Assessing the Suitability of ChatGPT and DeepSeek AI for Patient Education on Common Rheumatological Disorders

Abstract

Introduction Systemic lupus erythematosus (SLE), systemic sclerosis, and dermatomyositis are among the most prevalent rheumatological conditions. Using artificial intelligence tools (AI) like ChatGPT and Deepseek AI in health care can help in providing personalized patient education, resulting in improved health care literacy and treatment adherence. Aim To assess and compare ChatGPT and DeepSeek AI's effectiveness in generating understandable, accurate, and reliable patient education guide for three rheumatological conditions: systemic lupus erythematosus, systemic sclerosis, and dermatomyositis. Methodology ChatGPT 4.0 and DeepSeek AI were asked to write a patient education guide for "systemic lupus erythematosus", "systemic sclerosis," and " dermatomyositis". These materials were assessed using validated readability scores (Flesch-Kincaid, grade level, ease score), linguistic complexity analysis (average syllables per word, words per sentence), and similarity metrics to standard rheumatological resources. Finally, the reliability was rated using " discern score, a structured evaluation framework formed based on evidence-based guidelines from the British Society of Rheumatology and the American College of Rheumatology. Results There was no significant difference in the word count (p=0.775), sentence count (p=0.802), average word per sentence (p=0.349), average syllables per sentence(p=0.101), grade level (p=0.193), similarity% (p=0.481), reliability score (p=0.742), and ease score (p=0.097) between ChatGPT and Deepseek AI. Conclusions Both ChatGPT and DeepSeek AI offer promising avenues for augmenting patient education in rheumatology, but due to the limitations, their output should be used as a complement - not a replacement - for verified, expert-reviewed educational materials.

Ähnliche Arbeiten