AI chatbots are getting better with healthcare questions

Interest in using AI in healthcare has grown significantly over the last few months, especially as the technology continues to improve with more data and refinement. In fact, several recent studies have found that AI chatbots, including ChatGPT, perform just as well — or sometimes even better — than human doctors in certain medical areas.

In April, researchers at the University of California, San Diego (UCSD) published a study in JAMA Internal Medicine that found that ChatGPT may be better than doctors at answering patients' questions. Using patient questions and doctor responses taken from Reddit's AskDocs subreddit, a panel of three healthcare professionals scored ChatGPT higher than the human doctors on both quality and empathy (4 vs. 3.33 on quality and 4.67 vs. 2.33 on empathy).

ChatGPT was also able to generate clinical notes on par with those written by senior internal medicine residents, often leaving reviewers unable to distinguish between the AI- and human-generated notes, according to a July study published in JAMA Internal Medicine.

Also in July, computer scientists at Google published a study in Nature that found that responses generated by the company's medical-specific AI chatbot Med-PaLM "compare favorably with answers given by clinicians."

Similarly, a recent preprint published in medRxiv found that ChatGPT performed comparably to human doctors who reviewed and diagnosed eye-related health complaints. The AI also performed significantly better than WebMD, an online symptom checker.

ChatGPT "is definitely an improvement over just putting something into a Google search bar and seeing what you find," said Nieraj Jain, an assistant professor at the Emory Eye Center and one of the preprint's authors.

With these improvements, more healthcare organizations are beginning to partner with tech companies to incorporate AI chatbots into their work. For example, Mayo Clinic recently partnered with Google to integrate the Med-PaLM 2 chatbot into the system. Separately, WebMD in June announced a partnership with HIA Technologies to provide interactive "digital health assistants."

"We need physicians to start realizing that these new tools are here to stay and they're offering new capabilities both to physicians and patients," said James Benoit, an AI consultant. "They are accurate enough at this point to start meriting some consideration."

The future of AI chatbots in healthcare

Even with these improvements in AI chatbots' ability, researchers and other experts are still cautious about their use in healthcare, especially with patients. AI chatbots could potentially raise privacy, safety, bias, liability, transparency, and other issues for healthcare professionals.

According to critics, the fact that AI chatbots perform better than someone just searching up their symptoms online is an unconvincing argument for their broader use.

"That's a little bit of a disappointing bar to set, isn't it?" said Mason Marks, a physician and professor who specializes in health law at Florida State University. "I don't know how helpful it is to say, 'Well, let's just throw this conversational AI on as a band-aid to make up for these deeper systemic issues.'"

One of the biggest dangers with AI is that market incentives may push organizations to adjust their programs to push patients towards certain drugs or medical services. "Companies might want to push a particular product over another," Marks said. "The potential for exploitation of people and the commercialization of data is unprecedented."

Even OpenAI, ChatGPT's developer, has urged caution when using its chatbots for healthcare purposes. "OpenAI's models are not fine-tuned to provide medical information," a company spokesperson said. "You should never use our models to provide diagnostic or treatment services for serious medical conditions."

Currently, there are no regulations surrounding the safety and effectiveness of AI in healthcare, and it doesn't seem like this will change anytime soon, Leonard writes. In May, FDA Commissioner Robert Califf said the "the regulation of large language models is critical to our future," but offered few details about what that would entail aside from recommending regulators be "nimble" in how they approach the issue.

According to John Ayers, a computational epidemiologist and lead author of the UCSD study, regulators should focus on how AI chatbots can impact patient outcomes, much like they do with other medical interventions. "If regulators came out and said that if you want to provide patient services using a chatbot, you have to demonstrate that chatbots improve patient outcomes, then randomized controlled trials would be registered tomorrow for a host of outcomes," he said.

"One hundred million people have ChatGPT on their phone and are asking questions right now," Ayers added. "People are going to use chatbots with or without us." (Leonard, KFF Health News, 9/12)

AI has been met with suspicion in the healthcare industry, but is this justified? We unpack the four most commonly cited risks of AI, and why some of them may be overstated.