I work as a cardiologist for a multi-specialty group that sees patients both in capitated payer and fee-for-service environments. This diverse practice gives me the experience to know where insurance providers will find value or what they will likely pay for. As these protocols become more scientifically validated, AI is going to help the entire healthcare system to identify at-risk patients quickly and accurately.
By DR. JONATHAN A. ALIOTA

As a cardiologist, I see this kind of scenario play out on a daily basis. A patient comes in for a routine procedure and a heart murmur is detected. This murmur could have several different diagnoses.
While they can be a sign of something more serious, many of the heart murmurs we hear on a routine basis are normal, physiologic murmurs, such as mild mitral regurgitation. In addition, a healthy heart that had too much caffeine and not enough water that morning may generate what we call a systolic ejection murmur, but it’s not necessarily something more serious, such as aortic stenosis.
We waste a lot of money in the US investigating normal heart murmurs because there is no standardization to evaluating these at the primary care level. If a primary care physician hears a murmur, they may ignore it or they may order an echocardiogram (ECG); it is up to their own clinical judgment,
Anyone who’s been to the hospital knows that receiving care isn’t cheap, and costs can increase quickly when unnecessary medical interventions and tests are ordered as a result of false-positive diagnostic results. Experts estimate that in our $3.4 trillion healthcare system, at least $200 billion is wasted annually on excessive testing and treatment. Overly aggressive or un-indicated care also can harm patients, generating mistakes and injuries believed to cause 30,000 deaths each year.
But, with the emergence of AI, we have the power to do a better job at getting the right patients the right care at the right time. That could save the healthcare system billions.
AI can help care teams filter through data quickly to detect and decrease false activations. AI-enabled precision helps avoid unnecessary, costly interventions and decreases the chance a condition will be overlooked before it worsens, which could result in even pricier treatments.
Earlier this year, researchers at Harvard University and McKinsey projected that AI could save 5 to 10 percent of healthcare spend, roughly $200 billion to $360 billion a year, based on AI use cases employing current technologies attainable within the next five years, without sacrificing quality or access.
Now, when a patient presents with a heart murmur, AI analysis of the ECG in addition to my own clinical findings bring enhanced prognostic capability and new a compass to the direction of care. The AI solution helps me more rapidly detect when a patient needs care for conditions like ST elevation myocardial infarction (STEMI) or atrial fibrillation. This enables me to deploy appropriate therapy rapidly, which could mean the difference between life and death.
As cardiology workups become more expensive and there are delays in getting data back, AI becomes a powerful tool to prioritize tests and eliminate those that are unnecessary. The enhanced efficiency AI enables in individualized treatment protocols improves our ability to save patient lives, our first and foremost concern. As we move into value-based care and away from fee-for-service, early identification of patients who are at risk for a particular disease will mean earlier treatment. This provides not only better outcomes for the patient but will help the healthcare system avoid unneeded hospitalizations and costs of care that occur when a disease is diagnosed at a later stage.
There are those who urge caution in the use of AI. While I can understand reticence of AI adoption in other areas, in healthcare I overwhelmingly see tangible positives. I work as a cardiologist for a multi-specialty group that sees patients both in capitated payer and fee-for-service environments. This diverse practice gives me the experience to know where insurance providers will find value, or in other words, what they will likely pay for. I believe that as these protocols become more scientifically validated, AI is going to help the entire healthcare system to identify at-risk patients quickly and accurately. AI will help us build out those efficiencies, connecting us with insurance, providing more rapid authorizations for appropriate cases.
The use of AI healthcare will revolutionize the way we practice medicine. It will help us directly help patients. It will also increase efficiencies and decrease overall healthcare spend so that we can do more for our patients and do it better.
Photo: ismagilov, Getty Images