Table 3.
Demographic variables assessed across the included papers.
| Author(s) | Year | Study design | Key findings |
|---|---|---|---|
| Zhang et al[15] | 2023 | Narrative Review | Generative AI, such as OpenAI’s ChatGPT, plays a crucial role in medicine and healthcare by generating various content types. Key issues discussed include trust, veracity, clinical safety, privacy, and opportunities for AI-driven conversational interfaces. The review concludes that generative AI’s role will grow as it becomes better adapted to the medical domain and regulatory frameworks evolve |
| Martin-Sanchez et al[16] | 2023 | Proposal and Framework Development | A new competency framework for personalized precision medicine is proposed, defining 58 competencies across 5 essential domains and a cross-cutting domain. This framework addresses the training needs of 6 professional profiles in healthcare, suggesting progressive levels of training for optimal performance in personalized medicine |
| Lopes et al[17] | 2023 | Bibliometric, Descriptive, and Retrospective Analysis | This study analyzed publications related to AI in Health Professions Education from 1990 to 2023. It found that AI’s adoption in healthcare education is increasing, particularly in enhancing clinical competencies and attitudes. AI also plays a significant role in patient education, contributing to health literacy and healthcare management. The study underscores the growing impact of AI in healthcare education |
| Lazarus et al[18] | 2022 | Critical Analysis | The study explores the integration of AI into anatomy education, highlighting 5 main tensions: human variations, healthcare practice, diversity and social justice, student support, and student learning. It suggests that while AI offers opportunities, it also comes with significant challenges that need careful management. Recommendations include enhancing transparency, embracing diversity among AI developers, and incorporating anatomical variations and uncertainties, all aimed at improving the use of AI in anatomy education |
| Talan et al[19] | 2021 | Bibliometric Analysis | This bibliometric study investigates the use of AI in education through the Web of Science database, identifying 2686 publications primarily from the USA. Key findings include the prominence of institutions like Carnegie Mellon University and researchers such as Vanlehn, K. and Chen, C.-M. in AI educational research. The analysis highlights the frequent use of terms like “intelligent tutoring systems” and “machine learning,” indicating significant trends in AI application in higher education contexts |
| Masters et al[20] | 2023 | Ethical Guide | The guide addresses the ethical implications of AI in Health Professions Education (HPE), focusing on issues such as data gathering, privacy, consent, bias, and transparency among others. It aims to prepare HPE teachers and administrators to understand and manage these ethical challenges effectively, ensuring that AI applications in HPE adhere to crucial ethical principles. The guide also suggests proactive measures to deal with potential ethical dilemmas in the use of AI in HPE |
| Clausmann et al[21] | 2023 | Systematic Overview | Clausmann et al explore the potentials and limitations of large language models (LLMs) like ChatGPT in clinical practice, medical research, and medical education. They discuss how LLMs can democratize medical knowledge, improve accessibility, and actively engage in medical education, yet caution about risks of misinformation and scientific misconduct due to issues like lack of accountability and transparency. This dual potential underscores the need for careful integration of LLMs in health professional education to harness their benefits while mitigating risks |
| Gilson et al[22] | 2023 | Comparative Performance Evaluation | This study evaluated ChatGPT’s ability to respond to medical licensing examination questions, comparing its performance with other models like GPT-3 and InstructGPT. ChatGPT showed varying accuracies across different datasets but generally outperformed InstructGPT. The findings highlight ChatGPT’s capability to provide logical justifications and incorporate relevant information, suggesting its potential use as an educational tool in preparing students for exams like the USMLE. This aligns with the review’s interest in using generative AI to personalize and enhance learning experiences in health professional education, particularly in high-stakes testing scenarios |
| Preiksaitis et al[23] | 2023 | Scoping Review | Preiksaitis et al conducted a scoping review to identify the opportunities and challenges of generative AI in medical education. They found themes indicating generative AI’s potential for self-directed learning, simulation scenarios, and academic support. However, concerns about academic integrity and data accuracy were also prominent. The study proposes critical areas for future research, including developing critical evaluation skills, rethinking assessment methodologies, and exploring human-AI interactions. These findings are essential for understanding how generative AI can be tailored to enhance personalized learning and address challenges in health professional education |
| Das et al[24] | 2023 | Empirical Analysis | Das et al assessed ChatGPT’s ability to answer microbiology questions based on a competency-based curriculum. ChatGPT demonstrated a high level of accuracy in answering both first- and second-order questions, although there was variability across different topics. The study suggests that while ChatGPT can be an effective tool for automated question-answering in microbiology, continuous improvements are necessary. This supports the potential of generative AI as a personalized learning assistant in health professional education, capable of providing accurate responses and adapting to various educational needs |