Patient Trust in Healthcare AI Relies on Use Case, But Familiarity Is Lacking

Do patients trust AI in healthcare? It depends on what it’s being used for, 44 percent of patients said in a recent athenahealth/Dynata poll that was emailed to PatientEngagementHIT.

But even as patient trust in AI depends on the use case, nearly half said they don’t actually know any of those use cases.

A sizeable 43 percent of poll respondents said they don’t know how AI is being used in healthcare, with Baby Boomers being the generation least likely to say they are knowledgeable about AI’s use in healthcare and Millennials saying they have the most knowledge about the technology.

AI’s surge in the healthcare scene has raised numerous questions, ranging from the ethics of healthcare AI to implicit bias perpetuated by AI algorithms to consumer acceptance of AI in medical care. While artificial intelligence can help automate low-level patient-provider communication and even supplement clinical decision support, experts are concerned about whether patients trust AI and are okay with it being used in their care.

This recent survey showed that patient trust in healthcare AI depends on the use case, with 44 percent saying as much, while 33 percent of patients reported overall trust in AI. Broken down by generation, only 12 percent of Baby Boomers said they trust the use of AI in healthcare, while 32 percent of Millennials said the same.

The biggest hang-up for consumers is the potential to lose the human touch in healthcare.

Nearly one in five (17 percent) of respondents said they are worried that integrating more AI into healthcare could threaten the humanized nature of their healthcare, and 43 percent said they think that AI will result in their provider spending less time with them. Millennials (52 percent) and Gen Z (50 percent) are more worried about this than Baby Boomers (37 percent) and Gen X (34 percent).

Most patients want there to be safeguards for AI use in healthcare, with 57 percent saying there should be laws and regulations from the government advising how providers can use AI. Another 50 percent said there should be similar regulations for how health technology companies can use AI.

Far fewer patients think AI development should be allowed to flourish freely, with only 11 percent saying there should be no rules around AI in healthcare, the survey showed.

Looking forward, patients are hopeful about the use of AI in healthcare. Half of respondents (52 percent) said they think AI will be part of the solution to healthcare’s biggest problems, not a problem that further exacerbates issues. Millennials and high-income folks were more likely to consider AI a potential solution to healthcare’s biggest problems.

In particular, 42 percent of patients think AI’s greatest potential is in improving clinical care and outcomes; 13 percent disagreed. Another quarter were more equivocal, saying AI could both help and harm healthcare, while the remaining 21 percent were unsure.

Still, patients aren’t exactly sure how AI is being used in healthcare, with 43 percent reporting as much. Another 66 percent are unsure if their provider is currently using AI as part of their clinical practice.

But if patients could choose, they’d like to see AI being used to remove some of the work on their providers’ plates (44 percent), pointing to an automation use case. Patients said this could help their providers spend more time with them. Another 40 percent said they’d like to see AI being used to provide some diagnostic support to providers, like analyzing X-rays or interpreting some test results.

Far fewer (17 percent) patients said they’d like to see AI being used to augment or replace patient-provider interactions. This was unsurprising, the survey authors said, considering patient fears that AI could damage the personalized human touch they value in their healthcare.

This isn’t the first survey to demonstrate a lukewarm consumer response to healthcare AI.

In May 2023, a report from the University of Arizona Health Sciences showed that around half of patients don’t fully trust AI-powered medical advice, like the information issued from chatbots like ChatGPT. Instead, patients still trust and prefer their traditional healthcare providers.

Another consumer survey from October 2023 found similar results, showing that patients were split on whether they trusted generative AI in healthcare. The survey showed that patients had some data privacy concerns, while others said they were already happy with the clinical information their clinicians provided to them.

Experts are still considering the way AI in healthcare, including large language models integrated into AI chatbots, could help remove some administrative burden for providers. Chatbots could help respond to low-level patient queries, leaving providers with more time to dive deeply with patients with complex needs.

But to see that reality, patients need to be open to using AI. The University of Arizona poll indicated that patients could warm to AI chatbots if their providers recommended them. Meanwhile, a separate report from Deloitte indicated that patient trust in AI could increase if the health tech sector more deeply considered the wants and needs of patients. At present, that’s not happening, the Deloitte report showed.