Nearly half of U.S. doctors say they are anxious about using AI-powered software: survey

Doctor computer
A new physician survey indicates artificial intelligence applications are still in their infancy and have not affected mainstream physician practice at scale. (Getty/andrei_r)

Artificial intelligence gets a lot of buzz as a leading-edge technology in healthcare. But it still has a long way to go when it comes to adoption, with only 20% of physicians saying AI has changed the way they practice medicine, according to a recent survey.

In fact, the majority of physicians are anxious or uncomfortable with AI, according to Medscape’s survey of 1,500 doctors across Europe, Latin America and the U.S. Physicians in the U.S. voiced the most skepticism (49%), while 35% of physicians in Europe said they are uncomfortable with AI and 30% of physicians in Latin America said the same.

Although the number of physicians who say AI has changed the way they practice medicine grew from last year’s survey, 15% to 20%, the results indicate that AI applications are still in their infancy and have not affected mainstream physician practice at scale.

Whitepaper

Elevate Health Plan Member Engagement Through Call Center Transformation

Learn how health plans can rapidly transform their call center operations and provide high-touch, concierge service to health plan members.

The survey also found that physicians are more comfortable with using AI-powered tools in their personal lives—50% of U.S. physicians use Google Home, Alexa or something similar—but only 7% use it for professional purposes. Adoption numbers are even lower in Europe and Latin America, where more than 75% of doctors don't use voice-controlled technology for any reason.   

RELATED: To avoid repeating EHR missteps, involve physicians in artificial intelligence projects from the start

Only 19% of physicians said they would be comfortable using voice technology during a patient consultation, and nearly 1 in 3 physicians believe their role could be threatened by AI-powered software. 

Still, there is some interest from physicians in AI technology, with 70% of responding doctors indicating they believe it could make their decisions more accurate, according to the survey results

Two-thirds of physicians said they are likely to use AI in the future if it is better than humans at some diagnostic tasks, and 68% believe AI-powered software will allow them to spend more time on other important tasks. But less than half (44%) think AI will be as good or better than human physicians at diagnostic tasks.

Physicians expressed interest in using AI to look up drug information (54%), check for drug interactions (53%) and look up treatment guidelines (52%).

Despite the availability of digital tools to communicate with patients such as online appointment booking, video chat and text or image messaging, doctors are rarely using these tools, according to the survey. Only a quarter of U.S. doctors use online appointment booking, with doctors in Latin America and Europe reporting lower usage. Around 2% of doctors in all countries surveyed use video chat to communicate with patients. Twenty-three percent of U.S. physicians communicate with patients via text or image messaging compared to 12% of European doctors and 7% of Latin American doctors.

RELATED: Health systems launch new HIPAA-compliant Amazon Alexa voice tools

One exception is that 63% of Latin American physicians use WhatsApp to communicate with patients.

The survey points to the need for AI software developers to figure out how these digital tools can be integrated in a meaningful way into practice to improve the physician's workflow and day-to-day tasks.

Physician express concerns about bias, data privacy

The survey results also indicated that physicians are concerned about how these AI tools will handle sensitive patient information as well as potential bias, and doctors seem particularly distrustful of AI-powered software offered by pharmaceutical companies.

Physicians lean more toward using a voice-controlled clinical decision support system offered by a government institution (48%), a hospital system (47%) or an independent medical information provider (46%) compared to one offered by a technology company (31%).

Fewer than 1 in 5 would use a voice-controlled clinical decision support system that was offered by pharma, with 63% citing concerns about bias and the handling of sensitive personal information.

RELATED: Health IT Roundup—Google, IBM team with industry groups on AI standards

Physicians overall cited more trust in venture-backed independent technology startups with data privacy than technology startups owned or backed with investment from pharmaceutical companies. Doctors in Latin American demonstrated the most trust in independent tech startups (61%) compared to pharma-backed startups (35%). U.S. physicians seem more skeptical of either type of startup being able to handle sensitive personal data—31% of U.S. doctors trust independent venture-backed technology startups, and only 19% trust pharma-backed technology startups.

The report notes that technology vendors and life sciences companies need to lay strong user experience and product foundations first for new AI technologies to truly provide incremental value to physicians. Further, technology companies need to develop relevant applications that fill a needs gap, improve practice workflow, physician-patient interaction or provide wider health outcomes, the report said.

Suggested Articles

Nearly 10,000 patients involved in research studies were impacted by a third-party privacy breach that may have exposed their medical diagnoses, test results…

Payers have made strides digitizing and automating many core processes, yet prior authorization remains a largely manual, cumbersome process.

Veterans Health Administration medical facilities currently have a paper medical record backlog that if stacked up would be 5.15 miles high, according to the…