Dr James Januzzi: AI Gives Us the Chance to Explore Care-Improvement Opportunities

We want to use artificial intelligence (AI) to promote the best-case scenario in patient care, emphasized James Januzzi, MD, staff cardiologist at Massachusetts General Hospital.

We want to use artificial intelligence (AI) to promote the best-case scenario in patient care, to optimize medical therapy to overcome gaps in care and inadequate treatment adherence, and to support—not replace—clinical judgement at the point of care, emphasized James Januzzi, MD, Hutter Family Professor of Medicine, Harvard Medical School; staff cardiologist at Massachusetts General Hospital; trialist at the Boehm Institute for Clinical Research; and trustee of the American College of Cardiology.

In the TRANSFORM family of studies, which he references below, HealthReveal will use its personalized AI solution to continuously monitor patients and, in turn, apply the most recent care guidelines. This will enable clinicians to make real-time, as-needed treatment recommendations.

Transcript

Why do so few patients with heart disease get the correct drug interventions?

Factors that determine inadequate adherence to optimal medical therapy for patients are numerous. Clinical inertia is something that we often talk about—but that’s only one part of the problem. There are so many things that a busy clinician must attend to within the 20-minute visit with their patients. Staying abreast of new developments, new guideline recommendations, new drugs, new means by which to apply older drugs all tend to result in a gap in the up-to-date adherence to what we believe to be optimal guideline-directed medical therapy.

So, in a way, inertia and the other various reasons for treatment gaps are very intertwined. Means by which to address not only clinicians recognizing what therapies they should be prescribing, but also keeping them up-to-date with new developments and new means by which to apply older drugs is one way that we hope to address in the TRANSFORM studies.

What challenges and skepticism exist toward the use of AI in certain patient populations?

Some skepticism exists regarding the role of artificial intelligence for guiding decision-making in clinical medicine, and this speaks to the complexity of medical treatment for patients. It’s not just book knowledge, it’s not just what’s on paper, but also understanding the socioeconomic factors that patients are facing, the personal issues that may come along in each individual case. And so we clearly recognize at the American College of Cardiology that there will always need to be a clinician at the point of care seeing patients, taking care of patients, and that’s where there’s an optimal opportunity to evaluate how AI-leveraged approaches can support clinical judgement at the point of care rather than replacing it.

To be sure, evidence clearly points to the fact that gaps exist in application of optimal therapies for various heart disease states, such as heart failure. And so what we believe will be an optimal balance will be the combination of the information provided by artificial intelligence, which, by the way, is leveraging the knowledge articulated in the American College of Cardiology Expert Consensus Decision Pathway Documents. It’s not as if the AI is thinking independently on its own; it’s essentially speaking the consensus and guideline recommendations that are the bedrock of optimal medical care. But that’s used together with clinical judgement as well.

How will using AI to help develop guidelines differ from traditional clinical guideline development?

AI, as a component to optimal medical therapy, is best thought of as promoting the best-case scenario for patient care. Of course, there’s going to be differences of opinion at the point of care—because there may be unmeasured variables that may influence a clinician’s decision-making—but when we examine, again and again, the difference between situations where pathway-delivered care following essentially the logic that artificial intelligence would vs usual care, repeatedly we see that usual care falls short. So there’s an opportunity to accept our limitations and explore opportunities to improve the care that we deliver to our patients, which is so critically, critically important.