AI in patient education: a cross-sectional comparative analysis of patient information guides produced by ChatGPT versus Google Gemini for common haematology conditions

Background: Haematological disorders encompass a diverse range of conditions affecting blood and blood-forming organs. Educating Patients is crucial in managing these disorders, enabling individuals to understand their conditions, associated complications, and preventive measures. As artificial intelligence (AI) continues to progress, chatbots such as ChatGPT (CG) and Google Gemini (GG) offer readily available patient information. Methods: This cross-sectional comparative study was performed from January 15, 2025, to January 22, 2025. Six haematological conditions were selected. Patient education materials were created using CG (version 4.0) and GG, following standardized prompts. The generated texts were analysed for readability, similarity, and reliability with the use of Flesch-Kincaid Calculator (FKC), Quill Bot Plagiarism Tool (QPT) and Modified DISCERN score, respectively. Results: The study demonstrated no significant difference between the CG and GG in word count (WC), sentence count (SC), readability score, or reliability score in the statistical evaluation. The CG demonstrated a slightly higher ease-of-understanding score, but this difference was not statistically significant (p=0.175). Conclusions: AI tools including CG and GG have an important role in educating patients by delivering quick and structured information on haematological conditions. Future studies should explore the role of AI in other medical domains and investigate ways to enhance AI-generated patient education materials.