Large Language Models in Surgery: Promise, Pitfalls, and Practical Use

Area of Science:

• Artificial Intelligence in Medicine
• Surgical Technology
• Clinical Informatics

Background:

• AI-related medical publications surged 36-fold (2000-2022), yet practical surgical guidance lags.
• Large Language Models (LLMs) are emerging AI tools with broad medical applications.
• This review focuses on LLMs' pragmatic use in surgery, addressing limitations and ethics.

Purpose of the Study:

• To delineate pragmatic applications of LLMs in surgical practice.
• To address key limitations, implementation considerations, and ethical concerns of LLMs for surgeons.
• To evaluate LLM benefits, constraints, and risk mitigation for surgical workflows and communication.

Main Methods:

• Reviewed contemporary LLM platforms and their integration into clinical workflows.
• Assessed LLM utility in patient communication, surgical research, and academic writing.
• Evaluated benefits, constraints, and risk mitigation strategies for surgeons.

Main Results:

• LLMs enhance clinical workflows via ambient documentation and data summarization.
• LLMs improve patient communication by simplifying information and tailoring messages.
• LLMs aid surgical research through literature review and study design optimization.
• Constraints include prompt engineering needs, hallucination risks, PHI protection, and ambiguous liability.

Conclusions:

• LLMs provide valuable tools for surgical efficiency and patient communication with proper oversight.
• Successful LLM integration requires prompt engineering, fact-checking, privacy protection, and human judgment.
• Surgeons must understand LLM limitations and ethical considerations for effective deployment.