The current study reviews the challenges of using ChatGPT within medical education to address challenges and proposed solutions given by experts in the medical field. The scoping review uses PubMED using Levac, Colquhoun, and O'Brien's Framework as a reference. A total of 13 studies, out of 102 communication, viewpoints, comments, editorials, brief reports, and letters were included and reviewed. Two reviewers independently conducted the selection process, and the decision was consensually made. The study excluded the review, original, and irrelevant articles. The scoping review did identify concerns about the ethical and rational use of ChatGPT. Several experts have raised questions on how to use ChatGPT so that it does not violate the epistemological methods, which generally involve the critical examination of the sources, methods, and justifications for knowledge claims.
TopBackground
Several important potentials for improving medical education are presented by artificial intelligence (AI) and generative language models (GLMs), including the availability of realistic simulations, digital patients, tailored feedback, evaluation techniques, and the removal of language barriers (Karabacak et al., 2023). AI chatbots, like ChatGPT, examine previous AI applications in healthcare and preview some potential AI chatbot-assisted in-basket assistance, such as communicating test results with patients, providing patient education, and supporting clinical decision-making during history-taking, reviewing the characteristics of previous diagnostic tests, and common management scenarios (Matulis & McCoy, 2023).
OpenAI's ChatGPT large language model (LLM), made available to the public in November 2022, quickly passed the milestone of having more than 100 million users. Since the start of the ChatGPT revolution, AI tools have become accessible, and this study considers how they can alter physicians' opinions of this ground-breaking technology (Mesko, 2023).
Figure 1.
Benefits and challenges of ChatGPT Use in medical education
TopMethodology
Levac, Colquhoun, and O'Brien's Framework: The current study has adapted the Levac, Colquhoun, and O'Brien proposed framework (Levac et al., 2010). In order to better understand how ChatGPT (a language model) will be used in education in 2023, this research topic will examine the points of view shared in communication, viewpoints, comments, editorials, brief reports, and letters. It attempts to discover what different academic and educational community stakeholders believe are the challenges of incorporating ChatGPT in educational situations.
The study has systematically searched PubMed for Communication, Viewpoints, Comments, editorials, brief reports and letters published in 2023 that focus on integrating ChatGPT in educational settings. The key search terms consist of ChatGPT and education to identify communication, viewpoints, comments, editorials, brief reports, and letters to narrow the search results. The research screening identified studies based on their relevance to the research question, ensuring they address the perceived benefits of integrating ChatGPT in educational contexts. Include studies that present authors' opinions, viewpoints, or experiences regarding the challenges of using ChatGPT in educational settings. Exclude studies that do not directly discuss the challenges of ChatGPT integration or those that focus on other topics unrelated to the research question. The study on the rational use of ChatGPT focuses on the main theme, such as “What challenges are raised by researchers? Moreover, What will be the possible solution to solve the challenges?
The current scoping review will focus on the publication year (2023), the focus on ChatGPT integration in education, and discussions on perceived challenges. One hundred and two articles were screened, and only thirteen articles were included in the study. Irrelevant articles that do not have clear challenges or solutions were excluded. The selected articles will be evaluated based on the given theme in Table 1. The content extracted from Table 1 was used to develop the word cloud (https://www.wordclouds.com/).