Heart health interventions: Get with the times and use AI!

Published 2 October 2023

Melissa Pelly

Behaviour researcher | Focus on chronic conditions & digital health | PhD candidate

LinkedIn: Melissa Pelly

If you don’t already have cardiovascular disease, chances are you may soon develop this debilitating condition.

Almost half of Australians aged over 20 will have heart disease, heart failure, stroke or hypertension (Tsao et al., 2023). But there is hope. AI has the potential to spot heart irregularities before you realise there’s a problem and let you know when to call an ambulance.

Our new paper, soon to be published in the IOS Press, has found that research on mobile health apps targeting heart attacks don’t tend to use artificial intelligence. This substantial gap in the market is baffling, considering that, in addition to being a potential lifesaver, artificial intelligence can:

  1. Reduce anxiety and preoccupation about when people may experience another event.
  2. Provide specialised information by personalising the intervention to each user’s specific context and condition. e.g., “I can see you are walking at a steep incline. Considering your ‘x’ heart condition, you should take breaks every minute to not over-exert yourself”.
  3. Predict each user’s needs and offer timely guidance. e.g., “at the rate you are exercising, you will reach your max heart rate in 5 minutes. To reduce the risk of damage to your heart, slow down to a more comfortable pace”.

Perspectives from the heart: What do patients and health professionals think?

So why aren’t there more AI-based interventions for everyday people? The answer may be many people are still scared of AI.

My new paper builds on a previous study I published earlier this year reporting interviews with heart attack professionals and patients (Pelly et al., 2023). I found that while there is some intrigue on the use of artificial intelligence to support patients, many are sceptical about how AI can tailor information for their needs, if it’s accurate and if advice would be too superficial, to name a few.

Some patients were open to using AI, but not for serious health concerns such as heart health:

“You wouldn’t want anything semi-serious. You know, like I wouldn’t really want an automated response for like ‘why is my heart racing.’” (Patient 2).

Some would use AI for health concerns, but check everything with their doctor because of concerns about the accuracy of AI:

“I would chat but still go to the doctor because I’m not sure how reliable it is.” (Patient 10)

Apps that have been created to personalise to each user show more promise:

“If it knew my specific details then yes, yeah, I would trust it more.” (Patient 2)

Apps which could direct the user to activities or information based on provided information were viewed favourably:

“For example, you’re very anxious today, would you like to watch a module about anxiety?” (Health Professional 2).

For a deeper dive into patient and health professional perspectives, explore the complete article published in the International Journal of Medical Informatics (Pelly et al., 2023).

What can you do today?

If society hesitates to take advantage of the potential lifesaving power of AI, industry and government won’t fund the research we need to develop it.

This article is a call to action – next time you hear about AI in healthcare, instead of immediately knocking it back, give it a try instead. Who knows, it might just become your long-term ally!

Emerging leaders in digital health