Artificial Intelligence-Assisted Patient Education: Unlocking ChatGPT’s Potential in Congenital Hydronephrosis Awareness (Preprint)

Abstract

<sec> <title>BACKGROUND</title> Artificial intelligence (AI) has emerged as a transformative tool for patient education, bridging gaps in understanding and accessibility. ChatGPT, a state-of-the-art Large Language Model (LLM), holds potential for enhancing awareness of congenital hydronephrosis, a condition affecting 1–5% of pregnancies and representing a leading cause of congenital kidney anomalies </sec> <sec> <title>OBJECTIVE</title> This study evaluates ChatGPT’s ability to deliver clear, actionable, and user-friendly information for congenital hydronephrosis. </sec> <sec> <title>METHODS</title> ChatGPT 4.0 answered 12 structured questions spanning diagnosis, treatment, and postoperative care of congenital hydronephrosis. Eight pediatric urologists independently assessed the responses using the CDC Clear Communication Index, a validated tool for evaluating the clarity and usability of health communication. Scores above 90% denote optimal communication quality. </sec> <sec> <title>RESULTS</title> The total score for this study was 61.92%, with the understandability scores for diagnostic, treatment, and postoperative care being 64.10%, 60.92%, and 60.75%, respectively. According to these results, the information sources ought to be updated, especially in relation to the elements that received the lowest scores. Identified limitations included overly technical language and insufficient tailoring for diverse patient needs. </sec> <sec> <title>CONCLUSIONS</title> While ChatGPT shows promise in advancing patient education on congenital hydronephrosis, achieving superior clarity and reliability requires further refinement. Incorporating iterative learning, domain-specific training, and expert oversight can enable AI tools to better bridge the gap between complex medical knowledge and patient comprehension, paving the way for equitable and impactful health communication. </sec>

Ähnliche Arbeiten