Novel Breast Reconstruction Generative Pretrained Transformer Promotes Decisional Confidence and Clinical Efficiency

Area of Science:

• Plastic Surgery
• Artificial Intelligence in Medicine
• Patient Decision Making

Background:

• Breast reconstruction involves complex patient decisions.
• Patient education and engagement are crucial for successful outcomes.
• Current consultation methods may not fully address patient concerns.

Purpose of the Study:

• To evaluate the impact of a custom generative pretrained transformer (GPT) tool on patient decision-making and satisfaction during breast reconstruction consultations.
• To assess the effect of GPT use on consultation duration and decisional conflict.
• To explore patient perceptions of an AI-driven tool in a surgical setting.

Main Methods:

• A randomized controlled trial involving 20 patients undergoing immediate breast reconstruction.
• Ten patients interacted with a custom GPT for a maximum of 10 minutes prior to surgeon consultation.
• Standardized surveys assessed decisional conflict, healthcare satisfaction, and GPT usability.

Main Results:

• GPT cohort patients reported significantly less decisional conflict (P=0.03) and shorter consultation times (P=0.02).
• Patients using the GPT showed increased satisfaction with shared decision-making and preoperative communication (P=0.04).
• All patients found the GPT helpful, with common questions focusing on recovery, autologous reconstruction, and complications.

Conclusions:

• Custom-built generative pretrained transformer (GPT) tools can enhance patient understanding and reduce decisional conflict in breast reconstruction.
• Integrating AI tools like GPTs into surgical practices can improve efficiency and patient satisfaction.
• This study highlights the potential of AI to innovate patient care in reconstructive surgery.