Online platform vs. doctors: a comparative exploration of congenital cataract patient education from virtual to...

This study aimed to assess the quality and readability of patient education on congenital cataracts provided by Google, ChatGPT, and clinical doctors. Given the rarity of congenital cataracts and the need for accessible, accurate information for parents, we sought to evaluate the platforms' effectiveness in delivering relevant health information.

We developed two question banks related to congenital cataracts from different sources. Responses from Google, ChatGPT, and two doctors were evaluated across five criteria: correctness, completeness, readability, helpfulness, and safety. An ophthalmologist panel used a five-point Likert scale to score these responses. The readability of responses was also assessed using passage and readability statistics, with additional readability enhancements applied to ChatGPT responses.

The ChatGPT responses demonstrated similar quality to those from experienced doctors, particularly excelling in readability, which was enhanced further with simplification techniques. Resident doctors provided the most readable doctor responses, while Google results scored the lowest across all five evaluative criteria. Post-enhancement, ChatGPT responses showed significant improvements in readability and maintained response quality.

ChatGPT is a promising tool for delivering accessible, accurate information on congenital cataracts, especially for populations with lower health literacy. This study underscores the value of AI in healthcare education for rare conditions and highlights the need for consulting multiple information sources for comprehensive health guidance. ChatGPT, with readability enhancements, stands out as a particularly effective resource for public health information on congenital cataracts.