Is AI a Reliable Source for Health Information?

Share On Social!

Artificial intelligence, better known as AI, is part of everyday life for many people.

In fact, experts say people interact with AI almost constantly or several times a day, from looking up information to creating something to making decisions.

That goes for health decision-making, too.

“AI is rapidly becoming a key player in many people’s medical decisions. According to OpenAI, the maker of ChatGPT, more than 40 million people consult the platform every day for health information,” NPR reports.

Let’s dive deeper into what this means and how trustworthy AI is for health advice.

Making the Right Medical Choices with AI?  

One in six American adults consult AI chatbots for health information at least once a month, according to a recent study in Nature Medicine.

The study sought to test the reliability of AI for health information.

Researchers gave participants medical scenarios, such as identifying a health condition or whether to go to the emergency room, and asked them to consult AI chatbots or a source of their own choosing.

“Participants using the [AI chatbots called large language models, or LLMs] identified relevant conditions in fewer than 34.5% of cases and disposition in fewer than 44.2%, both no better than the control group [who used their own sources],” researchers wrote.

When using AI and asking questions, wording is very important, according to Andrew Bean, one of the study authors.

“Doctors are trained to ask you questions about symptoms you might not have realized you should have mentioned,” Bean said

AI Diagnosis and Advice?

Another study that tested ChatGPT Health performance in a structured test of triage recommendations found that, “in 52% of emergency cases, the bots ‘under-triaged,’ meaning treated the ailment as less serious than it was.”

In circumstances where AI is able to correctly identify the condition, it often does not present the next steps with the appropriate amount of urgency.

Additionally, while ChatGPT Health did respond correctly to a “textbook medical emergency,”  it both over- and under- estimated the amount of time a patient could wait until pursuing care once an “element of time” was a factor.

“A spokesperson from OpenAI said this study did not represent the way people actually use ChatGPT, and that the previous study used an older version of ChatGPT that the company argues has since been corrected for some of the concerns that surfaced,” according to NPR.

How Can AI Improve Health?

While caution is needed when consulting AI for health advice, some believe it can improve the quality of health information that people usually get.

Rober Watcher, a doctor at UC San Francisco and author of the recently published book, A Giant Leap: How AI Is Transforming Health Care and What That Means for Our Future, suggests consulting AI can gather information that otherwise is hard to find or afford.

“The advice you get from the tools is substantially better than nothing and better than what you would get from your second cousin,” Wachter told NPR.

Similarly, Adam Rodman, a hospitalist who researches AI programs at Harvard Medical School, suggests that AI may not be reliable for emergency situations but can be valuable when it comes to patient’s interaction with a human medical practitioner, allowing patients to walk into appointments better informed.

“A good time to use a large language model is when you’re about to go see a doctor — or after you see your doctor,” Rodman said.

Considering the upsides and downsides of AI for health information, some doctors are optimistic that AI can serve as an extension of the human relationship between doctors and patients, rather than a replacement.

“What I hope is that this technology can be used in a way that enhances humanity in medicine,” Rodman said. “And not in a way that cuts out the doctor-patient relationship.”

Using AI and its Reflection of the US Population

Fairness is a remaining issue with using AI for health and other information.

For example, both experts and the public believe White adults’ views are better-represented than other groups’ when it comes to AI design, according to Pew Research.

“73% of experts we surveyed say the people designing AI take the experiences and views of White adults into account at least somewhat well,” according to Pew. “Half say the same for Asian adults and even smaller shares about Black (27%) or Hispanic (25%) adults.”

Improve Health for All People in Your Community

Learning more about technological developments in healthcare is a reminder of the importance of staying healthy.

But what about the health of your community?

Learn more by downloading a Salud America! Health Report Card for your town!

Enter your county name and get auto-generated local data with interactive maps and comparative gauges on several health indicators. This can help you visualize and explore local issues in education, housing, transportation, food, health, and more.

See how your county stacks up compared to the rest of your state and nation.

Then email the Report Card to local leaders to raise awareness, include the data in a presentation or grant proposal, or share it on social media to drive healthy changes in your community!

GET YOUR HEALTH  REPORT CARD!