For many specialties, health technology provides a way to improve access to care or enhance treatment while a patient is in the office.
The unique challenge of emergency cardiac care is that it greatly benefits from improved foresight. It’s an exciting opportunity for digital health to fill the void, but that has yet to fully materialize, according to John Rumsfeld, M.D., chief innovation officer with the American College of Cardiology.
Using models and sensors for early detection of a heart attack or stroke could provide the missing link between the general public and health systems that care for those patients, he said in an interview with AJMC TV. But he also acknowledged that those technologies, while promising, need a lot more validation in the form of scientific evidence.
“There are already some apps, and there are already some tools for this, and there's, frankly, a lot of hype,” he said. “There’s a lot of, ‘Wow, this is amazing—artificial intelligence is going to predict who is going to have a stroke or heart attack.’ Most of that is unproven. We're still at the beginning, where we need a lot more evidence on using these digital health tools to improve that so-called chain of survival.”
There have been pockets of success in predicting cardiac illnesses. Engineers at Boston University have been working with local hospitals to develop algorithms that can predict hospitalizations associated with heart disease and diabetes with as much as 82% accuracy. A 10-year collaboration between Partners Health Care and GE is focused, in part, on developing AI platforms that can predict a patients risk of stroke.
But there’s still plenty of trepidation about whether digital tools can actually live up to their promise. Renowned surgeon Atul Gawande recently said healthcare is “massively far” from using AI to allow patients to obtain an accurate diagnosis. Stanford University researchers have also identified a number of challenges for digital health innovation in cardiology specifically.