JMIR Medical Education最新文献

{"title":"Estimating the Prevalence of Generative AI Use in Medical School Application Essays: Cross-Sectional Study.","authors":"Nicholas C Spies, Valerie S Ratts, Ian S Hagemann","doi":"10.2196/96673","DOIUrl":"10.2196/96673","url":null,"abstract":"<p><strong>Background: </strong>Generative AI tools became widely available to the public in November 2022. The extent to which these tools have been used by medical school applicants during the admissions process is unknown.</p><p><strong>Objective: </strong>We aimed to estimate the extent of generative AI use among cohorts of applicants spanning the rollout of these tools.</p><p><strong>Methods: </strong>We retrospectively analyzed 6000 essays from 2364 applicants submitted to a US medical school in 2021 to 2022 (baseline, before the wide availability of AI) and 2023 to 2024 (test year) to estimate the prevalence of AI use and its relation to other application data. We used GPTZero, a commercially available detection tool, to generate a metric (P_human) reflecting the predicted probability that each essay was completely human generated, ranging from 0 (the essay appears to be entirely AI generated) to 1 (the essay appears to be entirely human generated).</p><p><strong>Results: </strong>Fully human-generated negative controls demonstrated a median P_human of 0.93 (range 0.89-0.97), while fully AI-generated positive controls demonstrated a median P_human of 0.01 (range 0.00-0.01). The \"Personal Comments\" essays submitted in the 2023 to 2024 application cycle had a median P_human of 0.77 (95% CI 0.76-0.78) compared with 0.83 (95% CI 0.82-0.85) during the 2021 to 2022 cycle. Approximately 12.3% and 2.7% of essays were evaluated as having P_human <0.5 in the test and baseline years, respectively. Essays submitted as part of the secondary application demonstrated lower P_human values than those of the American Medical College Application Service (AMCAS) \"Personal Comments\" essays. In applicant-clustered, multivariable generalized estimating equation analyses, supplementary essay type and younger age were significantly associated with lower P_human. Application completion date, self-reported gender, program type (MD vs MD-PhD), grade point average (GPA), Medical College Admission Test (MCAT) score, socioeconomic status, and undergraduate major were not significant predictors after false discovery rate correction. P_human was not predictive of interview invitation or acceptance in adjusted applicant-level logistic regression analyses.</p><p><strong>Conclusions: </strong>An AI detection algorithm identified signs of increased use of generative AI in 2023 to 2024 medical school admission applications compared to those in the 2021 to 2022 baseline period, before AI was widely available. AI use did not appear to confer an admissions advantage. Although these results provide information about the applicant pool as a whole, AI detection is imperfect. We do not recommend deploying AI detection for individual applications in live admissions cycles.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e96673"},"PeriodicalIF":13.9,"publicationDate":"2026-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13586704/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148888687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating the Influence of MCAT Scores on Medical Student Selection and Performance Using a Machine Learning Approach: Retrospective Cohort Study.","authors":"Michele Cherfane, Marc Ghanem, Georges Choueiry, Selim M Nasser, Tamina Elias-Rizk, Sola Aoun Bahous","doi":"10.2196/94197","DOIUrl":"10.2196/94197","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;The Medical College Admission Test (MCAT) has been central to medical school admissions in North America, though its necessity in holistic selection processes remains debated. The COVID-19 pandemic's suspension of MCAT testing sessions allowed institutions to explore alternative admission criteria. Additionally, global challenges in test administration underscore the vulnerability of systems dependent on a single standardized test.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aimed to (1) assess whether including MCAT scores in admissions decisions improves the prediction of medical school performance compared with grade point average (GPA) and interview-based selection and (2) evaluate whether machine learning (ML) can generate viable MCAT score predictions when testing is unavailable.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We conducted a retrospective cohort study of 1898 applicants to the Lebanese American University School of Medicine (2009-2023). Among 349 admitted students with MCAT scores, we compared admission composites including vs excluding the MCAT in relation to academic outcomes (Med 1-4 final grades) and clinical performance (Med 1-2 objective structured clinical examination [OSCE] scores). Students were stratified into tertiles to examine tier-specific effects. To model an MCAT alternative, we trained an ensemble ML model (least absolute shrinkage and selection operator, kernel ridge, gradient boosting, elastic net, and light gradient boosting machine [LightGBM]) using data from 1583 applicants (2009-2020) to predict MCAT scores from cumulative GPA, core GPA, interview scores, merit points, and honors participation. The model was validated on 315 applicants admitted during the MCAT suspension (2021-2023), and its impact on admission rankings and performance correlations was evaluated.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Including the MCAT in the admissions composite did not meaningfully improve prediction of overall academic performance (&lt;i&gt;r&lt;/i&gt;=0.63, 95% CI 0.56-0.69 with MCAT vs &lt;i&gt;r&lt;/i&gt;=0.62, 95% CI 0.55-0.68 without MCAT; &lt;i&gt;P&lt;/i&gt;=.63). However, it significantly weakened prediction of clinical skills (OSCE: &lt;i&gt;r&lt;/i&gt;=0.37, 95% CI 0.27-0.46 with MCAT vs &lt;i&gt;r&lt;/i&gt;=0.48, 95% CI 0.40-0.56 without MCAT; &lt;i&gt;P&lt;/i&gt;&lt;.001). In tertile analyses, the MCAT modestly improved prediction among top-performing students but eliminated predictive validity in the bottom tertile (&lt;i&gt;r&lt;/i&gt;=0.12, 95% CI -0.07 to 0.30; &lt;i&gt;P&lt;/i&gt;=.21 vs &lt;i&gt;r&lt;/i&gt;=0.28, 95% CI 0.10-0.44 without MCAT; &lt;i&gt;P&lt;/i&gt;=.003). The ML model explained 36% of MCAT variance (&lt;i&gt;R&lt;/i&gt;²=0.36 with 95% CI 0.32-0.41; root mean squared error ≈10% of score range). Incorporating predicted MCAT scores reduced correlations with subsequent performance across all outcomes. Although rankings based on predicted MCAT scores strongly correlated with original rankings (&lt;i&gt;r&lt;/i&gt;=0.80, 95% CI 0.77-0.83), 6.3% (4/64) to 9.4% (6/64) of admission decisions would have changed.&lt;/p&gt;&lt;p&gt;&lt;stron","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e94197"},"PeriodicalIF":13.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13533314/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148876102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cost-Effectiveness of a Blended Learning Training Approach for Health Care Workers Conducting HIV Index Case Testing in Malawi: Secondary Economic Analysis of a Cluster-Randomized Controlled Trial.","authors":"Poonam Rawat, Nora E Rosenberg, Katie R Mollan, Tapiwa A Tembo, Dhrutika Vansia, Jiayu Wang, Mike J Chitani, Maria H Kim, Ireen Chirombo, Caroline Kumbuyo, Victor Mwapasa, Katherine R Simon, Saeed Ahmed, Sarah E Rutstein","doi":"10.2196/95450","DOIUrl":"10.2196/95450","url":null,"abstract":"<p><strong>Background: </strong>Index case testing (ICT) is an effective strategy for HIV case finding, but implementation in low- and middle-income countries (LMICs) is often limited by cost and logistical challenges. Traditional ICT training-centralized and in-person-is costly, disrupts service delivery, and varies in quality.</p><p><strong>Objective: </strong>This study evaluates the cost-effectiveness of a blended learning (BL) implementation package designed to build health care worker capacity for ICT, combining tablet-guided teaching and practice sessions, phone-based feedback, and tablet-guided continuous quality improvement, compared with standard of care (SOC) training.</p><p><strong>Methods: </strong>The Package of Resources for Assisted Contact Tracing: Implementation, Costs, and Effectiveness (PRACTICE) cluster-randomized controlled trial included 33 clusters in southern Malawi from May 2022 to September 2023, randomized 2:1 to SOC (n=22) or SOC + BL implementation package (n=11). Our cost-effectiveness analysis, from the health system perspective, used microcosting, time-and-motion assessments, and observed trial outcomes. The decision tree model estimated total program costs, contact testing outcomes, and incremental cost-effectiveness ratios (ICERs) per contact tested and per person diagnosed with HIV across 1 year of implementation. Sensitivity analyses assessed parameter uncertainty, and a scenario analysis modeled a nationwide scale-up under a decentralized, Ministry of Health-led approach.</p><p><strong>Results: </strong>Our model simulated 100,000 index clients eligible for contact elicitation over 1 year across 2 districts (50,000 per arm). The BL implementation package arm yielded 891 additional contacts tested and 54 more HIV diagnoses. The ICERs were US $125 per contact tested and US $2045 per person diagnosed with HIV; excluding training development costs reduced these to US $69 and US $1136, respectively. Nationwide scale-up under a Ministry of Health-led model further reduced ICERs to US $43 per contact tested and US $698 per person diagnosed with HIV. Probabilistic sensitivity analyses showed the BL implementation package was cost-effective in most simulations.</p><p><strong>Conclusions: </strong>The BL implementation package improved ICT delivery and HIV case finding. Scenario analyses suggested that a decentralized, government-led scale-up could substantially reduce costs and may represent an efficient case-finding strategy, particularly in Malawi's mature epidemic. BL implementation packages provide a scalable, system-integrated, cost-effective approach that could strengthen health care worker capacity across other service delivery areas in low-resource settings.</p><p><strong>International registered report identifier (irrid): </strong>RR2-10.1136/bmjopen-2023-077706.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e95450"},"PeriodicalIF":13.9,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13563035/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148841572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Documentation Quality and Educational Value in AI-Assisted Feedback.","authors":"Amane Endo, Takeshi Kimura, Yuki Kataoka","doi":"10.2196/105029","DOIUrl":"10.2196/105029","url":null,"abstract":"","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e105029"},"PeriodicalIF":13.9,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13515355/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148881840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Inclusive Medical Spanish Education to Address Bias in Clinical Communication: Tutorial on Patient-Informed Module Development.","authors":"Alexandra Lopez Vera","doi":"10.2196/98398","DOIUrl":"10.2196/98398","url":null,"abstract":"<p><strong>Unlabelled: </strong>Spanish-speaking patients who are lesbian, gay, bisexual, transgender, queer, and other minority sexual orientations and gender identities (LGBTQ+) may encounter overlapping barriers in clinical care related to language discordance, misgendering, heteronormative questioning, and culturally unresponsive communication. These barriers can undermine trust, limit disclosure, and negatively affect care experiences and continuity of care. At the same time, medical Spanish curricula in the United States remain highly variable in content, structure, and assessment, and inclusive, bias-aware communication has not been consistently integrated into language training for health professions learners. Our prior qualitative work with Latinx, Hispanic, and Spanish-origin LGBTQ+ adults identified communication priorities directly relevant to educational design, including respectful forms of address, open-ended relationship language, broader recognition of family structures, and reduced reliance on binary or heteronormative assumptions. Our prior curricular work has also shown that structured medical Spanish instruction using self-study, faculty-led teaching, peer support, standardized patient encounters, and performance-based assessment can be implemented in undergraduate medical education. This tutorial presents a patient-informed framework for developing an inclusive medical Spanish module to address bias in clinical communication. Rather than reporting a new intervention study, it synthesizes prior qualitative findings and prior curricular experience to provide practical guidance for educators. The framework describes how to translate patient-reported communication priorities into curricular design principles, learning objectives, technology-enhanced preparatory activities, standardized patient and live-practice components, feedback and assessment strategies, and implementation planning. It argues that inclusive communication should not be treated as optional cultural content or as a narrow language add-on but as a core component of competent clinical performance in Spanish-language care. This tutorial is intended for medical educators, language educators, course directors, and simulation teams seeking to strengthen language-concordant and bias-aware communication training in medical Spanish. By linking patient-informed findings to concrete educational design choices, the framework offers a practical model for integrating inclusive communication throughout instruction, simulation, and assessment. A patient-informed approach of this kind may help health professions programs move beyond general statements about inclusion and toward more deliberate preparation of learners to communicate respectfully and effectively with Spanish-speaking LGBTQ+ patients.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e98398"},"PeriodicalIF":13.9,"publicationDate":"2026-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13496390/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Beyond Pattern Recognition: Call for Functionally Aware AI for Anatomical Illustration.","authors":"Agata Maria Kawalec-Rutkowska, Marian Simka","doi":"10.2196/92700","DOIUrl":"10.2196/92700","url":null,"abstract":"<p><strong>Unlabelled: </strong>This paper critically assesses the role of generative AI in anatomical illustration, identifying fundamental barriers that currently preclude AI from replacing human medical illustrators. Despite the promise of unprecedented efficiency, contemporary models exhibit persistent anatomical inaccuracies and \"hallucinations\" of nonexistent structures-flaws stemming from statistical pattern-matching rather than genuine anatomical understanding. These systems further lack pedagogical intent, clinical context, and the capacity for deliberate visual judgment, while raising unresolved ethical and copyright concerns regarding training data. Although a specialized AI for this purpose is theoretically feasible, its development as a standalone goal remains economically nonviable given the niche nature of the profession. Rather than replacing human illustrators, AI's future role will be augmentative, with the requisite anatomical intelligence likely emerging as a byproduct of broader advances in clinical applications such as surgical planning and personalized medicine. For AI-generated imagery to become educationally and clinically reliable, it will require rigorous human supervision, curated gold standard datasets, and a foundation of genuine anatomical comprehension.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e92700"},"PeriodicalIF":13.9,"publicationDate":"2026-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13493040/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital Standardized Patients: Conceptual Framework for Generative AI-Empowered Medical Education.","authors":"Chuanlin Jia, Lulu Qi","doi":"10.2196/91050","DOIUrl":"10.2196/91050","url":null,"abstract":"<p><strong>Unlabelled: </strong>Generative AI is driving medical education from digital support toward intelligent, interactive learning environments. The cultivation of doctor-patient communication skills requires not only technical proficiency but also communication competence, emotional sensitivity, and ethical judgment. This paper proposes a conceptual framework for AI-driven digital standardized patients (AI-SPs) to provide new approaches for communication training, humanistic education, and emotional engagement in medical curricula. This viewpoint integrates research findings from educational technology and medical education, elaborating the framework from 4 dimensions: system architecture, multimodal interaction, personality modeling, and ethical considerations. AI-SPs can provide adaptive, emotionally responsive interactions in repeatable, controllable simulated scenarios, enabling learners to experience diverse patient characteristics and clinical situations. The proposed \"future learning\" framework emphasizes personalization, contextualization, and reflective learning. During implementation, attention must be paid to data privacy, algorithmic bias, and human supervision. AI-SPs represent an extension of the traditional standardized patient model. They facilitate human-AI collaborative learning, support the cultivation of empathy, and provide a new pathway for the appropriate application of generative AI in medical education.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e91050"},"PeriodicalIF":13.9,"publicationDate":"2026-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13492630/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Olfactory Cues and Patient Safety Behavior in Virtual Reality Medical Simulation: Controlled Trial.","authors":"Anna Junga, Pascal Kockwelp, Jennifer Dabel, Sönke Helmut Scherzer, Benjamin Risse, Markus Holling, Bernhard Marschall, Hendrik Friederichs","doi":"10.2196/93148","DOIUrl":"10.2196/93148","url":null,"abstract":"<p><strong>Background: </strong>Simulation-based medical education is essential for improving patient safety. In virtual reality (VR)-based simulation, immersion is primarily generated through visual and auditory cues, while other sensory modalities are typically absent. This sensory limitation may reduce the emergence of authentic safety-relevant behaviors. Olfaction plays an important role in clinical reasoning, risk perception, and self-protective behavior and is closely linked to memory and emotion. Although olfactory cues have been shown to influence hand hygiene behavior in real or simulated environments, their targeted integration into fully immersive VR-based medical simulation has not been systematically examined.</p><p><strong>Objective: </strong>This study aimed to investigate whether adding a real olfactory cue (disinfectant scent) to a fully virtual clinical simulation increases patient safety-relevant behavior, specifically hand hygiene compliance (hand disinfection and glove usage).</p><p><strong>Methods: </strong>In a controlled study at the University of Münster (during the winter term 2025-2026), 95 medical students participated in a VR-based clinical simulation. Study rooms were preassigned to either an olfactory intervention or a control condition, and participants selected their room without knowledge of the assigned condition (quasi-random allocation by room). Hand hygiene and glove use were automatically tracked as outcomes. Odds ratios (ORs) were calculated to assess the effect of the intervention on these behaviors.</p><p><strong>Results: </strong>The olfactory intervention nearly tripled the odds of hand disinfection (8/43, 18.6% vs 20/52, 38.5%; an absolute risk increase of 19.9 percentage points; OR=2.73, 95% CI 1.09-7.40; <i>P</i>=.04; number needed to treat=5), while no significant difference was observed for glove use (OR=1.44, 95% CI 0.62-3.45; <i>P</i>=.40).</p><p><strong>Conclusions: </strong>The integration of a real olfactory cue into a fully immersive VR medical simulation significantly increased hand disinfection behavior, particularly after patient contact, but did not affect glove use. These findings suggest that olfactory augmentation can selectively reinforce safety-relevant behaviors in digital training environments. Incorporating real-world sensory cues into VR may represent a simple yet effective design strategy to enhance behavioral authenticity and patient safety outcomes in simulation-based medical education.</p>","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e93148"},"PeriodicalIF":13.9,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13489282/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multiagent Large Language Model Framework for Psychotherapy Fidelity Assessment in Motivational Interviewing and Cognitive Behavioral Therapy Training: Cross-Sectional, Simulation-Based Evaluation Study.","authors":"Mohammad Amin Kamaleddin, Mina Mirjalili, Reza Barzegar, Nghia Trung Le, Zachary Cote, Olga Winkler, Lisa Burback, Yanbo Zhang, Osmar R Zaiane, Candice Monson, Divya Sharma, Sri Krishnan, Andrew J Greenshaw, Richard Zeifman, Peter Selby, Venkat Bhat","doi":"10.2196/92964","DOIUrl":"10.2196/92964","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Psychotherapy training is difficult to scale because manual rating of motivational interviewing (MI) and cognitive behavioral therapy (CBT) sessions is time-intensive, requires trained raters, and is subject to rater variability. Large language models (LLMs) may support simulation-based training and rubric-guided scoring, but early-stage evidence is needed before such systems can be applied to real learners.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aimed to conduct a simulation-based evaluation of a multiagent LLM framework for generating and scoring stylized MI and CBT training encounters.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We conducted a cross-sectional evaluation of a multiagent framework comprising Student, Patient, Evaluator, and Feedback agents. The Student agent conducted synthetic MI or CBT encounters with Patient agents derived from structured profiles. The Evaluator agent scored transcripts using study-specific MI and CBT scoring forms. Internal discrimination was tested across prompt-engineered novice, intermediate, and expert Student agent profiles using 133 MI and 102 CBT Patient profiles. Preliminary external grounding was assessed using 133 annotated motivational interviewing (AnnoMI) transcripts with coarse, metadata-derived, high/low session-level quality labels. Agreement with a pragmatic human-rater benchmark was evaluated using 16 independent human raters per modality. Criterion-level comparisons used paired Wilcoxon signed-rank tests with Benjamini-Hochberg false discovery rate correction at &lt;i&gt;q&lt;/i&gt;&lt;0.05. Agreement analyses used intraclass correlation coefficients (ICCs) with bootstrap 95% CIs. A secondary prompt-augmentation sensitivity analysis tested whether appending criterion-referenced feedback text to the novice Student-agent prompt shifted subsequent Evaluator-assigned scores.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Evaluator scores increased across prompt-defined Student-agent competence levels. For MI, overall mean scores increased from 1.18 (95% CI 1.16-1.20) for novice profiles to 1.75 (95% CI 1.69-1.81) for intermediate profiles and to 3.57 (95% CI 3.48-3.66) for expert profiles. For CBT, overall mean scores increased from 0.83 (95% CI 0.78-0.88) to 2.18 (95% CI 2.10-2.26) and 4.38 (95% CI 4.28-4.48), respectively. Criterion-level planned contrasts were significant after false discovery rate correction. On AnnoMI transcripts, Evaluator scores aligned with metadata-derived high/low session labels, with 91.7% classification accuracy at the prespecified threshold. Human interrater reliability was an ICC(2,1) of 0.866 for MI and 0.769 for CBT. Evaluator-vs-human-consensus agreement was an ICC(2,1) of 0.959 for MI and 0.934 for CBT. In the secondary prompt-augmentation analysis, overall MI scores shifted from 1.18 to 1.44, and CBT scores shifted from 0.83 to 1.01.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;This proof-of-concept study suggests that a rubric-guided multiagent LLM framework can score s","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e92964"},"PeriodicalIF":13.9,"publicationDate":"2026-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13483752/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Medical Students' Attitudes, Perceptions, and Self-Reported Familiarity With AI in Health Care: Systematic Review and Meta-Analysis.","authors":"Wei Zhang, Jiaxue Han, Xin Han, Hang Xu, Langkun Wang, Tianhai Lin, Ping Tan, Peng Zhang, Xiaonan Zheng","doi":"10.2196/89411","DOIUrl":"10.2196/89411","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;AI is increasingly encountered in clinical care and medical education, but medical students' attitudes, perceptions, and self-reported familiarity have been assessed using heterogeneous survey instruments, AI referents, and response scales. Prior reviews often combined mixed health profession populations or summarized central estimates without fully showing variation across settings.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aimed to synthesize quantitative evidence on medical students' AI-related attitudes, perceptions, and self-reported familiarity while examining construct harmonization, participant independence, heterogeneity, prediction intervals, risk of bias, and certainty of evidence.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We searched PubMed (MEDLINE), Embase, Web of Science, Scopus, PsycINFO, and Cochrane CENTRAL from inception to April 1, 2026; supplementary searches are described in the appendices. Eligible studies enrolled students in MD, MBBS, MBChB, or DO-equivalent medical programs, or reported separable medical student data from mixed samples. Proportion outcomes were harmonized into 9 domains and synthesized using random-effects meta-analysis with Freeman-Tukey double-arcsine transformation, Hartung-Knapp-Sidik-Jonkman-adjusted CIs, and prediction intervals. Subgroup analyses and meta-regressions were exploratory because of multiple testing, ecological confounding, and construct heterogeneity. Risk of bias and certainty were assessed using the Joanna Briggs Institute analytical cross-sectional checklist and the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) framework, respectively.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Ninety-six cross-sectional studies from 37 countries (&gt;45,000 medical students) were included. Summary estimates suggested favorable attitudes but wide between-setting dispersion. Positive attitude toward AI was 76.9% (95% CI 72.2%-81.4%; prediction interval 42.2%-98.3%; &lt;i&gt;I&lt;/i&gt;²=98.3%; 44 studies; N=20,806), perceived career benefit was 78.4% (95% CI 69.5%-86.2%; prediction interval 45.3%-98.3%; &lt;i&gt;I&lt;/i&gt;²=98.0%; 16 studies; N=9799), and support for curricular integration was 76.6% (95% CI 71.8%-81.1%; prediction interval 47.8%-96.1%; &lt;i&gt;I&lt;/i&gt;²=97.2%; 38 studies; N=16,308). Concern about physician replacement was 39.9% (95% CI 33.6%-46.5%; prediction interval 6.6%-80.1%; &lt;i&gt;I&lt;/i&gt;²=98.8%; 32 studies; N=16,642), willingness to learn about or adopt AI was 71.5% (95% CI 64.8%-77.8%; prediction interval 37.9%-95.4%; &lt;i&gt;I&lt;/i&gt;²=97.7%; 22 studies; N=9199), and ethical concerns were endorsed by 62.8% (95% CI 53.9%-71.3%; prediction interval 21.9%-95.0%; &lt;i&gt;I&lt;/i&gt;²=98.8%; 28 studies; N=14,571). Self-reported familiarity or knowledge was 63.3% (95% CI 55.9%-70.3%; prediction interval 7.8%-100.0%; &lt;i&gt;I&lt;/i&gt;²=99.5%; 52 studies; N=27,817), and trust in AI-assisted decisions was 50.6% (95% CI 28.5%-72.6%; prediction interval 7.3%-93.3%; &lt;i&gt;I&lt;/i&gt;²=98.1%; 8 studies; N=3007). All doma","PeriodicalId":36236,"journal":{"name":"JMIR Medical Education","volume":"12 ","pages":"e89411"},"PeriodicalIF":13.9,"publicationDate":"2026-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13484950/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148798981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}