JMIR Medical Education最新文献

Background: Research capacity building (RCB) among health care professionals remains limited, particularly for those working outside academic institutions. Japan is experiencing a decline in original clinical research due to insufficient RCB infrastructure. Our previous hospital-based workshops were effective but faced geographical and sustainability constraints. We developed a fully online Scientific Research Works Peer Support Group (SRWS-PSG) model that addresses geographical and time-bound constraints and establishes a sustainable economic model. Mentees use online materials, receive support from mentors via a communication platform after formulating their research question, and transition into mentors upon publication.

Objective: We evaluated whether our model's theoretical benefits translated into actual program effectiveness in RCB among health care professionals.

Methods: We conducted a retrospective cohort study of health care professionals who participated in the SRWS-PSG program between September 2019 and January 2025. Mentees progressed through a structured modular curriculum covering systematic review methodology, from protocol development to manuscript preparation, with personalized mentoring support. We evaluated manuscript submission, program discontinuation, promotion to a mentor status, and mentor response time. We collected data from program records and chat logs. Manuscript submission was defined as mentor-confirmed submission of a systematic review manuscript to a peer-reviewed journal. Program discontinuation referred to formal withdrawal before manuscript submission. Mentor promotion was defined as acceptance of an invitation to serve as a junior mentor after manuscript submission. Mentor response time was the elapsed time from a mentee's question in the chat to the first reply by an assigned mentor.

Results: Of 85 mentees analyzed, 31 (36.5%) held academic degrees (PhD or MPH), and 68 (80%) were medical doctors. During a median follow-up of 10 months, 51 (60%) submitted manuscripts and 46 (90%) became mentors. Ten mentees (12%) discontinued the program. The median mentor response time was 0.8 hours, with 90% responding within 24 hours.

Conclusions: A majority of participants of SRWS-PSG submitted manuscripts. This fully online RCB program might address geographical barriers and provides an adaptable approach for RCB across diverse health care contexts.

Unlabelled: The integration of large language models into medical education has significantly increased, providing valuable assistance in single-turn, isolated educational tasks. However, their utility remains limited in complex, iterative instructional workflows characteristic of clinical education. Single-prompt AI chatbots lack the necessary contextual awareness and iterative capability required for nuanced educational tasks. This Viewpoint paper argues for a shift from conventional chatbot paradigms toward a modular, multistep artificial intelligence (AI) agent framework that aligns closely with the pedagogical needs of medical educators. We propose a modular framework composed of specialized AI agents, each responsible for distinct instructional subtasks. Furthermore, these agents operate within clearly defined boundaries and are equipped with tools and resources to accomplish their tasks and ensure pedagogical continuity and coherence. Specialized agents enhance accuracy by using models optimally tailored to specific cognitive tasks, increasing the quality of outputs compared to single-model workflows. Using a clinical scenario design as an illustrative example, we demonstrate how task specialization, iterative feedback, and tool integration in an agent-based pipeline can mirror expert-driven educational processes. The framework maintains a human-in-the-loop structure, with educators reviewing and refining each output before progression, ensuring pedagogical integrity, flexibility, and transparency. Our proposed shift toward modular AI agents offers significant promise for enhancing educational workflows by delegating routine tasks to specialized systems. We encourage educators to explore how these emerging AI ecosystems could transform medical education.

Background: Large language models such as ChatGPT (OpenAI) have shown promise in medical education assessments, but the comparative effects of prompt engineering across optimized variants and relative performance against medical students remain unclear.

Objective: This study aims to systematically evaluate the impact of prompt engineering on five ChatGPT variants (GPT-3.5, GPT-4.0, GPT-4o, GPT-4o1-mini, and GPT-4o1) and benchmark their performance against fourth-year medical students in midterm and final examinations.

Methods: A 100-item examination dataset covering multiple choice questions, short answer questions, clinical case analysis, and image-based questions was administered to each model under no-prompt and prompt-engineering conditions over 5 independent runs. Student cohort scores (N=143) were collected for comparison. Responses were scored using standardized rubrics, converted to percentages, and analyzed in SPSS Statistics (v29.0) with paired t tests and Cohen d (P<.05).

Results: Baseline midterm scores ranged from 59.2% (GPT-3.5) to 94.1% (GPT-4o1), and final scores ranged from 55% to 92.4%. Fourth-year students averaged 89.4% (midterm) and 80.2% (final). Prompt engineering significantly improved GPT-3.5 (10.6%, P<.001) and GPT-4.0 (3.2%, P=.002) but yielded negligible gains for optimized variants (P=.07-.94). Optimized models matched or exceeded student performance on both exams.

Conclusions: Prompt engineering enhances early-generation model performance, whereas advanced variants inherently achieve near-ceiling accuracy, surpassing medical students. As large language models mature, emphasis should shift from prompt design to model selection, multimodal integration, and critical use of artificial intelligence as a learning companion.

Background: Traditional Chinese medicine (TCM) education in China has evolved significantly, shaped by both national policy and social needs. Despite this, the academic community has yet to fully explore the long-term trends and core issues in TCM education policies. As the global interest in TCM continues to grow, understanding these trends becomes crucial for guiding future policy and educational reforms. This study used cutting-edge deep learning techniques to fill this gap, offering a novel, data-driven perspective on the evolution of TCM education policies.

Objective: This study aimed to systematically analyze the research topics and evolutionary trends in TCM education policies in China using a deep learning-based topic modeling approach, providing valuable insights to guide future policy development and educational practices.

Methods: TCM policy-related documents were collected from major sources, including the Ministry of Education, the National Administration of Traditional Chinese Medicine, PKU Lawinfo, and archives of TCM colleges. The text was preprocessed and analyzed using the BERTopic model, a state-of-the-art tool for topic modeling, to extract key themes and examine the policy development trajectory.

Results: The analysis revealed 27 core topics in TCM education policies, including medical education, curriculum reform, rural health care, internationalization, and the integration of TCM with modern education systems. These topics were clustered into 5 stages of policy evolution: marginalization, standardization, specialization, systematization, and restandardization. These stages reflect the ongoing balancing act between modernizing TCM education and preserving its traditional values, while adapting to national political, social, and economic strategies.

Conclusions: This study offers groundbreaking insights into the dynamic and multifaceted evolution of TCM education policies in China. By leveraging the BERTopic model, it provides a comprehensive framework for understanding the forces shaping TCM education and offers actionable recommendations for future policy making. The findings are essential for educators, policymakers, and researchers aiming to refine and innovate TCM education in an increasingly globalized world.

Background: Training of health care professionals and their participation in continuous medical education are crucial to ensure quality health care. Low-resource countries in Sub-Saharan Africa struggle with health care disparities between urban and rural areas concerning access to educational resources. While e-learning can facilitate a wide distribution of educational content, it depends on learners' engagement and infrastructure.

Objective: This study aims to assess knowledge, attitude, practice, and access to infrastructure related to e-learning among health care professionals in primary health care settings in Ethiopia.

Methods: In April 2023, we carried out a quantitative, questionnaire-based cross-sectional study guided by the knowledge, attitudes, and practice framework, including additional items on available infrastructure. The scores in each category are defined as "high" and "low" based on the median, followed by the application of logistic regression on selected sociodemographic factors. We included health care professionals working in general and primary hospitals, health centers, and health posts.

Results: Of 398 participants (response rate 94.5%), more than half (n=207, 52%) reported feeling confident about their understanding of e-learning and conducting online searches, both for general (n=247, 62.1%) and medical-related content (n=251, 63.1%). Higher levels of education were associated with better knowledge (adjusted odds ratio [AOR] 2.32, 95% CI 1.45-3.68). Regardless of financial and personal efforts, we observed a generally positive attitude. Almost half of the participants (n=172, 43.2%) reported using the internet daily, compared to 16.8% (n=67) of participants who never used the internet. Higher education (AOR 2.56, 95% CI 1.57-4.16) and income levels (AOR 1.31, 95% CI 1.06-1.62) were associated with higher practice scores of e-learning-related activities. Women, however, exhibited lower practice scores (AOR 0.44, 95% CI 0.27-0.71). Regular access to an internet-enabled device was reported by 43.5% (n=173) of the participants. Smartphones were the primarily used device (268/393, 67.3%). Common barriers to internet access were limited internet availability (142/437, 32.5%) and costs (n=190, 43.5%). Higher education (AOR 1.56, 95% CI 0.98, 2.46) and income (AOR 1.50; 95% CI 1.21-1.85) were associated with increased access to infrastructure, while it was decreased for women (AOR 0.48, 95% CI 0.30-0.77).

Conclusions: Although Ethiopian health care professionals report mixed levels of knowledge, they have a positive attitude toward e-learning in medical education. While internet use is common, especially via smartphone, the access to devices and reliable internet is limited. To improve accessibility, investments in the digital infrastructure and individual digital education programs are necessary, especially targetin

Unlabelled: This paper aims to describe the cocreation and development processes of an educational ecosystem-centered Bachelor's degree in Digital Health and Biomedical Innovation (SauD InoB). This program is shaped by a multidisciplinary, intersectoral, and collaborative framework, involving more than 60 organizations in teaching activities, internship supervision, or hosting, most of which collaborated in needs assessment, curriculum development, and public promotion of the degree. In the context of health care digital transformation, this comprehensive Bachelor's degree will respond to unmet demands of the labor market by training students with technological, research, and management skills, as well as with basic clinical and biomedical concepts. Graduates will become transdisciplinary, creative professionals capable of understanding and integrating different "languages," reasoning, clinical processes, and scenarios.

Background: The integration of digital technology in dental education has been recognized for its potential to address the challenges in training removable partial denture (RPD) design. RPD framework design is crucial to long-term success in the treatment of dentition defects, but traditional training methods often fall short of adequately preparing students for real-world applications.

Objective: This study aimed to evaluate the efficacy of a 3D simulation-based preclinical training software for RPDs in enhancing learning outcomes among first-year stomatology master's students, while also assessing user perceptions among students and faculty.

Methods: RTS (Yikchi Siu) is a preclinical training software that simulates the clinical process of treating patients with partial edentulism. In this study, 26 newly enrolled master's degree students in stomatology who volunteered to participate were randomly divided into a control group (n=13) and a training group (n=13). The training group used the RTS for 2 credit hours (90 min) of self-study, while the control group received theoretical lessons and case practice from an instructor. After 2 hours, both groups completed the theoretical knowledge and drawing tests for RPD simultaneously. Test results were evaluated and graded by 2 experts in prosthodontics. Both users and teachers filled out a questionnaire afterward about their training experience.

Results: Participants in the training group obtained better final grades compared to controls (theoretical test: 88.8, SD 2.3; 85.7, SD 3.3, respectively; P=.01; drawing test: 89.8, SD 4.5; 85.1, SD 4.3, respectively; P=.01). The training group had a shorter completion time in the drawing test (12.6, SD 19 min; 17.7, SD 3 min, respectively; P<.001) but there were no significant differences in the completion times in the theoretical test (23.2, SD 2.2 min; 24.9, SD 2.8 min, respectively; P=.14). Students and faculty generally had a favorable opinion of the RTS.

Conclusions: The effectiveness of the RTS for newly enrolled master's degree students in stomatology to understand and apply their knowledge of RPD framework design was validated; the system was well received by both students and faculty members, who reported that it improved the effectiveness and convenience of teaching.

Background: Artificial intelligence (AI) has significantly impacted health care, medicine, and radiology, offering personalized treatment plans, simplified workflows, and informed clinical decisions. ChatGPT (OpenAI), a conversational AI model, has revolutionized health care and medical education by simulating clinical scenarios and improving communication skills. However, inconsistent performance across medical licensing examinations and variability between countries and specialties highlight the need for further research on contextual factors influencing AI accuracy and exploring its potential to enhance technical proficiency and soft skills, making AI a reliable tool in patient care and medical education.

Objective: This systematic review aims to evaluate and compare the accuracy and potential of ChatGPT-3.5 and 4.0 in medical licensing and in-training residency examinations across various countries and specialties.

Methods: A systematic review and meta-analysis were conducted, adhering to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Data were collected from multiple reputable databases (Scopus, PubMed, JMIR Publications, Elsevier, BMJ, and Wiley Online Library), focusing on studies published from January 2023 to July 2024. Analysis specifically targeted research assessing ChatGPT's efficacy in medical licensing exams, excluding studies not related to this focus or published in languages other than English. Ultimately, 53 studies were included, providing a robust dataset for comparing the accuracy rates of ChatGPT-3.5 and 4.0.

Results: ChatGPT-4 outperformed ChatGPT-3.5 in medical licensing exams, achieving a pooled accuracy of 81.8%, compared to ChatGPT-3.5's 60.8%. In in-training residency exams, ChatGPT-4 achieved an accuracy rate of 72.2%, compared to 57.7% for ChatGPT-3.5. The forest plot presented a risk ratio of 1.36 (95% CI 1.30-1.43), demonstrating that ChatGPT-4 was 36% more likely to provide correct answers than ChatGPT-3.5 across both medical licensing and residency exams. These results indicate that ChatGPT-4 significantly outperforms ChatGPT-3.5, but the performance advantage varies depending on the exam type. This highlights the importance of targeted improvements and further research to optimize ChatGPT-4's performance in specific educational and clinical settings.

Conclusions: ChatGPT-4.0 and 3.5 show promising results in enhancing medical education and supporting clinical decision-making, but they cannot replace the comprehensive skill set required for effective medical practice. Future research should focus on improving AI's capabilities in interpreting complex clinical data and enhancing its reliability as an educational resource.

Background: Artificial intelligence (AI) has revolutionized medical education by delivering tools that enhance and optimize learning. However, there is limited research on the medical students' perceptions regarding the effectiveness of AI as a learning tool, particularly in Sri Lanka.

Objective: The study aimed to explore students' perceived barriers and limitations to using AI for learning as well as their expectations in terms of future use of AI in medical education.

Methods: An exploratory qualitative study was conducted in September 2024, involving focus group discussions with medical students from two major universities in Sri Lanka. Reflexive thematic analysis was used to identify key themes and subthemes emerging from the discussions.

Results: Thirty-eight medical students participated in 5 focus group discussions. The majority of the participants were Sinhalese female students. The perceived benefits included saving time and effort and collecting and summarizing information. However, concerns and limitations centered around inaccuracies of information provided and the negative impacts on critical thinking, social interactions (peer and student teacher), and long-term retention of knowledge. Students were confused about contradictory messages received from educators regarding the use of AI for teaching and learning. However, participants showed an enthusiasm for learning more about the ethical use of AI to enhance learning and indicated that basic AI knowledge should be taught in their undergraduate program.

Conclusions: Participants recognized several benefits of AI-assisted learning but also expressed concerns and limitations requiring further studies for effective integration of AI into medical education. They expressed openness and enthusiasm for using AI while demonstrating confusion and reluctance due to the perspectives and stance of educators. We recommend educating both the educators and learners on the ethical use of AI, enabling a formal integration of AI tools into medical curricula.