherwise-healthy people run the risk of generating a disproportionate share of false positives.

Goldenberg said his study of HeartTrends showed that heart-rate variability is an independent predictor of myocardial ischemia in healthy people even after adjusting for traditional risk factors and that it improves diagnostic results when added to those other factors.

Between 30% and 40% of those identified as at risk in the study turned out to be false positives. But taking the test twice should reduce the incidence, as is the case for other tests, he said in an email response to questions. The test, however, was 97% accurate in ruling out myocardial ischemia, he added.

Photo: ismagilov, Getty Images