Improving the Readability of Spasmodic Dysphonia Patient Education Materials Using ChatGPT-4o Mini: A Cross-Sectional...

Area of Science:

• Health Communication
• Medical Informatics
• Artificial Intelligence in Healthcare

Background:

• Online patient education materials (PEMs) for spasmodic dysphonia (SD) often present complex information exceeding recommended readability levels.
• This accessibility barrier hinders patient comprehension and engagement with crucial health information.

Purpose of the Study:

• To assess the current readability of SD-related PEMs.
• To evaluate if the AI tool ChatGPT-4o mini can enhance PEM readability to a sixth-grade level.
• To determine the impact of AI revision on the preservation of essential clinical information in SD PEMs.

Main Methods:

• A systematic search identified 14 websites with 38 SD PEMs across core domains (definition, causes, treatment).
• ChatGPT-4o mini generated sixth-grade level revisions for each PEM.
• Readability was analyzed using Flesch-Kincaid, Flesch Reading Ease, Gunning Fog, and SMOG indices.
• Content fidelity analysis (specificity reduction, dropped/added content) assessed information preservation.

Main Results:

• Original SD PEMs required reading levels from 10th grade to early college.
• ChatGPT-4o mini significantly improved readability across all metrics (P <0.001).
• AI revision resulted in information loss: 57% dropped content, 29% reduced specificity; only 21% of PEMs were fully preserved.

Conclusions:

• AI tools like ChatGPT-4o mini can enhance the readability of health information for conditions like SD.
• AI-driven revisions may compromise essential clinical details, necessitating careful expert oversight.
• Future research should investigate patient comprehension and decision-making outcomes with AI-revised educational materials.