Artificial intelligence in undergraduate medical education: an updated scoping review

Abstract

Background: The irrevocable alteration of medical education due to widespread access to large language models (LLMs) in 2022, and the concomitant surge in AI-related literature, has prompted us to update the evolving impact of AI on undergraduate medical education (UGME).

Methods: The scoping review adhered to the framework of Arksey and O'Malley. A literature search was conducted in April 2024 on PubMed, Scopus, Web of Science Core Collection, ERIC, and Google Scholar using the terms "UGME", "medical students", "AI", "NLP", "ML", "ChatGPT", and "LLM", and included publications that appeared from January 2020 to April 2024. The inclusion criteria were UGME and AI-related topics. The exclusion criteria were postgraduate education, continuing medical education, and non-AI technologies.

Results: After screening 3,238 identified publications, 310 were ultimately included in the review. One hundred sixty-one publications (52%) related to AI use solely in UGME appeared in eight months between the time the last general medical education scoping review on AI took place and the current study. The use of AI is rapidly increasing in UGME, both in basic and clinical courses, with applications ranging from autonomous tutoring, self-assessment, and simulation-based learning to assessment generation and grading, clinical assessment, procedural skills evaluation, and predictive analytics, among others. No publications assessed AI's impact on critical thinking or clinical reasoning in medical students. While students strongly demand the acquisition of AI literacy during UGME, and some institutions have begun integrating AI into their curricula, there is neither a standardized approach for doing so nor a consensus on AI competencies or ethical frameworks in UGME.

Conclusions: This review highlights the dramatic increase in the use of AI in UGME, presenting both benefits and challenges. While AI can enhance learning experiences, the best evidence for its implementation is unclear and requires, as key priorities, the definition of AI competencies, pedagogical methods, and ethical guidelines. Further research is needed to assess the impact of AI on ethics, empathy, critical thinking, and clinical reasoning. Faculty development in AI is vital, as is the need for collaborative and international endeavors.

Keywords: AI competencies; Artificial intelligence (AI); Large language models (LLMs); Scoping review; Undergraduate medical education (UGME).

Fig. 1

Publications included in the corpus. The bars show the yearly count of examined publications of each type published from 1 January 2020 to 31 December 2023. The connected blue squares represent the annual sum across all publication types

Fig. 2

Author collaboration network graph. The blue dots represent publications, and the orange dots represent the UN country groups: the African Group (AG), Asia and Pacific Group (APG), Eastern European Group (EEG), Latin America and Caribbean Group (LAC), and Western European Group and Others Group (WEOG). The graph visualizes the connections between publications and their respective author(s) originating UN country group(s)