面向基层医疗人员的AI工作坊：应用大语言模型减少老年人不当多重用药

This descriptive study was based on an AI workshop for local community primary care providers. Polypharmacy with inappropriate polypharmacy is a very common geriatric syndrome in older adults. It is still hard to resolve. As the application of Large Language Models (LLMs) to clinical practice is growing, LLMs could offer a potential solution to reduce inappropriate polypharmacy in older adults. Our 1.5-hour AI workshop demonstrated how to use LLMs to reduce inappropriate polypharmacy in older adults. We found that ChatGPT identified all potentially inappropriate medications (PIM) for one geriatric vignette which was consistent with the responses by most attendees. Additionally, 80-100% of attendees agreed or strongly agreed with ChatGPT outputs to three additional geriatric pharmacology vignettes and questions. Our AI workshop was rated as good, very good and excellent by 95% of attendees. Our preliminary study showed awareness of LLMs among the community primary care providers was low. This pilot AI workshop was a pioneer to integrate LLMs with continued medical education and to help primary care providers learn how to apply LLMs to reduce potentially inappropriate polypharmacy in older adults as an assistant instrument. More AI workshops are needed to help community primary care providers, particularly in rural areas.