Assessing the Utilization of Large Language Model Chatbots for Educational Purposes by Medical Teachers: A Nationwide Survey From India.

Background Large language models (LLMs) are increasingly explored in healthcare and education. In medical education, they hold the potential to enhance learning by supporting personalized teaching, resource development, and student engagement. However, LLM use also raises concerns about ethics, accuracy, and reliance. Understanding how educators leverage LLMs can help assess their role and implications in medical education. Methods This cross-sectional online survey was conducted among medical teachers in India from December 2023 to March 2024. A validated questionnaire with acceptable internal consistency and test-retest reliability was used. It collected data on LLM chatbot usage patterns, as well as teachers' knowledge, attitudes, and practices regarding LLMs for educational purposes. Results A total of 396 medical teachers with an average teaching experience of 4.12±2.47 (minimum six months, maximum 13 years) years participated from different parts of India. The majority of the teachers heard about ChatGPT (OpenAI, San Francisco, CA, USA) (85%), followed by Copilot/Bing (Microsoft, Washington, DC, USA) (53%), and Gemini/Bard (Google, Mountain View, CA, USA) (45%) (p-value < 0.0001). However, 29% of the respondents never used it and 47% rarely use LLMs for educational purposes (p-value < 0.0001). The majority of the teachers use it for making any topic simple (55%), generating text for PowerPoint slides (55%), generating multiple-choice questions (MCQs) (52%), and finding answers to student's queries (35%). Knowledge (3.4±0.47) showed the highest score, followed by practice (3.3±0.81) and attitude (3.14±0.46) (p-value = 0.0023). Conclusion While awareness of LLMs was high among medical teachers in India, their actual usage for educational purposes remains limited. Despite recognizing the potential of LLMs for simplifying topics, generating teaching materials, and addressing student queries, a significant proportion of educators seldom integrate these technologies into their teaching practices. Institutions may provide training to help medical educators effectively integrate LLMs into teaching practices.