I. Yunusa

Comparative performance of large language models for appraising bias in real-world evidence studies.

LLMs hold considerable promise for automating methodological appraisal of RWE studies; however, their performance is variable and model dependent.