269P Performance of publicly available large language models as decision-aids for lung cancer screening: An evaluation of quality, safety, and readability

Abstract

Uptake of lung cancer screening (LCS) is suboptimal, creating a need for scalable, patient-centric communications. While large language models (LLMs) are increasingly used for health information, their performance as uninstructed decision-aids for LCS is unknown. This study evaluates the quality, safety, and content of three publicly available LLMs against clinical and linguistic benchmarks.

Ähnliche Arbeiten