Artificial Intelligence as a Consent Aid for Carpal Tunnel Release

Area of Science:

• Medical Informatics
• Surgical Education
• Artificial Intelligence in Healthcare

Background:

• The Montgomery ruling necessitates enhanced patient consent processes in hand surgery.
• Artificial Intelligence (AI) language models are being explored as tools for patient education and consent.
• Evaluating AI's role in providing clear and accessible patient information is crucial.

Purpose of the Study:

• To compare the quality and readability of AI-generated patient information with existing resources.
• To assess the suitability of AI language models for surgical consent, specifically for carpal tunnel release.

Main Methods:

• A comparative analysis was conducted between the Every Informed Decision Online (EIDO) leaflet and ChatGPT for carpal tunnel release.
• Information quality was evaluated using the DISCERN score.
• Readability was assessed using the Flesch-Kincaid scoring system.

Main Results:

• ChatGPT demonstrated significantly higher information quality (DISCERN score: 71/80) compared to EIDO (62/80).
• Interrater reliability for DISCERN scoring was high (kappa=0.65).
• ChatGPT's readability score (Flesch-Kincaid: 12.3) indicated a more complex reading age than EIDO (7.5).

Conclusions:

• AI language models, such as ChatGPT, can produce high-quality patient information for surgical consent.
• The enhanced quality of AI-generated content may come at the cost of reduced readability.
• Further research is needed to optimize AI tools for patient comprehension in surgical consent.