The article investigates the use of Large Language Models (LLMs) in implementing Just-in-Time Adaptive Interventions (JITAIs) in digital health. JITAIs are being explored for sustainable behavior change, adapting interventions to an individual’s context and needs. However, traditional models face scalability and reliability limitations. To counter these, the article tests the use of LLMs, particularly the GPT-4 model, in fostering heart-healthy physical activity in outpatient cardiac rehabilitation. The study finds that JITAIs generated by LLMs were superior to those by healthcare professionals and laypersons, indicating LLMs’ significant potential in implementing JITAIs for personalized health.
Publication date: 15 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2402.08658
