Different Forms of AI, Technology Can Be Beneficial in Preventive Cardiology

Artificial intelligence (AI) has been increasing in use across different forms of medicine, and preventive cardiology is no different. In a session held on August 1 at the ASPC 2025 Congress on CVD Prevention, experts discussed how both AI and technology can be used to improve the practice of preventing cardiovascular conditions—including language models, wearable technology, and mobile technology—to bring top care to patients.

Language Models Show Promise in Preventive Cardiology

Generative AI (genAI) is a popular form of AI that can produce content based on patterns learned from existing data, to generate text, video, audio, images, code, and other forms of content as prompted.

Ashish Sarraju, MD, FACC, the director of research at the preventive cardiology center at Cleveland Clinic, discussed how prevalent genAI already is in the preventive cardiology space, highlighting how findings already suggest responses to patient messages and clinic notes can be generated by gen AI. Also that it can be used to interpret data and images to give medical recommendations to patients.

“Cleveland Clinic, interested in this topic, did a survey last year of a nationally representative sample, asking how many people are using generative AI, and what do you think? A stunning 72% of patients said they would and have used generative AI to ask medical questions, and 65% said that they would trust the recommendations provided by the chatbot for their own health,” said Sarraju.

Sarraju hopes that genAI can not only be used for these methods but also to block accelerants of cardiovascular disease and improve prevention of these conditions. Accelerants of cardiovascular disease can include lack of health care access, underdiagnosis, poor risk stratification, delays in treatment, poor treatment implementation, and clinical trial underrepresentation, leading to incorrect assumptions regarding diagnosis or treatment. Lack of education from patients and clinician burnout can also contribute to the acceleration of cardiovascular disease.

genAI can help to close these gaps by allowing patients to get preventive cardiology recommendations from the chatbot, as previous research has shown that genAI can provide appropriate responses to common preventive cardiology questions 84% of the time.¹ Screening for clinical trial enrollment may also improve when using genAI, as it can be used to scan electronic medical records to determine eligibility. He noted that the current state of readability in preventive cardiology overall is low, which genAI may be able to help with.

“The current state with readability leaves much to be improved,” he said. “There were great data led by Keon Pearson, MD, a resident at Stanford, suggesting that of 27 unique websites reviewed addressing the question of lipoprotein, only 1 website even crossed the lower bound of the recommended reading level, below the sixth-grade level.”

Addressing medication and treatment use may benefit from the use of genAI. Studies have previously shown that more than half of those who had been prescribed statins had discontinued their use, at least temporarily.² genAI can be used to identify the reasons for these discontinuations as well as evaluate the public perceptions around medications.

With all of these impressive uses of genAI, it is easy to forget about the weaknesses of the platform. Public-facing misinformation regarding prevention is way more accessible and more numerous than accurate information, which may be used to train genAI and lead to the promotion of inaccurate information.

“I think we are at a point, especially with generative AI, where it has become democratized, where we can participate in the conversation in a much more robust manner than we could 10 years ago. You do not need to be a software engineer to understand the implications nor the penetration of these technologies potentially on our lives or our patients' lives,” said Sarraju.

Sarraju questioned whether efforts to mitigate propagation through genAI should be undertaken, with conversations around what can be done to mitigate inappropriate uses of the technology.

AI and wearable technologies can help in the practice of preventive cardiology. | Image credit: AntonioDiaz - stock.adobe.com

Wearable and Mobile Technologies Can Provide Take-Home Care

Wearables and mobile technologies also have a promising future in preventive cardiology, according to Seth Martin, MD, MHS, FASPC, professor of medicine at Johns Hopkins University. Wearable technologies, he said, can help to empower patients and providers, improve outcomes, improve experience of care, and reduce health care costs when used effectively.

“I think when it comes to the applications of this in cardiovascular disease, in prevention, it's really anything you can think of. It's hard to think of something where these tools will not have an impact,” he said.

Technologies that can be utilized in this way include smartphone apps, smartwatches, text messaging, telehealth, virtual reality, and even AI to assess health and wellness, coronary disease, cardiometabolic risk, and mental health among other health applications. Measuring steps, heart rate, exercise, and heart rhythm can help doctors make informed decisions surrounding treatment methods for their patients. Using these technologies alongside community input or engagement can help to increase physical activity, promote adequate sleep, encourage healthy dietary intakes, and quit smoking, which can reduce disparities.

Although all of these new gadgets are exciting, throwing them at problems is not always the way to go, said Martin. Instead, doctors should understand what problem they are trying to solve and work backward to create solutions. This is the approach that was taken with the smartphone app, Corrie, to encourage cardiovascular health, which proved effective. After patients used the app, 30-day all-cause readmissions were only 6.5% compared with 16.8% in those who did not use the technology.³

“We basically gave a comprehensive smartphone tool to educate and empower patients to take an active role in their care,” said Martin. “In the post–myocardial infarction setting, they started using this app in the hospital, and as they transitioned home, it helped with meds and their lifestyle and so forth.”

Technology can also be used to deliver rehab for those who require it, including through asynchronous, synchronous, and combination methods. In the case of some patients, discharge papers may be discarded or not followed after the patient is discharged from the hospital. Apps that provide the care from home and on a more accessible level can be beneficial in making sure the therapies are being followed. Although recent studies on whether mHealth applications were better than usual care in adults have had underwhelming results, Martin emphasized that studies have at least shown equivalence, which is promising.

Physicians should welcome wearable and mobile technologies in their practices as soon as possible, Martin said, as well as inform their patients about which data are the most reliable with these technologies, primarily resting heart rates, step counts, and exercise minutes. He also emphasized training future leaders in technology to be used in the cardiovascular space.

“We can meet people where they are by embracing these technologies that are in our hands and on our wrists, that are increasingly integrated into our lives. I don't think it's about technology replacing us, but rather the master clinician of tomorrow learning to balance technology with human touch,” he concluded.

References

1. Sarraju A, Bruemmer D, Van Iterson E, Cho L, Rodriguez F, Laffin L. Appropriateness of cardiovascular disease prevention recommendations obtained from a popular online chat-based artificial intelligence model. JAMA. 2023;329(10):842-844. doi:10.1001/jama.2023.1044

2. Zhang H, Plutzky J, Skentzos S. Discontinuation of statins in routine care settings: a cohort study. Ann Intern Med. 2013;158(7):526-534. doi:10.7326/0003-4819-158-7-201304020-00004

3. Marvel FA, Spaulding EM, Lee MA, et al. Digital health intervention in acute myocardial infarction. Circ Cardiovasc Qual Outcomes. 2021; 14(7):e007741. doi:10.1161/CIRCOUTCOMES.121.007741