AI and Wearables Shaping Hypertension’s Future

In part two of this hypertension discussion, Payal Kohli, MD, explores the evolving landscape of hypertension monitoring and management, highlighting the challenges of accurately diagnosing conditions like white coat and masked hypertension. Moreover, Dr. Kohli expresses optimism about emerging wearable technologies, such as reflective photoplethysmography patches —wearables that could offer continuous, noninvasive blood pressure monitoring. Dr. Kohli also discusses the immense potential in artificial intelligence to enhance diagnosis, personalize medication choices based on patient-specific risk profiles, and streamline chronic disease management through better data analysis.

Transcript:

Cardio Care Today: What is your perspective on the accuracy and utility of wearable blood pressure monitors compared to traditional cuff-based measurements?

Dr. Payal Kohli: This is fascinating because one of the biggest challenges I have as a physician is when the patient comes into the office with elevated blood pressure, and then they say, “Oh, Dr. Kohli, I promise it’s normal at home, and here’s my blood pressure log.” I’m left scratching my head; is it just white coat hypertension, or is there underdetection of high blood pressure at home? Sometimes the opposite can occur, what we call masked hypertension, where they’re normal in the office, but they end up running high at home. So I think wearables are the future.

Having said that, the technology needs to be validated. For now, I’m mostly relying on my patients to use the arm cuff, which is the one that we know performs the best with the least variability. Sometimes patients come in with a wrist cuff or a watch, for example, that measures blood pressure using the traditional method.  However, the wrist is more prone to noise, especially if it is small or large, or if the hand is in motion. So I rely less on that. But the newest technology, which I think is interesting, is something called reflective photoplethysmography, which is like a patch, a sticker, just like a Zio patch or a Holter monitor that you would wear that would check your blood pressure, along with other vital signs such as your pulse oximetry, your heart rate. Even measures such as stroke volume and cardiac output can be detected using this technology, based on reflective photoplethysmography.

I think it’s a cool technology. It has tremendous potential. Imagine if I could put a patch on my patient for a whole day, or a whole week, or two weeks, and have it continuously assess blood pressure and get an accurate read. However, it still needs a bit of validation, in my opinion, before I start to titrate my medications based on that. I’m just excited that that’s where the future is headed, because if this technology is validated, you could start to think about less invasive ways to check your blood pressure, because even the process of checking our blood pressure, which squeezes our arms, in some patients, can raise their blood pressure because it causes a little bit of pain and sympathetic activation when you’re checking the blood pressure. So stay tuned. It’s coming, and it holds a lot of promise.

With emerging AI and machine learning tools, how do you envision decision-support systems reshaping hypertension care in the next 5-10 years?

The potential is tremendous for AI to enhance our clinical capabilities, improve patient screening for hypertension, and even empower patients, pharmacists, nurse practitioners, and the rest of the healthcare team to more effectively titrate blood pressure medications. I see a few potential applications. The first is to help in detection. Hundreds of thousands of patients visit primary care facilities every day. And understanding if they’ve had a repetitive trend with multiple readings of stage one hypertension with blood pressures over one 30 over 80, or if they have other comorbidities that might suggest hypercortisolism or some other diagnosis as well, AI could help us be better at diagnosing, detecting and understanding that this patient didn’t just have a high blood pressure today because the parking lot was full, but they’ve had a high blood pressure many times before, and they do have stage one hypertension.

So I think diagnosis with AI is going to be tremendous, but what I’m optimistic about is AI helping me to better understand perhaps which medications I should use for a patient based on their risk factor profile, their age, their comorbidities, their history of side effects, their medication intolerances because right now it’s a free-for-all, right? We have a choice between angiotensin receptor blockers (ARBs), angiotensin-converting enzyme inhibitors (ACE inhibitors), calcium channel blockers, or thiazide diuretics as our first-line options, and we simply select one. We randomly pick. So, if AI could help direct me based on the patient’s risk factor profile, considering they’ve experienced edema with other medications, their tendency to be more hyponatremic, and their slightly older age, that would be a tremendous application as well.

So I think AI in hypertension is something that I’m very excited about, especially with the interface of those wearables as we discussed, if the AI could become the monitor that looks at this chronic disease management because I know a lot of practices have RCM programs where they look at chronic disease management and titrate blood pressure medicines, but if AI could make us more efficient with that, then it has a role, not just in diagnosis, not just in selecting the treatment, but even in ongoing titration and management of the treatment. So I’m very excited about that potential.

From a research perspective, we don’t completely understand hypertension, and an emerging body of literature suggests that our gut microbiome and the types of bile acids we produce determine our blood pressure. And that could be one of the reasons why we’re seeing more hypertension now than ever before. Of course, the sodium in our food, but even the processed foods change the way that our gut microbiome looks. So by altering that back in a positive direction, can we think about treating hypertension using different mechanisms, not just RAS and the endothelin receptors and the other receptors on the blood vessels, but thinking about if the gut could be involved in regulating our hypertension beyond just a sodium intake, how could that look for our patients as well?