New search suggests that healthcare systems are struggling to effectively use AI to improve patient engagement.
The study, published earlier this month, found that while investments in tools such as ambient scribes are booming, AI applications for patient engagement are lagging. For the study, patient engagement startup Lirio healthcare consultancy commissioned Sage Growth Partners to interview more than 75 healthcare leaders across the United States
Only 5% of these executives say they are satisfied with the tools they have to address common patient engagement challenges, such as medication adherence and missed appointments – issues that not only lead to poor health outcomes, but also billions of dollars spent each year on avoidable health costs.
To help hospitals fill these gaps, companies selling patient engagement tools need to move toward an “N-of-1 personalization” model, said Amy Bucher, chief behavioral officer at Lirio.
“In healthcare, standard approaches to personalization are not very personal,” she said.
Often, personalization begins and ends with certain form fields like first name or age range. These approaches essentially just segment people based on demographics instead of considering their individual motivations and behaviors, Bucher noted.
For example, a provider might send the same generic reminder email about mammogram services to all women ages 40 and older. But not all women in this broad age group need the same type of messaging — and Bucher explained that an N-of-1 approach goes a step further by generating personalized messages that take into account each patient’s unique needs, behaviors and barriers.
“If a woman hasn’t had a mammogram in a few years, N-of-1 personalization examines why that might be the case. Is it difficult to prioritize the appointment over work? Does she need child care? Hates the anxiety of getting cancer screened? Either way, personalization that doesn’t address this issue won’t be as effective,” Bucher noted.
Recent advances in AI have created the opportunity to scale N-of-1 personalization, she added.
She noted that humans are able to do this well in a 1:1 format: we can integrate complex information from what people tell us, as well as nonverbal cues and contextual cues, and adjust our approach in the moment. But humans aren’t scalable for a large patient population, nor is it affordable to use live assistance for every use case, Bucher said.
“Technology has long been capable of processing more complex and larger data sets than humans, but only recently, with the explosion of agentic AI and the use of techniques such as reinforcement learning, can it also produce meaningful N-of-1 results,” she said.
She also highlighted that better personalization can reveal new levels of efficiency and connection.
Take diabetes for example. The disease affects 1 in 10 Americans – but despite its prevalence, people with diabetes are often disengaged with their health care providers, Bucher said.
“Standard methods of trying to get people to make appointments and take medications clearly don’t work for everyone. Personalizing these outreaches can help generate interest and get people to think differently about the value proposition of taking action, and doing so with digital outreach can create operational efficiencies,” she noted.
The diabetes use case highlights why patient engagement may be one of the most promising – and most underutilized – applications of AI in healthcare. When personalization goes beyond demographics to address individual barriers to action, it can not only drive clinical improvements but also help health systems engage patients at scale, Bucher said.
Photo: Paul Bradbury, Getty Images