Artificial intelligence is not fooling patients

Artificial intelligence (AI) isn’t capturing the imagination of patients the same way as it’s charming healthcare’s digital visionaries, a new poll from Pew Research reveals.   

Six out of ten Americans state they would feel “uncomfortable” with AI being used as part of their healthcare, the survey found, with consumers expressing trepidation about the technology’s impact on data security, care quality, and the patient-provider relationship.   

“On a personal level, there’s significant discomfort among Americans with the idea of AI being used in their own healthcare,” the report says. “A majority of the public is unconvinced that the use of AI in health and medicine would improve health outcomes.” 

A cautious view of AI’s potential for revolutionizing healthcare 

Respondents are particularly worried that AI might erode the bonds between patient and provider, with 57% stating that the human element of healthcare might suffer from an infusion of artificial intelligence technology. 

That represents a contrast to the viewpoint of many providers and healthcare leaders, who see AI as a way to free up the time and attention of front-line clinicians so they can focus on the patient in front of them.   

More than a third (37%) are also worried about data security as sensitive datasets are used in novel ways to feed increasingly advanced algorithms. AI models typically require massive amounts of data for training, and both the growing financial opportunities around selling or renting data, as well as the act of moving data across disparate systems itself, could be cause for concern. 

Interestingly, the participants are basically split on whether or not AI could improve care quality and outcomes.  While 40% believe AI might reduce the occurrence of mistakes and errors, just 30% think artificial intelligence will improve care quality for people similar to themselves and 38% don’t believe AI will make much of a difference at all. 

Bright spots in the public opinion of AI for healthcare 

But the news isn’t all bad for AI aficionados.  Respondents to the survey, particularly those who consider themselves more familiar with AI and its applications, do see some utility in new technologies. 

Close to 40% feel that AI could help to reduce bias in healthcare and support providers in their mission to treat people of all backgrounds equally. Those numbers get even higher among those who self-identify as seeing a clear problem with bias in healthcare: among those individuals, more than 50% think that AI will be important for achieving greater health equity. 

However, for some respondents, this optimism is predicated on their preconceived notion that AI is not inherently biased and does not see race or ethnicity when making suggestions, which isn’t always the case. To rise to these expectations, stakeholders will need to carefully design and train their algorithms to avoid hidden biases that may unintentionally skew their results. 

Situation-specific comfort with certain AI applications 

Consumers are also more willing to engage with AI in certain circumstances than others.   

For example, 65% of respondents would want to use an AI skin cancer detection algorithm in their own care. But 67% would definitely or probably not want AI to be involved with managing their pain medication after surgery. 

Americans are split on the idea of AI-powered surgical robots, with 40% in favor and 59% unsure about the idea.   

But they are much clearer about the role of AI in mental healthcare: 79% would probably or definitely not use an AI chatbot to help manage their mental healthcare, despite the rapid proliferation of these tools in the consumer market, largely aimed at younger demographics.   

Forty-six percent think these tools should only be used by people who are also seeing a human therapist, while an additional 28% think mental health chatbots shouldn’t be available to the public in any capacity. 

The strong backlash against these applications may be surprising, but it does hearken back to consumers’ concerns about preserving patient-provider relationships. It also shows that the public has started to take mental health issues very seriously, and would not trust this critical aspect of their health to a technology that is still in its infancy. 

Overall, it’s a mixed bag for artificial intelligence right now. Personal knowledge about AI seems to make a difference, with more education equating to more enthusiasm for AI-driven results.  This could bode well for the tide of opinion in the future as the technology becomes more deeply integrated and more familiar in everyday life. 

At the moment, however, healthcare leaders may wish to proceed with caution when extolling the virtues of new AI tools to their patients. Some patients may benefit from one-on-one conversations with their providers about the potential benefits and drawbacks of making AI a feature of their care, particularly as AI begins to have an everyday influence on the way clinicians make important decisions.   

Jennifer Bresnick is a journalist and freelance content creator with a decade of experience in the health IT industry.  Her work has focused on leveraging innovative technology tools to create value, improve health equity, and achieve the promises of the learning health system.  She can be reached at jennifer@inklesscreative.com.