An artificial intelligence language model improves readability of burns first aid information

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Health Literacy

Background:

• Patient education materials (PEMs) on burns first aid often lack optimal readability.
• Improving the accessibility of health information is crucial for effective public health communication.

Purpose of the Study:

• To evaluate the effectiveness of an AI large language model in enhancing the readability of burns first aid information.
• To determine if AI can adapt existing content to meet recommended literacy levels.

Main Methods:

• An AI language model (ChatGPT-3) was employed to rewrite content from top English-language burns first aid webpages.
• Readability was assessed using five validated tools, targeting an 11-year-old literacy level.
• Original and AI-modified materials were compared using statistical analysis.

Main Results:

• Initially, only 4% of materials met the target readability level; the median grade was 6.9.
• After AI modification, 18% of materials reached the target readability level, with a median grade of 6.
• AI rewriting significantly improved all readability scores (p < .001).

Conclusions:

• Artificial intelligence large language models are a viable and effective tool for improving the readability of health information.
• AI-driven content modification can make critical burns first aid guidance more understandable and accessible to the public.