UPDF AI

Large Language Model−Based Chatbot vs Surgeon-Generated Informed Consent Documentation for Common Procedures

Hannah Decker,Karen Trang,6 作者,Elizabeth C. Wick

2023 · DOI: 10.1001/jamanetworkopen.2023.36997
JAMA Network Open · 引用 106 次

TLDR

Findings indicate that LLM-based chatbots are a promising tool for generating informed consent forms, easing the documentation burden on physicians while providing salient information to patients.

摘要

Key Points Question Can a large language model (LLM)-based chatbot outperform surgeons in generating readable, accurate, and complete procedure-specific risks, benefits, and alternatives (RBAs) for use in informed consent? Findings This cross-sectional study of 36 RBAs for 6 commonly performed surgical procedures found that the LLM-based chatbot generated more readable, complete, and accurate consent documentation than the surgeons. Meaning These findings indicate that LLM-based chatbots are a promising tool for generating informed consent forms, easing the documentation burden on physicians while providing salient information to patients.