The DOI matches its Crossref record; 21 of the 21 references it cites were found in a registry or an index; 7 of the 16 sources its claims cite were checked for support.
Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study
Open accessAbstract
Background: Artificial intelligence (AI) tools are designed to create or generate content from their trained parameters using an online conversational interface. AI has opened new avenues in redefining the role boundaries of teachers and learners and has the potential to impact the teaching-learning process. Methods: In this descriptive proof-of- concept cross-sectional study we have explored the application of three generative AI tools on drug treatment of hypertension theme to generate: (1) specific learning outcomes (SLOs); (2) test items (MCQs- A type and case cluster; SAQs; OSPE); (3) test standard-setting parameters for medical students. Results: Analysis of AI-generated output showed profound homology but divergence in quality and responsiveness to refining search queries. The SLOs identified key domains of antihypertensive pharmacology and therapeutics relevant to stages of the medical program, stated with appropriate action verbs as per Bloom’s taxonomy. Test items often had clinical vignettes aligned with the key domain stated in search queries. Some test items related to A-type MCQs had construction defects, multiple correct answers, and dubious appropriateness to the learner’s stage. ChatGPT generated explanations for test items, this enhancing usefulness to support self-study by learners. Integrated case-cluster items had focused clinical case description vignettes, integration across disciplines, and targeted higher levels of competencies. The response of AI tools on standard-setting varied. Individual questions for each SAQ clinical scenario were mostly open-ended. The AI-generated OSPE test items were appropriate for the learner’s stage and identified relevant pharmacotherapeutic issues. The model answers supplied for both SAQs and OSPEs can aid course instructors in planning classroom lessons, identifying suitable instructional methods, establishing rubrics for grading, and for learners as a study guide. Key lessons learnt for improving AI-generated test item quality are outlined. Conclusions: AI tools are useful adjuncts to plan instructional methods, identify themes for test blueprinting, generate test items, and guide test standard-setting appropriate to learners’ stage in the medical program. However, experts need to review the content validity of AI-generated output. We expect AIs to influence the medical education landscape to empower learners, and to align competencies with curriculum implementation. AI literacy is an essential competency for health professionals.
Overview
CoChat Summary
TL;DR
This proof-of-concept cross-sectional study queried three generative AI tools to produce pharmacology/therapeutics learning objectives and assessment items on hypertension for pre-clerkship and graduating medical students. AI outputs were broadly homologous and often usable, but required expert content-validity review because A-type MCQs and integrated case clusters showed construction defects, multiple correct answers, and variable standard-setting responses.
Key findings
- All three AI platforms generated specific learning objectives with Bloom’s taxonomy action verbs; pre-clerkship and graduation SLOs covered JNC-recommended antihypertensive classes, mechanism of action, pharmacokinetics, adverse effects, indications/contraindications, monitoring, and mono/combination therapy, with graduation SLOs using higher-cognitive verbs.
- A-type MCQs were often quality items across mechanism, pharmacokinetics, adverse reactions, and indications, but Table 2 reports errors: e.g., Sage Poe pre-clerkship pharmacokinetics had item construction defects in 5 items; ChatGPT pre-clerkship mechanism-of-action had 2 construction defects; multiple correct options occurred across platforms. ChatGPT supplied explanations useful for self-study, though some were incomplete or defective.
- Integrated case-cluster MCQs had focused clinical vignettes and horizontal integration, mostly targeting “knows” or “knows how” levels, but Table 3 shows they generally had at least one construction defect and deficient comprehensiveness; ChatGPT often included “None of the above.”
- SAQs were technically accurate and comprehensive but had construction defects; OSPEs were technically accurate, comprehensive, level-appropriate, and free of construction defects for ChatGPT and Claude-Instant (Sage Poe unavailable). Standard-setting simulation varied: Sage Poe estimated 60–90% correct for mechanism, 65–85% pharmacokinetics, 60–80% adverse reactions, and 75–90% indications; ChatGPT estimated 67–92%, 63–86%, 65–82%, and 64–89%, respectively; Claude-Instant declined a full simulation.
Methods
- Observational, cross-sectional proof-of-concept study conducted in the Department of Pharmacology & Therapeutics, College of Medicine and Medical Sciences, Arabian Gulf University, Bahrain, April–August 2023, with study procedures in May–June 2023; no human interaction or personal data, so ethics approval was not sought.
- Three AI platforms accessed via the Poe chatbot interface—Sage Poe, Claude-Instant, and ChatGPT version 3.5—were prompted to generate SLOs, A-type MCQs, integrated case-cluster MCQs, SAQs, and OSPE test items on systemic hypertension/antihypertensives for pre-clerkship and graduation-level students, plus standard-setting simulation estimates.
- Both authors independently evaluated AI outputs and reached consensus; answers were cross-checked against JNC-8 and Goodman & Gilman’s *The Pharmacological Basis of Therapeutics* (2023), and items were assessed using a preliminary conceptual framework covering technical accuracy, comprehensiveness, education level, and freedom from construction defects, categorized as complete or deficient.
Limitations
- Stated in the paper: this was a single case study (hypertension) pilot; AI-generated items were not administered to students, so test-item discrimination and difficulty indices could not be assessed; the conceptual framework needs validation in a larger population; flowcharts, algorithms, and figures were not evaluated; and generalizability to other disorders, disciplines, and curricula is uncertain.
- Evident from the text: only three AI tools/versions were tested at one time point, standard-setting estimates were hypothetical AI simulations rather than real student performance, and outputs required expert review for content validity before use.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
7 Valid · 4 Not valid · 8 Not checked · 93 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Data suggests that only a third of academics in medical schools have formal training in medical education [16] which is a limitation.”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Discussion
What the sources say
- Partly supportedcites [16] Trainor A, Richards JB. Training medical educators to teach: bridging the gap between perception and reality. Isr J Health Policy Res. 2021;10(1):75. sourceExperimental — accuracy not yet measured
The Abstract reports that 35% of surveyed teaching physicians in three specialties at four Israeli faculties of medicine had received some training in medical education skills. That supports the approximate one-third figure, but the source does not establish that only a third of medical-school academics generally have formal training; the reported measure was any training among the physicians who responded. (LLM, conf=0.88)
hedged- 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 support only part of this claim — read each source’s verdict below.
In the paper
“A recent scoping review evaluating the barriers to writing quality test items based on 13 studies has concluded that motivation, time constraints, and scheduling were the most common [19].”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Discussion
What the sources say
- Partly supportedcites [19] Karthikeyan S, O’Connor E, Hu W. Barriers and facilitators to writing quality items for medical school assessments - a scoping review. BMC Med Educ. 2019;19(1):123. sourceExperimental — accuracy not yet measured
The Abstract (Results; page not available) says the review selected 13 articles and identifies motivation, time constraints, and scheduling as barriers to quality item writing. It calls these key findings, but does not establish that they were the most common barriers. (LLM, conf=0.83)
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 support only part of this claim — read each source’s verdict below.
In the paper
“Hence, the AI platforms can commit errors due to extracting details from unauthenticated sources”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Discussion
What the sources say
- Partly supportedcites [22] Chanda SS, Banerjee DN. Omission and commission errors underlying AI failures. AI Soc. 2022;17:1–24. sourceExperimental — accuracy not yet measured
The Framework section (page unavailable) says AI can take improper actions after learning questionable inferences from website and repository content. This supports errors arising from external content, but the paper does not identify those sources as unauthenticated or say that extracting details from them is the mechanism. (LLM, conf=0.80)
plainstates a cause- 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.
- Partly 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
“In all instances, the test items generated had focused case descriptions in the form of a clinical vignette, and horizontal integration across the pathophysiology of hypertension and pharmacology of antihypertensive drugs.”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > Integrated case cluster MCQs
plainn = 20- 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.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The clinical scenarios identified by the three AI platforms were comparable”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > OSPEs
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.
32 more claims
Create a free account to see all 112 claims and every source verdict.
- Probably not a substantial claim. 64
- It may state two separate claims as one. 23
- The quote only partly states this claim. 1
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > OSPEs
“The clinical scenarios identified by the three AI platforms were comparable”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > OSPEs
“For reasons unknown, Sage Poe did not provide any response to this search query.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > A-type MCQs
“| More than one correct option | 1g | 1ah | 1z | 2aq, ar | None | None |”
§ Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study > Results > A-type MCQs
“z- hypertension with angina and amlodipine (item #17)”
Citations
Cites
James, P. A., Oparil, S., Carter, B. L., Cushman, W. C., Dennison-Himmelfarb, C., Handler, J., Lackland, D. T., LeFevre, M. L., MacKenzie, T. D., Ogedegbe, O., et al. (2014) · JAMA · doi:10.1001/jama.2013.284427 · Cited by 7,950
Goodman & Gilman's The Pharmacological Basis of Therapeutics
Brunton, L. L. (1985) · Cited by 7,321
Artificial Intelligence in Education: The Three Paradigms
Ouyang, F., & Jiao, P. (2021) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2021.100020 · Cited by 991
Artificial Intelligence in Health Care: A Report From the National Academy of Medicine
Matheny, M. E., Whicher, D., & Thadaney Israni, S. (2020) · JAMA · doi:10.1001/jama.2019.21579 · Cited by 628
Boscardin, C. K., Gin, B., Golde, P. B., & Hauer, K. E. (2024) · Academic Medicine · doi:10.1097/acm.0000000000005439 · Cited by 620 · cited 1×
“We have used generative AIs in the present study that could generate new content from their training dataset [17].”
MethodEthical use of Artificial Intelligence in Health Professions Education: AMEE Guide No. 158
Masters, K. (2023) · Medical Teacher · doi:10.1080/0142159x.2023.2186203 · Cited by 302 · cited 1×
“While applications of AIs in undergraduate medical training are being explored, huge ethical challenges remain in terms of data collection, maintaining anonymity, consent, and ownership of the provided data [4].”
BackgroundArtificial intelligence for precision education in radiology
Duong, M. T., Rauschecker, A. M., Rudie, J. D., Chen, P.-H., Cook, T. S., Bryan, R. N., & Mohan, S. (2019) · The British Journal of Radiology · doi:10.1259/bjr.20190389 · Cited by 205 · cited 2×
“Problem-based and interactive learning approaches are referred to as “bottom-up” where learners obtain first-hand experience in solving the cases first and then indulge in discussion with the educators to refine their understanding and critical thinking skills [18].”
BackgroundMir, M. M. (2023) · PubMed · doi:10.30476/jamp.2023.98655.1803 · Cited by 165
Learning theories for artificial intelligence promoting learning processes
Gibson, D., Kovanovic, V., Ifenthaler, D., Dexter, S., & Feng, S. (2023) · British Journal of Educational Technology · doi:10.1111/bjet.13341 · Cited by 155 · cited 1×
“More recently a three-level model that synthesizes and unifies existing learning theories to model the roles of AIs in promoting learning process has been proposed [26].”
BackgroundMehta, N., Harish, V., Bilimoria, K., Morgado, F., Ginsburg, S., Law, M., & Das, S. (2021) · MedEdPublish · doi:10.15694/mep.2021.000075.1 · Cited by 145
Research works printed but not linked: 1
- Lee S. AI tools for educators. EIT InnoEnergy Master School Teachers Conference. 2023. https://www.slideshare.net/ignatia/ai-toolkit-for-educators?from_action=save (Accessed on 24th September 2023).
Cited by
Mapping the use of artificial intelligence in medical education: a scoping review
Rincón, E. H. H., Jimenez, D., Aguilar, L. A. C., Flórez, J. M. P., Tapia, Á. E. R., & Peñuela, C. L. J. (2025) · BMC Medical Education · doi:10.1186/s12909-025-07089-8 · Cited by 145
“A variety of pedagogical methods were proposed across the studies, including lectures, e-learning modules, collaborative learning platforms, and hands-on experiences with AI tools [46, 47, 50, 54–56].”
InfluentialCritical thinking in the age of generative AI: implications for health sciences education
Naqvi, W. M., Ganjoo, R., Rowe, M. D., Pashine, A. A., & Mishra, G. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2025.1571527 · Cited by 52
Using ChatGPT for medical education: the technical perspective
Chan, K. Y., Yuen, T. H., & Co, M. (2025) · BMC Medical Education · doi:10.1186/s12909-025-06785-9 · Cited by 34
“Notably, one area of interest is how AI can assist students in enhancing learning efficiency, such as through the analysis of course evaluation comments [2], its application in assessments [53], and its influence on curricula in both undergraduate and postgraduate studies [13, 49, 52, 54, 55].”
Bolgova, O., Ganguly, P. K., Ikram, M. F., & Mavrych, V. (2025) · Medical Education Online · doi:10.1080/10872981.2025.2550751 · Cited by 42
“SAQs, while pedagogically superior for certain types of assessment, are often under-utilized due to the practical limitations of manual grading [7].”
Jacobs, C., Johnson, H., Tan, N., Brownlie, K., Joiner, R., & Thompson, T. (2025) · JMIR Medical Education · doi:10.2196/70766 · Cited by 20
“These findings align with previous research suggesting that AI tools can effectively support medical education by providing immediate feedback and opportunities for repeated practice [32,33].”
Metadata
- Authors
- Sridharan, K., & Sequeira, R. P.
- Affiliations
- Arabian Gulf University (BH)
- Corresponding author
- Sridharan, Kannan
- ORCID
- Sridharan, Kannan
- Venue
- BMC Medical Education
- Published
- 2024-04-22
- Volume
- 24
- Issue
- 1
- Pages
- 431-431
- ISSN
- 1472-6920 (ISSN-L)
- PMC
- PMC11034110
- OpenAlex
- W4395002516
- Semantic Scholar
- e6ab89d86c2f3201a0f5371b13f9c002973f2aac
- Cited by
- 114
- References
- 20
- Citations by year
- 2026: 56 · 2025: 47 · 2024: 10
- FWCI
- 9.61
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- BMC Medical Education (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Education › Innovations in Medical Education
- Keywords
- generative AI, medical education, test item generation, student assessment, hypertension, antihypertensive drugs, standard setting, open-ended questions, OSPE, multiple-choice questions, AI literacy, ChatGPT
- MeSH terms
- Proof of Concept Study, Artificial Intelligence, Cross-Sectional Studies, Curriculum, Education, Medical, Education, Medical, Undergraduate, Educational Measurement, Humans, Hypertension, Students, Medical
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study. https://app.cochat.ai/papers/10.1186/s12909-024-05365-7
- The paper
- Sridharan, K., & Sequeira, R. P. (2024). Artificial intelligence and medical education: application in classroom instruction and student assessment using a pharmacology & therapeutics case study. BMC Medical Education. https://doi.org/10.1186/s12909-024-05365-7
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.