The landscape of the modern medical office has changed significantly over the past decade, particularly with the emergence of artificial intelligence. The days of turning to Google for medical questions are becoming increasingly distant as large language models like ChatGPT or the Bing chatbot help create personalized and comprehensive answers to routine medical questions. As a family doctor, I wonder how this tool could be useful not only to the office, but also to my patients after their appointment. Previously, when my patients arrived for their visit with Google searches, much of the visit consisted of extracting pearls of reliable information from the mush of misinformation online. Patients went home with documents full of relevant but overly complex information, usually never to look at them again. Recent studies suggest that artificial intelligence could play a key role in bridging the gap between what patients take home and what their doctors think they know.
Artificial intelligence is revolutionary in the sense that, over time, it can create its own connections and form new narratives after building a knowledge base from the web. But how accurate are these stories? Given all the misinformation online, if medical questions were asked of a language model like ChatGPT, could it provide reliable information to our patients? The guidelines recommend that women receive a screening mammogram up to age 74, but recommendations for women over 74 are nuanced and vary between organizations. In recent study, a team of six clinicians – experts in general internal medicine, family medicine, geriatric medicine, population health, cancer control and radiology – evaluated the extent to which ChatGPT's responses were appropriate to questions regarding mammograms older than 74 years old. The study found that 64 percent of At that time, ChatGPT offered an appropriate response. It also showed that in 18 percent of cases the responses were inappropriate, with the rest of the responses being unreliable or without real consensus.
When it comes to simpler medical questions, large language models seem to perform much better. A study by a team of radiologists found that the Bing chatbot could easily answer questions about imaging studies with 93 percent of responses being entirely correct while 7 percent were mostly correct. The reliability of answers can depend on several different factors: 1) avoid stacking questions when asking them AI2) some AI platforms have been shown to create informationwhile others cite the sources for which they derived the information, 3) areas of medicine without clear answers may not produce reliable or appropriate answers. answers of AI.
Patient education is another area where AI has proven to flourish. Large language models can make patient educational materials simpler and more easily understood and sometimes have the ability to translate this information in different languages. Medical jargon can often overcomplicate patient instructions and lead to communication problems between patients and their clinicians. With the ability of large language models to interpret and clarify patient educational materials, communication barriers can be reduced without prohibitive time and expense. As a primary care physician, I'm always looking for different places to connect my patients to reliable medical information when they want to learn more, before or after our appointments. With some improvements in reliability, I believe artificial intelligence could be the glue that connects my patients to more meaningful office visits to address their health concerns together.
Olivia Hilal is a family medicine resident.