The DOI matches its Crossref record; 47 of the 47 references it cites were found in a registry or an index; 15 of the 47 sources its claims cite were checked for support.
Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study
Open accessAbstract
Background: As AI fundamentally transforms the healthcare landscape, medical education curricula have struggled to keep pace with these technological shifts. While current research has established a foundation for general AI literacy, less attention has been given to the role-specific competencies required for the diverse functions that healthcare professionals perform in an AI-integrated environment. Although existing tiered and domain-based competency models have clarified general AI competency requirements, they offer limited guidance on how such competencies should be differentiated according to the roles healthcare professionals perform in practice. Objective: This study aimed to explore how AI-driven changes in healthcare shape competency requirements across professional roles and to develop differentiated competency frameworks and curriculum implications for 3 distinct roles: users, developers, and leaders. Methods: Using a qualitative research design, we conducted in-depth interviews with 13 subject matter experts across academic medicine and dentistry, clinical practice, and the healthcare AI industry, who were recruited through 4 independent channels. Data were analyzed using reflexive thematic analysis with inductive coding organized around 3 research questions and published competency frameworks serving as sensitizing concepts. Trustworthiness was supported through investigator triangulation, member checking, an audit trail, and attention to researcher reflexivity and positionality. Results: We identified a role-differentiated competency framework with a layered structure as follows: (1) Users require machine learning and data literacy to understand where algorithms fail and scrutinize data origins and biases; the ability to select best-fit AI solutions and set appropriate thresholds for human-machine delegation; and a high degree of AI-related professionalism to work responsibly with AI by recognizing its limits, critically appraising outputs, maintaining clinical accountability, preserving patient-centered care, and updating knowledge continuously; (2) Developers add operational competencies, including validating systems across technical performance, clinical relevance, and regulatory compliance; bridging the language gap between clinical needs and technical refinements; and fostering a challenging, problem-solving mindset to address real-world clinical bottlenecks; and (3) Leaders focus on system-level competencies, including macro-level strategic planning, governing medical AI across its full lifecycle—selection, validation, deployment, and monitoring—through policy “roads” and governance frameworks, and optimizing systemic resource utilization while orchestrating trust-based collaboration that safeguards care quality. Rather than a strictly cumulative hierarchy, the roles represent analytic distinctions with shared foundational competencies. This framework translates these competencies into corresponding curriculum implications across the health professions education continuum. Conclusions: This study identifies role-differentiated AI competencies and proposes a corresponding curriculum scaffold for healthcare education. As a preliminary, hypothesis-generating starting point requiring multistakeholder validation, the proposed framework can help prepare the future healthcare workforce to utilize AI responsibly and to govern and lead the next generation of digital health innovation.
Overview
CoChat Summary
TL;DR
This qualitative study interviewed 13 subject matter experts to develop a role-differentiated AI competency framework for health professions education across users, developers, and leaders, with corresponding curriculum modules. It proposes a layered, hypothesis-generating framework in which foundational competencies are shared but role emphases differ.
Key findings
- Experts anticipated four AI-driven healthcare changes: greater patient empowerment and more horizontal doctor-patient relationships; redefinition/reconfiguration of medical expertise; transformation of clinical workflows through human-machine collaboration; and improvement in average healthcare quality through reduced operator variability.
- The study identified role-differentiated competencies: users require AI/ML literacy, data literacy, AI solution selection and utilization, and AI professionalism (U-1 to U-4); developers add advanced user foundations, effectiveness evaluation, field demand communication, and a challenging mindset/problem-solving (D-1 to D-4); leaders require systemic insight and strategic planning, AI governance and policy design, and resource management and leadership (L-1 to L-3).
- Curriculum implications were mapped to 4 user modules, 4 developer modules, and 3 leader modules; the roles are analytic distinctions with shared foundational competencies rather than a strictly cumulative hierarchy.
- Experts disagreed on the appropriate scope and depth of AI/ML instruction for all learners, though the authors tentatively conclude that a minimal level of ML and data literacy should be included.
Methods
- Qualitative design using in-depth semistructured interviews with 13 subject matter experts recruited through 4 independent channels; interviews were conducted from October 2022 to October 2025, lasted 45–90 minutes, and were held in person (12) or via Zoom (1).
- Data were analyzed using reflexive thematic analysis with inductive coding organized around 3 research questions, with published competency frameworks serving as sensitizing concepts; trustworthiness included investigator triangulation, member checking with all 13 SMEs (7 provided detailed written feedback), an audit trail, and attention to reflexivity and positionality.
- Curriculum implications were derived by mapping participant-proposed educational content to competency themes and organizing role-specific modules from foundational to applied content.
Limitations
- Stated: the framework was derived from a limited number of expert interviews situated largely within a single national context in East Asia, so findings should be interpreted with attention to healthcare delivery, reimbursement, and regulatory/policy differences; it is an adaptable template rather than one-size-fits-all.
- Stated: several dental experts were included, so findings may reflect AI use cases more visible in imaging-intensive, CAD/CAM-supported, and procedure-oriented workflows; transferability to other specialties and care contexts remains to be empirically established.
- Stated: interrater reliability was not formally assessed, limiting quantitative evidence of coding consistency; the framework is preliminary and hypothesis-generating, requiring multistakeholder validation, pilot implementation studies, and iterative refinement. Evident from the text: no formal pilot was conducted, and the first interview served only as a calibration check.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
7 Valid · 2 Not valid · 15 Not checked · 126 Own results and interpretations
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“However, they generally organize competencies by educational stakeholder, level of expertise, training phase, or institutional implementation level, rather than by the functional responsibilities healthcare professionals assume in AI-enabled practice.”
§ Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study > Introduction
What the sources say
- Partly supportedcites [8] Schubert, AI education for clinicians, EClinicalMedicine sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
- Disputedcites [15] Ng, Artificial intelligence education: an evidence-based medicine approach for consumers, translators, and developers, Cell Rep Med sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 1 partially supported, 1 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
Checked by a reasoning model — may be wrong.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Sample size was determined through iterative monitoring of analytic stabilization—sometimes termed data saturation, though we use this term cautiously given the constructivist orientation of reflexive thematic analysis [16]—rather than a prespecified target.”
§ Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study > Methods > Data Collection
What the sources say
- Could not verifycites [16] Braun, Toward good practice in thematic analysis: avoiding common problems and be(com)ing a researcher, Int J Transgend Health sourceExperimental — accuracy not yet measured
The paper's “Everything changes…” section says saturation is not a meaningful concept for reflexive thematic analysis, which explains the citing sentence's caution about that term. It does not say that sample size was determined by iteratively monitoring analytic stabilization rather than by a prespecified target; searches for those sample-size terms found no supporting passage. Page numbers were not available in our copy. (LLM, conf=0.74)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Figure 2 shows that recent literature has largely prioritized user-level competencies for medical students and practicing health professionals [2,6-8].”
§ Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study > Discussion > Principal Findings
What the sources say
- Not checked yetcites [2] Russell, Competencies for the use of artificial intelligence-based tools by health care professionals, Acad Med source
- Not checked yetcites [6] Lee, Defining medical AI competencies for medical school graduates: outcomes of a Delphi survey and medical student/educator questionnaire of South Korean medical schools, Acad Med source
- Partly supportedcites [7] Moldt, Assessing AI awareness and identifying essential competencies: insights from key stakeholders in integrating AI into medical education, JMIR Med Educ sourceExperimental — accuracy not yet measured
The Results section (Identification of AI Competencies and Implications for Medical Curriculum; page unavailable) identifies skills medical students need to use AI, and the Discussion (Key Competencies for Health Science Students; page unavailable) links those skills to future health care practice. But the paper does not establish that recent literature has largely prioritized user-level competencies, or that practicing professionals are a principal target of that literature. (LLM, conf=0.82)
- Partly supportedcites [8] Schubert, AI education for clinicians, EClinicalMedicine sourceExperimental — accuracy not yet measured
In “What are the AI skills that future clinicians should be equipped with?” and “How should medical schools incorporate AI training into their existing curricula?” (pages unavailable), the paper emphasizes practical AI-use and appraisal skills for clinicians and medical students. It does not establish that recent literature has largely prioritized those competencies; that characterization extends beyond what this source says. (LLM, conf=0.82)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Every source the paper cites for this claim was checked and supports it.
In the paper
“Faculty readiness, raised by several experts, represents a further barrier to implementation [41,44].”
§ Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study > Discussion > Principal Findings
What the sources say
- Supportedcites [41] Charow, Artificial intelligence education programs for health care professionals: scoping review, JMIR Med Educ source
In Results, under “What Were the Critical Implementation Factors?—Barriers” (page not provided), the review identifies varying AI literacy among faculty as a major barrier to implementing AI education programs. It reports that multiple papers discussed limited faculty knowledge, technical expertise, or time to teach. (LLM, conf=0.90)
- Supportedcites [44] Ngo, The cases for and against artificial intelligence in the medical school curriculum, Radiol Artif Intell source
The Conclusion identifies a paucity of faculty expertise on AI as a main obstacle to integrating AI into the medical school curriculum. The Introduction likewise describes a dearth of qualified instructors. These passages support the faculty-readiness barrier; the reference to several experts is not needed to establish that point. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and a high degree of AI-related professionalism to work responsibly with AI by recognizing its limits, critically appraising outputs, maintaining clinical accountability, preserving patient-centered care, and updating knowledge continuously;”
§ Role-Differentiated AI Competencies and Curriculum Implications for Health Professions Education: Qualitative Study > Abstract > Results
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
32 more claims
Create a free account to see all 150 claims and to verify the 32 references still open.
- Probably not a substantial claim. 87
- It may state two separate claims as one. 68
- The quote only partly states this claim. 1
Citations
Cites
Sample Size in Qualitative Interview Studies: Guided by Information Power
Malterud, K., Siersma, V. D., & Guassora, A. D. (2016) · Qualitative Health Research · doi:10.1177/1049732315617444 · Cited by 12,703
Standards for Reporting Qualitative Research: A Synthesis of Recommendations
O’Brien, B. C., Harris, I. B., Beckman, T. J., Reed, D. A., & Cook, D. A. (2014) · Academic Medicine · doi:10.1097/acm.0000000000000388 · Cited by 12,211
Thematic analysis of qualitative data: AMEE Guide No. 131
Kiger, M. E., & Varpio, L. (2020) · Medical Teacher · doi:10.1080/0142159x.2020.1755030 · Cited by 3,787
Braun, V., & Clarke, V. (2023) · International Journal of Transgender Health · doi:10.1080/26895269.2022.2129597 · Cited by 2,878
Inductive/deductive hybrid thematic analysis in mixed methods research
Proudfoot, K. (2023) · Journal of Mixed Methods Research · doi:10.1177/15586898221126816 · Cited by 934
A scoping review of artificial intelligence in medical education: BEME Guide No. 84
Gordon, M., Daniel, M., Ajiboye, A., Uraiby, H., Xu, N. Y., Bartlett, R., Hanson, J., Haas, M., Spadafore, M., Grafton-Clarke, C., et al. (2024) · Medical Teacher · doi:10.1080/0142159x.2024.2314198 · Cited by 679
AI competency framework for teachers
(2024) · doi:10.54675/zjte2084 · Cited by 661
Introducing artificial intelligence training in medical education
Paranjape, K., Schinkel, M., Nannan Panday, R., Car, J., & Nanayakkara, P. (2019) · JMIR Medical Education · doi:10.2196/16048 · Cited by 634
SHIFTing artificial intelligence to be responsible in healthcare: a systematic review
Siala, H., & Wang, Y. (2022) · Social Science & Medicine · doi:10.1016/j.socscimed.2022.114782 · Cited by 358
Artificial intelligence in health care: accountability and safety
Habli, I., Lawton, T., & Porter, Z. (2020) · Bulletin of the World Health Organization · doi:10.2471/blt.19.237487 · Cited by 357
Research works printed but not linked: 1
- Lincoln YS Guba EG . Naturalistic Inquiry. Sage Publications; 1985. ISBN: 9780803924314
Cited by
Still being checked.
Metadata
- Authors
- Park, Y., Han, Y., Kim, H., Ahn, S., Kron, F. W., & Lee, J.
- Affiliations
- Seoul National University (KR)
- Yeungnam University (KR)
- University of Illinois Urbana-Champaign (US)
- Old Dominion University (US)
- Corresponding author
- Lee, Jihyun
- Venue
- JMIR Medical Education
- Publisher
- JMIR Publications Inc.
- Published
- 2026-09-08
- Volume
- 12
- Pages
- e97608-e97608
- ISSN
- 2369-3762 (ISSN-L)
- PMC
- PMC13552830
- OpenAlex
- W7211973760
- Semantic Scholar
- ab9d9c8c794d77d89912490dd1ed0d4d2f5fb284
- Cited by
- 0
- References
- 47
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- JMIR Medical Education (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- ODU Digital Commons (Old Dominion University) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Funders
- National Research Foundation
- National Research Foundation of Korea
- Ministry of Science and ICT, South Korea (2021R1F1A1056465)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Innovations in Medical Education
- Health Sciences › Health Professions › General Health Professions › Health and Medical Education
- Keywords
- artificial intelligence in healthcare, AI competencies, health professions education, medical curriculum, AI literacy, AI developers, AI users, AI leadership, AI governance, data literacy, machine learning, clinical accountability, human-machine collaboration, competency framework, dental education, responsible AI use, career readiness
- MeSH terms
- Artificial Intelligence, Curriculum, Health Occupations, Humans, Professional Competence, Qualitative Research
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.