ChatGPT-4 Can Help Hand Surgeons Communicate Better With Patients.

Area of Science:

• Medical writing
• Health literacy
• Artificial intelligence in medicine

Background:

• Patient-focused materials from hand surgery societies often exceed recommended readability levels.
• Health equity necessitates accessible health information tailored to appropriate literacy levels.
• Large language models (LLMs) offer potential solutions for improving content readability.

Purpose of the Study:

• To assess the effectiveness of ChatGPT-4 in rewriting patient-focused hand surgery articles to a sixth-grade readability level.
• To evaluate the impact of AI-driven rewriting on various readability metrics.
• To determine if AI can enhance health literacy in patient education materials.

Main Methods:

• Readability of English articles from the American Society for Surgery of the Hand and British Society for Surgery of the Hand websites was calculated using seven common formulas.
• Chat Generative Pre-Trained Transformer version 4 (ChatGPT-4) rewrote each article to a sixth-grade readability level.
• Readability of the rewritten articles was recalculated and compared to the original versions.

Main Results:

• ChatGPT-4 successfully improved readability across all tested formulas.
• A mean sixth-grade readability level was achieved for Flesch Kincaid Grade Level and Simple Measure of Gobbledygook.
• The Flesch Reading Ease score increased, indicating improved material readability.

Conclusions:

• ChatGPT-4 is a capable tool for enhancing the readability of patient-focused hand surgery content.
• AI can assist in making complex medical information more accessible, promoting health equity.
• Surgeon evaluation is crucial to ensure information accuracy is maintained alongside improved readability.