SurgiBot: Large Language Model-Based Patient Simulator for Surgical Informed Consent Training

Area of Science:

• Medical Education
• Artificial Intelligence in Healthcare
• Surgical Training

Background:

• Informed consent discussions are crucial for patient understanding and ethical surgical practice.
• Surgical residents have limited opportunities to practice and receive feedback on these vital conversations.
• Current training methods may not adequately prepare residents for complex patient communication.

Purpose of the Study:

• To develop and evaluate SurgiBot, a novel training platform for surgical residents.
• To enhance the practice and feedback mechanisms for surgical informed consent discussions.
• To provide a realistic and adaptable simulation environment for resident training.

Main Methods:

• Developed SurgiBot, a large language model-based training platform.
• Incorporated participatory design with surgeons and patients to define requirements.
• Created simulated patient interactions for practicing informed consent conversations.
• Designed the platform for adaptability across various surgical specialties.

Main Results:

• SurgiBot offers a realistic and low-risk training environment for surgical residents.
• The platform facilitates practice of informed consent discussions with simulated patients.
• Participatory design ensured stakeholder relevance and usability.
• The system is adaptable for use in different surgical fields.

Conclusions:

• SurgiBot represents a significant advancement in surgical resident training for informed consent.
• The AI-powered platform addresses the need for improved communication skills practice.
• This tool can enhance patient safety and shared decision-making in surgery.