Mitigating physician communication overload and enhancing patient–provider communication with Large Language Models

Physician communication overload, driven by high volumes of electronic messages, patient inquiries, and the cognitive effort required to navigate complex information systems, is a significant challenge in modern healthcare. This overload contributes to burnout, reduces the time available for direct patient care, and increases the risk of safety issues. Large Language Models (LLMs) offer promising tools to help alleviate these burdens. This paper examines the sources of communication overload, including extensive Electronic Health Record time demands, information overload within clinical documentation, and saturation of inbox communication. It then reviews four LLM applications that directly address these challenges: message and transcript summarization, inbox prioritization, automated responses to routine inquiries, and the generation of personalized patient education materials. A human–machine communication framework is used to explain how these tools shape clinical communication processes. The paper also discusses how LLMs can enhance patient communication by improving clarity, consistency, and engagement. Implementation factors and ethical implications, including accuracy, workflow integration, transparency, oversight, and patient acceptance, are outlined to guide responsible adoption.