Evaluating the Effectiveness of Chatbot-Based Patient Education Compared to Traditional Patient Information Leaflets Related to Pediatric Anesthesia: A Pilot Cross-Sectional Study.
Uday Ambi, Mahima Gajanan Kamath, Sneha Jossie, Salman M Soomar
Background Clear, accurate, and empathetic communication is essential in pediatric anesthesia, where parental anxiety and information needs are high. Traditional patient information leaflets (PILs), while clinically robust, may lack emotional engagement. Large language model (LLM)-based chatbots, such as ChatGPT and Google Gemini, offer a novel, interactive approach to patient education, yet their role in pediatric anesthesia remains inadequately explored. Objective To evaluate and compare the readability, accuracy, completeness, sentiment, and parental satisfaction of artificial intelligence (AI)-generated patient education materials (ChatGPT and Google Gemini) with a clinician-authored departmental PIL (DPIL) for pediatric general anesthesia. Methods This pilot cross-sectional study evaluated responses generated by ChatGPT and Google Gemini to seven frequently asked questions derived from the departmental PIL. Three blinded leaflets were presented in randomized order using a computer-generated sequence and evaluated by 10 anesthetists for accuracy and completeness using 10-point Likert scales. Readability was assessed using Flesch Reading Ease and Flesch-Kincaid Grade Level scores. Sentiment analysis and parental satisfaction were also assessed. Both descriptive and inferential statistical analyses were performed. Results The DPIL demonstrated the highest readability, followed by ChatGPT, with Gemini scoring the lowest. All materials exceeded the recommended sixth-grade readability level. No significant differences were observed in accuracy or completeness among the three sources (p > 0.05). Parents consistently perceived ChatGPT responses as more reassuring and relatable, while the DPIL was viewed as informative but formal. Gemini responses were often considered linguistically complex. ChatGPT demonstrated a neutral and more empathetic sentiment compared with the other leaflets. Conclusion Clinician-authored PILS remain the most reliable source of pediatric anesthesia information. AI-generated content, particularly ChatGPT, may enhance clarity and emotional reassurance when used as a clinician-reviewed adjunct rather than a replacement.
Read on ELI