Back

Are older adults using ChatGPT for medical advice? Results from a cross-sectional survey study

Turchioe, M.; Shamnath, A.; Mayfield, J.; Xu, X.; Slotwiner, D.; Benda, N.

2026-01-04 health informatics
10.64898/2026.01.01.25342468 medRxiv
Show abstract

General-purpose large language models (LLMs) like ChatGPT are increasingly used for medical advice despite lacking medical training and frequently producing incorrect or unsafe output. Older adults health information-seeking behaviors using LLMs remain poorly characterized. We conducted a cross-sectional survey of 574 US adults aged 50+ recruited via Prolific, balanced by sex and race. Participants reported health information sources, ChatGPT and PubMed use, demographics, and health literacy. Most participants (92%) searched online for health information. All had heard of ChatGPT, and 63% used it for medical information, compared to 44% who had heard of PubMed and 39% who used it. Those with inadequate health literacy had higher odds of ChatGPT use for medical advice (AOR 2.36, 95% CI 1.30-4.52) versus those with adequate health literacy. In conclusion, with more than half of older adults using LLMs for medical advice, the development of safer, purpose-trained medical LLMs is warranted.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.