AI Is Changing the Landscape of Cardiovascular Care

In the realm of cardiorenalmetabolic disease, saying a patient either has or does not have a condition may simplify their circumstances too much, argued Jaime Murillo, MD, senior vice president and chief medical officer, Medical Affairs, UnitedHealth Group. In many cases, disease severity falls on a spectrum and requires more careful care approaches.

The implementation of artificial intelligence (AI) into clinical settings has enhanced patient screenings and assessments, redefined notions of “disease,” shed light on predictive pathways, and helped guide clinical decisions. In this interview with The American Journal of Managed Care^®, Murillo spoke more to these topics and the value of integrating AI into health care.

This transcript has been lightly edited; captions were auto-generated.

Transcript

What role(s) can artificial intelligence integration play in improving care delivery and outcomes?

Artificial intelligence is going to change the world [of] health care. Period. It's already changing in many ways, but I would submit that we haven't even seen the actual capabilities and extent of what AI can do in health care. Right now, we see efforts like predictive analytics to identify high-risk patients. We see work on what we call CDS, which is clinical decision support, where AI can come in and tell doctors, “Hey, this patient will benefit from a mammogram or a colon cancer screening,” and so on. We also see efforts around the use of AI to reduce administrative burden. For instance, payers use it for claim approvals, not denials. It's important to clarify that AI is not denying, it's approving, the ones that are streamlined.

We also see AI in imaging now. We can detect disease more accurately through AI reading EKGs [electrocardiograms], CT scans, echocardiograms, mammograms, and so on. So there's a lot of opportunity, but there are 2 areas where I see an opportunity for AI. Number 1, bringing personalized care to people, because now you can accumulate tons of data into the patient's information and then give the provider information about the patient that will allow them to be more personalized in terms of their care.

The second area where you see a major opportunity is in redefining disease through the AI. Let me explain that right now, we have an approach that is very binary to health care: someone has diabetes or doesn't have diabetes. The reality is, it's all within the spectrum. Disease goes from all the way normal to a very complicated condition. That spectrum is what AI has an opportunity to define through patterns. If today we say, if you have a hemoglobin A_1C of, let's say, 6.3, you're not diabetic. If it's 6.5, you're diabetic. The reality is that 6.3 may actually have a higher risk than the 6.5 because of a host of other conditions that we're still not familiar with.