Large Language Models (LLMs) are used to personalize health education by adapting complex medical information to the patient's literacy level and preferred language. For example, chatbots can educate newly diagnosed diabetes patients interactively, using health literacy techniques like the teach-back method to ensure understanding.
Studies evaluating LLMs like GPT-4o mini on large datasets of consumer health questions (e.g., from the AskDocs subreddit) show these models can provide high-quality, expert-aligned answers to real-world, layperson medical queries. This demonstrates LLMs' potential to serve as accessible sources of reliable health information for consumers outside clinical settings.
LLMs have also been applied to analyze large volumes of free-text data from social media, news, and emergency calls to support public health surveillance and crisis response. For instance, during the COVID-19 pandemic, LLM-powered tools analyzed public sentiment and trends to inform messaging and interventions.
AskEllyn is an AI-powered chatbot designed specifically for people navigating breast cancer, as well as their friends and family. Unlike diagnostic tools, AskEllyn offers empathetic, non-judgmental support, answers to common questions, and shares lived experiences from a breast cancer survivor. It is available in multiple languages, adapts its tone to the user (e.g., child, spouse), and is free to use.
Doro is an AI chatbot developed to serve as a “daily mental health coach.” It is deployed in university wellness centers and clinics, offering 30-minute chat sessions where users can discuss stress, self-esteem, or other concerns. Doro suggests evidence-based techniques such as journaling, mindfulness, and planning, making mental health support more accessible and affordable, especially for those exhibiting early signs of mental health conditions.