The chatbots eventually came up with the right diagnosis more than 90 percent of the time when provided with all pertinent information, the study found.<\/p>\n
Nevertheless, some people using such large language models such as ChatGPT and Grok may act on erroneous medical advice spit out in the earlier phase, according to the head of the research team. Those people may seek out biopsies when they don\u2019t have cancer or delay treatment for headaches that are actually signs of a stroke.<\/p>\n
\u201cYou can\u2019t just trust what the chatbot says,\u201d Dr. Marc Succi, executive director of the MESH Incubator, an innovation center at MGB, said in an interview. \u201cIt\u2019s a good start but there needs to be oversight and a human in the loop.\u201d<\/p>\n
By human, Succi meant a health care provider who can interview the patient, review the individual\u2019s medical history, run diagnostic tests, narrow down the potential diagnoses to the correct one, and start treatment right away if the problem is urgent.<\/p>\n
\u201cPart of being a doctor is forming an initial differential and then narrowing down the possibilities with pointed questions and things like a physical exam, the proper lab test \u2014 narrow, narrow, narrow until the final diagnosis,\u201d said Succi, an associate professor of radiology at Harvard Medical School.<\/p>\n
The research team compared 21 general-purpose large language models, including the latest versions of ChatGPT, DeepSeek, Claude, Gemini, and Grok. Researchers gradually fed each chatbot information on 29 published medical cases involving bread-and-butter conditions ranging from heart failure to ectopic pregnancies.<\/p>\n
The chatbots performed poorly when coming up with differential diagnoses based on age, gender, and symptoms. It was only after researchers provided the results of a physical examination and laboratory tests that the chatbots began to home in on the correct diagnosis.<\/p>\n
\u201cThese models are great at naming a final diagnosis once the data is complete, but they struggle at the open-ended start of a case, when there isn\u2019t much information,\u201d said Arya Rao, lead author, MESH researcher, and MD-PhD student at Harvard Medical School. <\/p>\n
Given that thousands of MGB patients lack primary care physicians amid a shortage of such practitioners<\/a>, the state\u2019s largest health care system last September launched an AI app called Care Connect that features a chatbot<\/a>. It take inquiries 24\/7, questions patients, reviews medical records, and sets up a telehealth appointment with a physician in as little as half an hour.<\/p>\n Dr. Rajesh Patel, vice president of digital patient experience at MGB, said the health system\u2019s chatbot is very different from those used by people seeking diagnoses.<\/p>\n \u201cThe AI chatbot in Care Connect is used to perform medical intake to help expedite appointment[s] with our clinicians — it does not diagnose,\u201d he said in a statement. \u201cPatients always see a real clinician who diagnoses, treats, and follows up with our patients …. This model actually supports the findings of this study, that physician involvement is still key in medical decision making.\u201d<\/p>\n