Read more

July 12, 2023
4 min read
Save

Q&A: 32% of adults comfortable with an AI-led primary care visit

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Key takeaways:

  • In a recent survey, most U.S. adults said they expect AI to play a greater role in health care in 5 years.
  • Among health care professionals, 42% said AI in health care would equally help and hurt.

Nearly one-third of adults said they would be comfortable with an artificial intelligence-led primary care visit, a recent survey found.

Outbreaks Near Me — a team of epidemiologists from Boston Children’s Hospital and Harvard Medical School — polled 3,317 adults, including 357 health care professionals, on their opinions of artificial intelligence (AI)-led care and how big a role it will play in the near future.

PC0723Gertz_Graphic_01_WEB

The researchers found that 32% and 25% of respondents would be comfortable with AI-led primary care visits and AI-led therapy, respectively. The acceptance of AI as part of care differed by age ranges: 40% of adults aged 18 to 34 years said they would be comfortable with an AI-led primary care visit, compared with 24% of adults aged 65 years and older.

Even though a large portion of respondents were uncertain of AI-led appointments, 66% said they expect AI will play a bigger role in health care in 5 years than it does now.

About 14% of health care professionals said they use AI to diagnose, treat or care for patients, but most (83%) are not. Additionally, 23% said AI will help more than it will hurt, 33% said AI will hurt more than it will help and 42% said AI will equally help and hurt.

Healio spoke with Autumn Gertz, MS, project manager of the Boston Children’s Hospital’s computational epidemiology lab, to learn more about the findings, the future of AI in health care and more.

Healio: Were the survey responses surprising?

Gertz: The results of our survey were a mix of expected and surprising. It was surprising to see that overall fair proportions (albeit not the majority) are comfortable with AI-led primary care and therapy, especially because such AIs do not exist in the market yet. What was more expected is the result that younger individuals seem more comfortable with the idea of AI-led health care than older individuals. Furthermore, a trend of preference for human-led medicine at this time was also expected.

Healio: What responses stood out the most?

Gertz: A few things stood out! The first being that overall, already one in three people are comfortable with AI-led primary care, and one in four are comfortable with AI-led therapy. Although these are not the majority, they are fair amounts of the population comfortable with a hypothetical AI leading their care. Another interesting result was just under half (42%) of health care professionals think the use of AI in medicine will equally help and hurt. Again, this survey asked respondents to think about hypothetical AIs in medicine, so a high proportion of “mixed feelings” speaks to underlying concerns and skepticism around AI development.

Healio: What should primary care physicians take away from these findings?

Gertz: One takeaway for primary care physicians is that health care professionals may have differing acceptance and comfort of using AI than their patients. Two-thirds of our overall respondents think AI will play a bigger role in health care within the next 5 years, yet over 80% of health care professionals do not currently use AI in their practice. Additionally, a higher proportion of health care professionals indicated a belief that AI will hurt more than help. This suggests a disconnect between perceptions of AI use in medicine between patients and providers.

Healio: Are there any responses you think will see a significant shift in the coming years?

Gertz: Older adults having increased acceptance is certainly something we could see in the coming years. Similarly, as the population ages, younger individuals not surveyed here may grow up in a world where AI-led medicine is the norm and have very high rates of acceptance. I also think we will see shifting attitudes among health care professionals, in both directions, as they start to use AI within their practices. Positive, neutral and negative experiences with using AI in medical practice will shift attitudes around its use for both patients and health care professionals.

Healio: Where does research on patient/provider attitudes on AI go from here?

Gertz: This survey asked its participants to base their responses on hypothetical medically trained AI. So, an important step will be testing actual AIs as they emerge on the market. Mirroring the rigor of other suggested changes to medical practice will be important to ensuring the utility, safety and acceptance of AI in medicine.

As we wait for AIs to emerge to market, I think research on attitudes and willingness to use AI among groups that have traditionally received disparate care is important. If created and used correctly, AI has immense potential to help reduce bias, disparities and improve access to quality health care. However, as with human medical professionals, AI can have bias. Studying and understanding the bias certain AIs are susceptible to will be important to providing equitable care and reducing harm.

Healio: Anything else to add?

Gertz: In my opinion, the best health care AI would mitigate barriers to providing and receiving health care, not “replace” providers. Viewing the emergence of AI-led or supported health care as a tool rather than an obstacle may help physicians and clinicians optimize their practice. I also think AI-led and AI-supported health care should be tested in a manner similar to clinical trials to ensure it both supports health care professionals and patient outcomes. AI should not bypass current testing and vigor that is needed for other changes to medical practice. It will take time and effort of development, trial, study and success to build trust in using AI-led or supported medical care. However, these results, and the history of the relationship between the public and health care, indicate it takes much less to dissuade public opinion of using AI in health care. Therefore, it is important that patient safety and medical professional satisfaction are at the forefront of medical AI as we progress in this new frontier.

Reference: