Wolters Kluwer survey finds patients, clinicians embrace AI but seek reassurance they can trust it

Artificial intelligence is changing the way patients prepare for clinical appointments, and in some cases enabling them to avoid visits altogether. Clinicians are on board for the most part, according to a recent Wolters Kluwer report called Future Ready Healthcare – in large part because they, too, are incorporating AI into the practice of medicine. 

The survey found that 42% of patients bring AI-generated information to their appointments, including 56% of patients 30 to 44 and 81% of patients 18 to 24. Of these patients, 59% said clinicians welcomed the information and engaged with it.  

All told, 69% of patients thought AI was helping improve their care, said Dr. Peter Bonis, Chief Medical Officer at Wolters Kluwer Health. “There’s little doubt that AI is already having an important impact on dynamic of clinical encounters. It’s accelerating health literacy and patient empowerment beyond web searches,” he said. 

Though AI use is widespread among patients and clinicians, with 72% using it in a professional capacity at least once a week, the industry needs to ensure individuals know how to use AI well, Bonis added. Health systems also face the challenge of recognizing this reality amid care models and business strategies that don’t account for what often motivates patients to use AI. 

An ecosystem of shared decision-making

Wolters Kluwer found that patients’ main use cases for AI tools include researching side effects, learning about a diagnosis, checking symptoms, or getting plain-language explanations of complex medical terminology. To that end, about 70% of patients and clinicians believe AI can help increase health literacy.  

While the idea of the more informed patient is hardly new – WebMD, for example, is nearly 30 years old – what’s changing is how quickly and easily information is available, Bonis said. “AI is interactive, immediate, personalized, convincing, and compelling. That’s why patients turn to it before, after, and in some cases instead of appointments.” 

This sets the stage for the patient-clinician dynamic to evolve, Bonis continued. Broad adoption of AI tools can create a healthcare ecosystem where clinicians and patients benefit from a “shared evidence base” that allows for collaborative decision-making. Equitable access to AI also has the potential to reduce variations in care quality based on social, geographic, or demographic factors, he added. 

Lean into shared AI use

Previous polls have shown patients use AI in healthcare because it’s convenient. As many as 14% of users have avoided an in-person encounter based on the information or advice they received. Meanwhile, a recent Commonwealth Fund report found that patients turn to AI “when there are gaps in communication or trust,” or when there’s a desire to ask questions without paying for another appointment. According to Wolters Kluwer, 19% of patients say AI provides answers faster than clinicians can. 

For patients, AI is filling a gap where convenient and affordable access is unavailable, Bonis said. For health systems, it’s worth leaning into patients’ expanding use of AI tools – “they’re using them for a reason” – especially given the potential cost and complexity of implementing other low-cost alternatives to in-person care. 

Offering AI tools backed by clinical expertise can help health systems overcome some of the friction that can arise when patients bring independent research into an appointment. In fact, Wolters Kluwer found that a majority of clinicians and patients want AI tools to explain how their findings do (or don’t) align with evidence-based resources.  

There will always be discrepancies that need to be resolved, Bonis said, but a mismatch between how patients and clinicians view AI use “might create more points of contention or uncertainty.” 

Governance and transparency are a team effort

The other challenge, of course, is that AI works well on textbook cases and in test environments but doesn’t perform as well in nuanced clinical situations. Many things can happen between an AI pilot and a large-scale rollout, Bonis noted; the most successful AI tools are built for a specific purpose and grounded in “well-synthesized, clinically validated evidence.” 

Governance and transparency are critical here. Wolters Kluwer found that 90% of patients and clinicians want humans in the loop to validate AI outputs, and 75% are concerned about accountability for an AI tool’s decisions. At the same time, only 27% of clinicians are aware of AI policies within their organizations. A recent Heidi survey told a similar story, with 83% of clinicians using AI “without guidance from their employer, a formal policy or a recommended tool.” 

Health systems shouldn’t have to tackle these issues alone, Bonis said. Organizations should lean on technology vendors already familiar with their governance structures and data requirements (among other factors) to develop guardrails for AI use and engage with the workforce on developing and enforcing policies.  

“AI is rapidly evolving, but understanding of its effectiveness and safety is lagging,” Bonis said. “The issue is truly understanding the real-world implications of AI and sensitizing stakeholders to the fact that, when they use it, they need to know its limitations.” 

Brian Eastwood is a Boston-based writer with more than 10 years of experience covering healthcare IT and healthcare delivery. He also writes about enterprise IT, consumer technology, and corporate leadership.