Medical Teacher最新文献

{"title":"Evaluating hybrid human-LLM coding workflows for qualitative research in medical education: A generalizability study.","authors":"Emily Rush, Md Nazmul Karim, George S Yacu, Jessica N Byram, Colleen N Garnett, Nicole DeVaul, Laura Smith, Margaret Checchi, Daniel Martin, Leslie A Hoffman, Kirsten M Brown, Daniel J Mumbower, Robert M Becker, Victoria A Roach, Alison F Doubleday, Danielle N Edwards, Rebecca S Lufler, Alexandra Wactor, Sophia Boxerman, Suzanne Smith, Hannah L Herriott, Megan E Kruskie, Kyle A Robertson, Elizabeth R Agosto, Christopher Facer, Abdel Metwally, Melissa Barbosa, Dahlia Chavez, Ali Akram, Truman Steele, Seth Adler, Joshua Samaniego, Sara Aqel, Chloie Flores, Yi Gao, Emily Nguyen, Melissa Petito, Adam B Wilson","doi":"10.1080/0142159X.2026.2721362","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2721362","url":null,"abstract":"<p><strong>Introduction: </strong>Large language models (LLMs) are increasingly proposed as deductive coders in qualitative research, but their measurement properties remain underexplored. This study applies generalizability theory to evaluate whether hybrid human-LLM workflow configurations can achieve reliable mode-outcomes as an alternative consensus-generating mechanism for deductive coding tasks in medical education research.</p><p><strong>Methods: </strong>Three commercial LLMs (GPT-5.2, Claude Opus 4.5, Gemini 3-Flash Preview) coded 741 excerpts from a published audit of AI-related policy documents at 146 U.S. medical schools against a 24-subtheme deductive framework. Mixed-effects logistic regression assessed variability in agreement with human consensus across coder type (human versus LLM), excerpt characteristics (complexity and length), and coding conditions (sequential independent versus batched processing). A simulation-based D-study forecasted agreement levels for various hybrid human-LLM configurations.</p><p><strong>Results: </strong>Sequential independent LLM coding showed agreement comparable to human coding (<i>β</i> = -0.14, <i>p</i> = 0.326). Batched processing showed significantly lower human-LLM agreement (<i>β</i> = -0.41, <i>p</i> = 0.007) and substantial batch-related variance (variance = 30.16). The three LLMs produced similar agreement (joint Wald χ² = 0.15, <i>p</i> = 0.930), and LLM-LLM agreement (<i>κ</i> = 0.750-0.755) substantially exceeded human-human agreement (<i>κ</i> = 0.422) and human-versus-consensus agreement (<i>κ</i> = 0.420-0.533). LLM disagreements were more systematically patterned across the codebook (Cramer's <i>V</i> = 0.554-0.585) than human disagreements (Cramer's <i>V</i> = 0.196). Simulations forecasted agreement ranging from 0.520 to 0.529 when two or more LLMs were paired with one to two human coders. These forecasted agreement levels for hybrid human-LLM workflows fell within the range of human-versus-consensus agreement.</p><p><strong>Discussion: </strong>D-study simulations support the use of hybrid human-LLM workflows to reach coding consensus for deductive reasoning tasks through mode responses, an alternative consensus-generating mechanism to traditional adjudication discussions. Future work should examine whether these patterns extend across additional deductive coding contexts and model families.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-13"},"PeriodicalIF":4.0,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148830201","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Response to: 'On the search sensitivity of a meta-analysis of AI-generated MCQs'.","authors":"Noor H S Alani, Jay Amaranathan, Abdulhalim Dandoush, Sally Rye, Paula Toko King, Marshall H Chin","doi":"10.1080/0142159X.2026.2716263","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2716263","url":null,"abstract":"","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1"},"PeriodicalIF":4.0,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148813524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Two-point regression standard setting, improving discrimination at both ends of the grading scale in OSCEs.","authors":"Andrew M Lunn, Christopher J Harrison, John C McLachlan","doi":"10.1080/0142159X.2026.2721367","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2721367","url":null,"abstract":"<p><strong>Background: </strong>Standard setting in Objective Structured Clinical Examinations (OSCEs) typically identifies a single cut score to distinguish competent from non-competent candidates. However, this approach does not help defensible differentiation between 'competent' and 'excellent' students.</p><p><strong>Objective: </strong>This study introduces a Two-Point Regression model, an extension of the borderline regression method, to establish two cut scores, one for competence and one for excellence, enabling more nuanced classification of student performance.</p><p><strong>Methods: </strong>A retrospective analysis was performed on 10 OSCEs delivered to UK medical students. For each station, linear regressions of global ratings (0-3) against checklist scores were used to generate the traditional pass cut score (cX = 1). Additional thresholds for excellence were generated at global thresholds eX = 2, 2.25, 2.5 and 3. Outcomes included: proportion graded as excellent (≥70% after scaling), mean scaled grades, and mean global scores of 'excellent' students. Statistical analyses included chi-square, ANOVA, and post-hoc testing as appropriate.</p><p><strong>Results: </strong>Pass/fail classifications were unchanged across all methods. The Two-Point regression model substantially reduced the proportion of students achieving excellence from 47% to 17.2% when the excellence cut score was set to a global threshold of eX = 2.25, aligning with national benchmarks. Mean global scores of 'excellent' students increased in line with threshold choice, demonstrating strong internal consistency (R²=0.9939).</p><p><strong>Conclusion: </strong>The Two-Point Regression method offers an easily implementable and defensible approach for identifying excellence in numerically graded OSCEs, improving alignment between examiner global judgements and awarded grades. Threshold choice should be context-dependent, but the Two-Point Regression method provides a robust and adaptable framework.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-11"},"PeriodicalIF":4.0,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148819044","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Twelve tips for educators supporting learner-centered AI integration in undergraduate medical education.","authors":"Diego F Niño, Ken Masters, Rakesh Patel, Sola Aoun Bahous, Davy Cheng, Alexis L Rossi","doi":"10.1080/0142159X.2026.2718319","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2718319","url":null,"abstract":"<p><p>Artificial intelligence is reshaping medical education at a pace that outstrips most institutions' capacity to respond, yet practical, multi-domain guidance for educators remains limited. This article translates the first three levels of the International Advisory Committee for Artificial Intelligence (IACAI) Learner Matrix, the Intrapersonal, Micro, and Meso levels of a five-level socio-ecological framework, into 12 practical tips for educators and curriculum designers. The tips span foundational AI culture and literacy, ethical practice, critical tool evaluation, learning design, clinical skills integration, assessment, career preparation, curriculum co-design, research integrity, workforce readiness, and learner well-being. Each tip identifies a core challenge and offers concrete implementation strategies at the learner and institutional levels. Taken together, the tips provide a coherent, learner-centered scaffold to help educators worldwide prepare learners for AI-augmented practice.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-16"},"PeriodicalIF":4.0,"publicationDate":"2026-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795385","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Different feedback practices across health professions in a shared clinical workplace.","authors":"Stefanie Lea Mosimann, Maurizio Alen Trippolini, Karen Dominique Könings","doi":"10.1080/0142159X.2026.2712420","DOIUrl":"10.1080/0142159X.2026.2712420","url":null,"abstract":"<p><strong>Introduction: </strong>Feedback in clinical health professions workplaces often lacks the features known to promote competence development. A constructive interprofessional feedback culture calls for professionals to have aligned ideas about feedback giving and receiving but little is known about feedback characteristics across professions within the same workplace. This study explores perceptions and challenges of feedback in nursing, physiotherapy, and medicine using the R2C2 feedback model as a framework.</p><p><strong>Methods: </strong>We conducted a qualitative study in a medical department of a Swiss hospital. Purposefully sampled physicians (both residents and attendings), nurses, and physiotherapists participated in audio-recorded, semi-structured interviews. Transcripts were analyzed iteratively alongside data collection.</p><p><strong>Results: </strong>We identified various promoting and inhibiting factors for effective feedback conversations related to the four components of the R2C2 model. Across professions, feedback quality was depending on interpersonal relationships, reluctance to provide feedback to avoid conflict, and limited time. We recognized distinct differences in feedback orientation in the different professions, with nurses emphazising team feedback and the culture of niceness, physiotherapists perceiving feedback as a routine learner-centered coaching opportunity, and both physician groups experiencing feedback as a one-way transmission of information and understanding feedback as <i>of</i>, not <i>for</i> learning.</p><p><strong>Discussion: </strong>Distinct profession-specific feedback orientations were identified within a shared clinical workplace, rather than shared feedback practices among professions. Feedback was relationally framed in nursing, coaching-oriented in physiotherapists, and evaluative in medicine. These findings suggest that professional socialisation shapes feedback practices even within a common organizational environment and warrant further research into how such differences influence feedback across professional boundaries.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-11"},"PeriodicalIF":4.0,"publicationDate":"2026-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795350","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence in health professions education publishing: A content analysis of journal and publisher guidance.","authors":"Sruthi Saravanan, See Chai Carol Chan, Mohammed Ahmed Rashid","doi":"10.1080/0142159X.2026.2712427","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2712427","url":null,"abstract":"<p><strong>Introduction: </strong>Artificial intelligence (AI) is being widely used by authors, reviewers and editors in the academic publishing process. However, there remains ambiguity about the boundaries of AI use. Amid this growing uncertainty, editors and publishers face an ongoing tension between a desire to advance the academic publishing process and a need to uphold scientific integrity. This study examined advisory and guideline statements about AI use by health professions education (HPE) journals.</p><p><strong>Methods: </strong>We collected a corpus of texts relating to the use of AI in HPE journals from 22 HPE journals, 6 respective publishers and the Committee on Publication Ethics (COPE) policies in January 2026 relating to the use of AI in academic publishing. The corpus was examined using content analysis.</p><p><strong>Results: </strong>Five main themes were identified from the analysis: 1) Accountable governance, 2) Protection of ethical standards, 3) Safeguarding of research integrity, 4) Maintenance of human oversight, and 5) Balancing of opportunities and risk.</p><p><strong>Discussion: </strong>Overall, HPE journals predominantly frame the use of AI in academic publishing as a risk rather than an opportunity, with the overarching concern being a potential compromise of scientific integrity. Editors need to manage various tensions related to AI use and disclosure, including ensuring that policies keep up with the speed of technological advancement, and potentially broader impacts on the scholarly HPE field. As guidance on AI use continues to evolve, authors, reviewers and editors must ensure that they actively maintain and optimise their policies and practices in response to emerging evidence, including of unintended consequences.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-6"},"PeriodicalIF":4.0,"publicationDate":"2026-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Beyond admissions: Using situational judgment test scores to predict learner performance throughout undergraduate medical education.","authors":"Alexandra H Vinson, Suzy McTaggart, Kevin W Walsh, Carol A Teener, Rebecca E Weeks, Deborah R Berman, Seetha U Monrad","doi":"10.1080/0142159X.2026.2718314","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2718314","url":null,"abstract":"<p><strong>Introduction: </strong>Situational judgment tests (SJTs) have been introduced in medical school admissions across multiple national contexts over the past three decades. However, ongoing research is needed to understand SJTs' role in predicting outcomes across undergraduate medical education (UME), particularly in the context of institution-specific approaches to medical student assessment. This study investigated the role of the Casper SJT for understanding student performance in professionalism, communication and self-awareness at three key timepoints in admissions and UME: interview day scores, the preclinical phase, and the clinical phase.</p><p><strong>Method: </strong>We conducted a retrospective exploratory analysis of associations between the Casper score, interview day scores, and medical student assessment data from 508 students at the University of Michigan Medical School. Using sequential logistic regression models, we linked student data across admissions and UME phases, using milestone assessments and professionalism referrals as outcome variables. To better understand the role of the composite SJT score in understanding construct-specific UME assessments, we mapped Casper constructs to institution-specific assessments of professionalism, communication skills, and self-awareness.</p><p><strong>Results: </strong>When modeling the relationship between Casper SJT scores and communication competency domains in UME, we found that including the SJT score in the communication and self-awareness domains models modestly improved the model outcomes. However, our modeling analysis demonstrated limited utility for predicting professionalism lapses during medical school using SJT scores.</p><p><strong>Discussion: </strong>As SJTs are adopted in selection processes, it is crucial to investigate whether SJT scores predict future student success in the context of real-world institutional variation in applicant pool, admissions processes, and UME assessment programs. Our analysis contributes to efforts to establish the effectiveness of the Casper SJT for predicting temporally distant and competency-specific outcomes by demonstrating the utility of sequential logistic regression modeling methods. Our discussion critically considers the use of composite SJT scores to predict for construct-specific UME assessments.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-10"},"PeriodicalIF":4.0,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Serious games and gamification in emergency and critical care training: Implications for health professions education.","authors":"Kun Li, Xuejia Pan, Xueling Fang","doi":"10.1080/0142159X.2026.2710216","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2710216","url":null,"abstract":"<p><strong>Background: </strong>Serious games and gamification are increasingly used in emergency and critical care training, but the literature is heterogeneous and often overlaps conceptually with digital simulation. For medical education, it remains unclear how these interventions are being designed and how deeply their outcomes are being evaluated.</p><p><strong>Methods: </strong>We conducted a scoping review following PRISMA-ScR guidance and registered the protocol on the Open Science Framework. PubMed, Web of Science Core Collection, and EBSCOhost (CINAHL Plus with Full Text and ERIC) were searched from inception to 22 February 2026 without language restrictions. We included empirical studies of serious games, gamification, or game-based learning in emergency, resuscitation, trauma, intensive care, and related high-acuity training contexts. Outcomes were mapped to the highest reported Kirkpatrick level.</p><p><strong>Results: </strong>Fifty-two unique studies were included from 669 records, with marked growth in recent years. Interventions were grouped into six primary modality families: digital serious games and virtual-patient environments (27/52, 51.92%), gamification-enhanced teaching or platform-based interventions (9/52, 17.31%), escape rooms (6/52, 11.54%), gamified simulation competitions or deliberate-practice formats (4/52, 7.69%), analogue board or card games (3/52, 5.77%), and virtual reality or augmented reality gamified simulations (3/52, 5.77%). Evaluation remained predominantly proximal: the highest reported Kirkpatrick level was Level 2b in 29 studies (55.77%), Level 2a in nine (17.31%), Level 1 in six (11.54%), and Level 3 in eight (15.38%); no study reported Level 4 outcomes.</p><p><strong>Conclusions: </strong>The field is expanding rapidly, but most evaluations remain focused on engagement and short-term learning rather than behavioural transfer or downstream clinical impact. In high-acuity health professions education, game-based approaches appear most useful as adjuncts for pretraining, reinforcement, and distributed practice. Future research should improve conceptual clarity and align educational objectives, mechanisms, game mechanics, and outcome measures more explicitly.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-12"},"PeriodicalIF":4.0,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"When the profession itself is contested: Transformative learning and divergent identity pathways among first-year medical students during a macro-social crisis.","authors":"Ye Ji Kang, Jaehee Rho, Do-Hwan Kim","doi":"10.1080/0142159X.2026.2718306","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2718306","url":null,"abstract":"<p><strong>Introduction: </strong>Professional identity formation (PIF) research in medical education has largely examined micro-level dilemmas within relatively stable training environments. Less is known about how macro-social crises destabilizing the profession's public legitimacy shape early PIF. This study explored how first-year medical students narrated the impact of the 2024-2025 doctor-government conflict in South Korea on their professional identity development, using Transformative Learning Theory (TLT) as an interpretive lens.</p><p><strong>Methods: </strong>We conducted a theory-informed qualitative study using written reflective essays submitted by first-year medical students as a professionalism course assignment upon returning to the formal curriculum. Using Big Q reflexive thematic analysis, we combined deductive use of TLT as a sensitizing concept with inductive latent analysis to examine students' pre-crisis assumptions, disorienting dilemmas, critical reflection, and evolving identities.</p><p><strong>Results: </strong>Students' narratives were interpretable through four TLT-informed phases: pre-crisis frames of reference, disorienting dilemma, critical reflection, and evolving identity. Pre-crisis configurations included idealized, perfectionistic, system-naive, meritocratic, and hierarchical assumptions about medicine. Disorienting dilemmas centered on collapsing trust in the healthcare system, perceived patient harm, moral ambiguity, social rejection, and internal fragmentation within the medical community. Critical reflection involved deconstructing heroic and individualistic views of doctoring, reattributing problems to structural conditions, and renegotiating meanings of responsibility, competence, and solidarity. Many students described reconstructive identity work, culminating in more socially embedded, system-aware, and pragmatically collaborative identities. Others showed defensive retrenchment, reinforcing grievance, distrust, and a sense of professional victimhood.</p><p><strong>Discussion: </strong>Macro-social crisis can function as a collective disorienting dilemma in early medical education, extending PIF and transformative learning beyond micro-level clinical experiences. Identity development during legitimacy shocks is not uniformly growth-promoting; it depends on reflective and relational conditions under which students process disruption. Medical schools should support early identity work through structured dialogue, attention to belonging, and engagement with healthcare policy.</p>","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1-13"},"PeriodicalIF":4.0,"publicationDate":"2026-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148795374","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Students with disabilities - the examiners struggle.","authors":"Marc Vorstenbosch, Ellemieke Rasenberg, André Bremers","doi":"10.1080/0142159X.2026.2708988","DOIUrl":"https://doi.org/10.1080/0142159X.2026.2708988","url":null,"abstract":"","PeriodicalId":18643,"journal":{"name":"Medical Teacher","volume":" ","pages":"1"},"PeriodicalIF":4.0,"publicationDate":"2026-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148761426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"教育学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}