Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
ChatGPT as a Virtual Patient: Written Empathic Expressions During Medical History Taking
17
Zitationen
4
Autoren
2025
Jahr
Abstract
Objective: Virtual patients are already utilized in the teaching of medical history taking. Since its emergence, ChatGPT has been integrated into several areas of medical education. This study aimed to examine whether ChatGPT can be used to train empathic history taking while fostering students' subjective autonomy. Methods: Third-year medical students took histories with ChatGPT 3.5 after entering a predefined prompt covering cardiological diseases. Afterwards, students answered a questionnaire regarding their experienced autonomy. All chats were analyzed using the Empathic Communication Coding System measuring ChatGPT's given empathic opportunities as well as students' responses. Results: Out of 659 interactions, 93 were identified as empathic. ChatGPT provided opportunities mostly through reporting emotional statements or challenges. Students sometimes missed reacting adequately to ChatGPT's opportunities but more often responded by implicit recognition of patient perspective and reported a high level of experienced autonomy. Conclusions: The study yielded preliminary results that ChatGPT might be suitable as a tool mimicking a virtual patient while enabling an empathic history taking. To date, ChatGPT seems valid as a supplement to training with simulated patients. Medical faculty could consider integrating ChatGPT into teaching, such as through a flipped classroom approach, to guide students in its use as ChatGPT continues to gain attention.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.764 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.674 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 8.234 Zit.
BioBERT: a pre-trained biomedical language representation model for biomedical text mining
2019 · 6.898 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.