Health IT

Where do population health and AI intersect? ATA panelists weigh in

There are numerous ways AI is being used at the patient bedside, and even big companies like Google are leveraging the technology in healthcare. But it’ll be some time before machines completely take over for clinicians.

Composite image of a large group of diverse people smiling

Though overhyped, artificial intelligence generates plenty of interest in healthcare. Its applications range from clinical trials to radiology. And at this year’s American Telemedicine Association conference in Chicago, three experts shared their thoughts on how it relates to population health.

Kamal Jethwani, senior director at Partners Connected Health Innovation, outlined a few examples of how his organization is leveraging AI to assist patients.

One is around predictive modeling and using the technology to figure out the chances of a patient returning to the hospital post-discharge. Another effort involves coupling wearable data with medical record data to predict whether a patient will have COPD exacerbations. Additionally, the organization is using AI to predict patient engagement as far as remote monitoring is concerned. The technology “gives us a notification that a person’s on their way down” and “gives us a window of time to intervene,” Jethwani said during an ATA session on April 30.

Though these features enhance the doctor-patient relationship, Clarify Health Solutions founder and CEO Jean Drouin posed an interesting question: “How do you go from the terrific things going on at Partners … and make those available across the industry in a way that’s affordable and implementable?”

It’s certainly a challenge, and different organizations are at different points as far as the implementation of these tools is concerned.

The decently advanced level of AI technologies at some healthcare entities has piqued the curiosity of parties who wonder if algorithms will replace clinicians. But Jethwani doesn’t think so.

“AI Has the power to equalize the decision-making process. With AI, we have the opportunity to move that decision-making process lower down the chain,” he said. “That’s going to democratize the process a little bit. I don’t think it’s going to become completely automated.”

Alexandra Pelletier, senior manager of global innovation at Deloitte, said this application of AI isn’t extremely different than the classical decision support tools we’ve seen thus far.

“When these things really take root, the system might transform and actually look different,” she said.

Plus, there are bigger fish to fry. “Before we get to the machines taking over, we have some really thorny trust issues around data that we have to solve first,” Pelletier added.

Applying AI to population health is such a hot topic that big-name companies like Google are making moves.

As Drouin mentioned, the tech giant is working to map patients across cardiac care, as well as a few other issues. “They are unique in having the resources to be able to do that,” he said.

Pelletier referenced the recently released CB Insights analysis regarding Google and mentioned she believes Amazon is going to leverage AI in the pharmaceutical realm.

Editor’s note: Pelletier later emailed to correct her comments regarding Google and Amazon. At ATA, she intended to say she believes Amazon (not Google) will leverage AI in the pharmaceutical space. This article has been updated to reflect the change.

Photo: PeopleImages, Getty Images