# A Cross-Sectional Study to Compare AI-Generated Educational Content Using Google Gemini for Medical Professionals With UpToDate on Pediatric Asthma

## TL;DR

This study compares AI-generated medical content from Google Gemini with UpToDate for pediatric asthma, finding differences in readability and accessibility.

## Contribution

The study evaluates the readability of AI-generated medical content versus established resources for pediatric asthma management.

## Key findings

- Google Gemini is more accessible for individuals with lower literacy skills compared to UpToDate.
- UpToDate scored higher on the SMOG index, indicating it is harder for the general population to understand.
- Google Gemini used more complex vocabulary, which may challenge younger or lower socio-economic populations.

## Abstract

Introduction: Accurate and up-to-date educational resources are crucial for medical professionals to deliver effective patient care, particularly in conditions like pediatric asthma, which has a high disease burden in children. Timely interventions are essential to manage this condition appropriately and to ensure better outcomes. With the rapid advancement of artificial intelligence in healthcare, AI tools like Google Gemini are being explored as quick and accessible alternatives for generating medical content.

Methods: A cross-sectional observational study was conducted to focus on four core topics related to the management of pediatric asthma. Prompts for each of the core topics were entered in Google Gemini and UpToDate to generate responses. The WebFx Readability Tool was used to assess readability utilizing metrics such as Flesch Reading Ease (FRE), Flesch-Kincaid Grade Level (FKGL), SMOG Index, word count, sentence count, words per sentence, difficult word count, and percentage. The collected data were analyzed using the Mann-Whitney U test, and a p-value of < 0.05 was considered statistically significant.

Results: When comparing the readability characteristics between UpToDate and Google Gemini, statistically significant differences were found, indicating that Google Gemini is more accessible for individuals with lower literacy skills. UpToDate received higher scores on the Simple Measure of Gobbledygook (SMOG) index across all four core topics, denoting it as hard to understand for the normal population. Google Gemini scored a greater difficulty word percentage across all four topics.

Conclusion: Google Gemini was found to use more complex vocabulary while still maintaining overall accessibility, making it appropriate for patients with lower literacy levels. Although certain readability parameters demonstrated Google Gemini to be a more reader-friendly tool for assessing and understanding medical content, the high percentage of difficult words may make it more challenging for younger individuals and lower socio-economic populations to access.