Leveraging LLMs to Change Healthcare Provider Behavior

Healthcare providers are asked to change their behavior in numerous ways – engage differently in shared decision making, counsel patients on risk, educate prior to surgery, or identify implicit biases they may have. In order to achieve behavior change, interventions and tools need to be tailored to a specialty or provider. Tailored communication has been used effectively to improve the quality of education material or customize a decision aid so that it is more relevant and accessible. However, the tailoring process requires answering numerous questions and iteratively responding to prompts that refine content. Large language models (LLMs) are well positioned to automate the development of tailored behavior change interventions for providers and optimize culturally sensitive content. In this session we will discuss how LLMs can be used for a range of provider behavior change interventions and explore a case study related to a recently submitted R01 on how they can be adapted to address behavior change related colorectal surgery for low health literacy patients in the Midwest and Deep South.

Zoom information will be sent to e-HAIL members.