The DOI matches its Crossref record; 22 of the 23 references it cites were found in a registry or an index; 3 of the 4 sources its claims cite were checked for support.
Minimum information about clinical artificial intelligence modeling: the MI-CLAIM checklist.
Open accessAbstract
Here we present the MI-CLAIM checklist, a tool intended to improve transparent reporting of AI algorithms in medicine.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
3 Partly supported · 2 Own interpretation · 1 No basis given · accuracy not yet measured
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Furthermore, the same methods that are enabling new analyses of clinical data may also enable the re-identification of patients in sometimes unpredictable ways17.”
§ Minimum information about clinical artificial intelligence modeling: the MI-CLAIM checklist > General principles of the MI-CLAIM design
What the sources say
- Partly supportedcites [17] Schwarz, C. G. et al. N. Engl. J. Med. 381, 1684–1686 (2019). sourceExperimental — accuracy not yet measured
The paper reports that face-recognition software matched photographs to reconstructed cranial MRI faces for 70 of 84 participants, demonstrating a potential route to re-identification. It does not establish that the same methods enable new analyses of clinical data or describe the risk as unpredictable. (LLM, conf=0.82)
hedgedconditional- 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
“Newer methodologies have the danger of becoming more complex and less interpretable, even when sophisticated interpretation technique are used11.”
What the sources say
- Partly supportedcites [11] Lipton, Z. C. Queue 16, 31–57 (2018). sourceExperimental — accuracy not yet measured
The discussion of simulatability (§3.1.1, p. 4) links complexity to difficulty understanding a model, and the discussion of post-hoc interpretability (§3.2, p. 5) says such techniques often do not reveal precisely how a model works. But §4.1 (p. 7) cautions that deep models are not uniformly less interpretable than simpler ones and can have advantages for some post-hoc interpretations. The paper supports the stated risk, not a general trend that newer methodologies are less interpretable. (LLM, conf=0.84)
hedgedconditional- 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
“In the past, ‘minimum information’ guidelines have substantially improved the downstream utility, transparency, and interpretability of data deposited in repositories and reported in publications across many other research domains, including data on randomized control trials15, RNA (gene) expression, diagnostic accuracy15, observational studies16, and meta-analyses.”
What the sources say
- Partly supportedcites [15] Moher, D. et al. Br. Med. J. 340, c869 (2010). sourceExperimental — accuracy not yet measured
In the section on improving RCT reporting, the paper identifies CONSORT as a checklist of essential reporting items. Its closing discussion says CONSORT seems to have improved reporting in journals that adopted it and notes that other groups use its template for diagnostic tests. It does not establish improved utility of data deposited in repositories or demonstrate that diagnostic-accuracy reporting has improved. (LLM, conf=0.84)
- Not checked — no full text of the sourcecites [16] von Elm, E. et al. Ann. Intern. Med. 147, 573–577 (2007). source
plaindirect- 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
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Unclear documentation on training and test-cohort selection, development methodology, and how systems were validated has added to the confusion.”
plaindirect- 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.
- Probably not a substantial claim. 2
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Real-world circumstances may prevent the complete sharing of code in certain situations, such as for results published by commercial entities that view their code base as a proprietary trade secret.”
§ Minimum information about clinical artificial intelligence modeling: the MI-CLAIM checklist > The six parts of the MI-cLAIM > Part 6: reproducible pipeline
hedgedconditional- 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.
Probably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Furthermore, no single examination or interpretation methodology is perfect.”
§ Minimum information about clinical artificial intelligence modeling: the MI-CLAIM checklist > The six parts of the MI-cLAIM > Part 5: model examination
plain- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.
The paper relies on this without support.
Overfitting and information leakage are the strongest drivers of poor generalization out of sample and of misrepresentations of model efficacy.
“These two factors are the strongest drivers of poor generalization out of sample and misrepresentations of model efficacy.”
The paper relies on this without support.
Testing with an external cohort from an independent clinic or hospital system is the highest level of validation.
“Testing with an external cohort from an independent clinic or hospital system is the highest level of validation.”
The paper relies on this without support.
Cross-validation is not a replacement for a separate test cohort.
“Cross-validation is not a replacement for a separate test cohort and should not be described as such.”
The paper relies on this without support.
Machine-learning models in other fields have shown serious shortcomings when researchers have attempted to generalize across populations.
“as these models in other fields have shown serious shortcomings when researchers have attempted to generalize across populations.”
The paper relies on this without support.
Sharing raw clinical data is often not possible because of institutional patient-privacy policies.
“Sharing of raw clinical data is often neither possible, due to institutional patient-privacy policies, nor advisable without such safeguards in place.”
The paper relies on this without support.
Validation of exact results is generally less important than whether the results are validated in a new cohort of patients.
“In any case, validation of the exact results is generally of less interest than whether or not the results are validated in a new cohort of patients.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Code for the complete model building pipeline should be published in well-documented scripts with computer environment requirements when possible7,8,13,14,16,19,20,24,25,27,33,50,62,64,65, including statements about any restrictions to access or re-use.”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
The cited source says what the claim attributes to it.
In the citing paper
“To ensure the generalizability of the AIPM to the intended healthcare setting, developers are advised to extensively validate the model for representative data from that setting6–8,10,13–15,24,27,56,64,66,68,72,77,79,87,88 (see phase 1, Representativeness).”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
In the citing paper
“The subsequent evidence review underscored the limits of current knowledge and research about health care algorithms in the literature.”
From Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care (2023) · cites this paper as [11]
In the citing paper
“It has been widely recommended to report any measures taken to prevent overfitting3,7,8,13,15.”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
In the citing paper
“When possible, the literature recommends the comparison of current best practice (e.g., an existing prediction model or medical decision rule) to the AIPM performance7,8,14–16.”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
In the citing paper
“Performance analysis by population subgroups or specific problematic use cases is recommended to identify algorithmic bias6,7,23,24,26,42,61,72,79,91.”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
In the citing paper
“The literature recommends to assess generalizability on external data from a different time period, place, or healthcare setting3,7,8,10,16,17,27,56,68,72,79.”
From Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review (2022) · cites this paper as [7]
In the citing paper
“Medical AI models should not only have high performance, but should also be interpretable 28,29 .”
From Swarm learning for decentralized artificial intelligence in cancer histopathology (2022) · cites this paper as [29]
In the citing paper
“The MI-CLAIM (minimum information about clinical artificial intelligence modeling) checklist is available as Supplementary Table 7 (ref.29 ).”
From Swarm learning for decentralized artificial intelligence in cancer histopathology (2022) · cites this paper as [29]
Citations
Cites
von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., Vandenbroucke, J. P., & Initiative, F. T. S. (2007) · PLoS Medicine · doi:10.1371/journal.pmed.0040296 · Cited by 23,020
von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., Vandenbroucke, J. P., & Initiative, F. T. S. (2007) · Annals of Internal Medicine · doi:10.7326/0003-4819-147-8-200710160-00010 · Cited by 17,796
Moher, D., Hopewell, S., Schulz, K. F., Montori, V., Gotzsche, P. C., Devereaux, P. J., Elbourne, D., Egger, M., & Altman, D. G. (2010) · BMJ · doi:10.1136/bmj.c869 · Cited by 8,883
Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K., Madams, T., Cuadros, J., et al. (2016) · JAMA · doi:10.1001/jama.2016.17216 · Cited by 7,697
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137
Self-citationLipton, Z. C. (2018) · Queue · doi:10.1145/3236386.3241340 · Cited by 5,427
A guide to deep learning in healthcare
Esteva, A., Robicquet, A., Ramsundar, B., Kuleshov, V., DePristo, M., Chou, K., Cui, C., Corrado, G., Thrun, S., & Dean, J. (2019) · Nature Medicine · doi:10.1038/s41591-018-0316-z · Cited by 5,312
Scalable and accurate deep learning with electronic health records
Rajkomar, A., Oren, E., Chen, K., Dai, A. M., Hajaj, N., Hardt, M., Liu, P. J., Liu, X., Marcus, J., Sun, M., et al. (2018) · npj Digital Medicine · doi:10.1038/s41746-018-0029-1 · Cited by 2,648
Self-citationOpportunities and obstacles for deep learning in biology and medicine
Ching, T., Himmelstein, D. S., Beaulieu-Jones, B. K., Kalinin, A. A., Do, B. T., Way, G. P., Ferrero, E., Agapow, P.-M., Zietz, M., Hoffman, M. M., et al. (2018) · Journal of The Royal Society Interface · doi:10.1098/rsif.2017.0387 · Cited by 2,310
Self-citationScalable and accurate deep learning with electronic health records.
Rajkomar, A., Oren, E., Chen, K., Dai, A. M., Hajaj, N., Hardt, M., Liu, P. J., Liu, X., Marcus, J., Sun, M., et al. (2018) · eScholarship (California Digital Library) · Cited by 2,242
Self-citation
Research works printed but not linked: 3
- Subbaswamy, A. & Saria, S. Biostatistics 21, 345–352 (2020).
- Pan, J., McGuinness, K., Sayrol, E., O’Connor, N. & Giro-i-Nieto, X. arXiv https://ui.adsabs.harvard.edu/abs/2016arXiv160300845P (2016).
- Lundberg, S. & Lee, S.-I. arXiv https://ui.adsabs.harvard.edu/abs/2017arXiv170507874L (2017).
Cited by
Sorted by citations
Items 601–620 of 620
Vijayaraj, T. D., Karpagarajan, R., Subramaniam, S., Annamalai, D., & Kumari, S. (2026) · Biosensors · doi:10.3390/bios16100537
Topcu, D., Bayraktar, N., & Devrim, E. (2026) · Clinica chimica acta; international journal of clinical chemistry · doi:10.1016/j.cca.2026.121355
Grant, O., Clark, D., & Nguyen, S. (2023) · Journal of Health Informatics and Digital Systems · doi:10.68159/w681883871
Teti, F. S. L., Keke, K. J.-B., Yao, K. S., Dongo, K. Y. S., Fionko, Y. B., & Haïdara, A. (2026) · International Journal of Scientific Research and Modern Technology. · doi:10.38124/ijsrmt.v5i9.1664
Selecting the optimal study design!
Albabtain, M. A., AlRasheed, M. M., Arafat, A. A., Breitinger, H., Breitinger, U., Elsisi, G. H., Gaber, A., Gumaa, M., Hozien, M., Abdelghany, R. M., et al. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-45595-7.00015-7
Marunych, R. Y., Bartusik‐Aebisher, D., Dynarowicz, K., Smolak, B., & Aebisher, D. (2026) · International Journal of Molecular Sciences · doi:10.3390/ijms27198871
Huang, Y., Sun, Z., Chen, J., Yang, T., Xie, L., & Lin, X. (2026) · Frontiers in Medicine · doi:10.3389/fmed.2026.1955863 · Cited by 1
Nasir, M. U., Rehmat, U., Saudagar, A. K. J., et al. (2026) · Frontiers in Medicine · doi:10.3389/fmed.2026.1959427
“…(Figure 26) and demonstrate that the FFT-Nasir GA-BC pipeline correctly identifies the absence of signal on the chosen synthetic benchmark — a publication norm increasingly demanded by the medical-AI reproducibility literature (22, 23) but rarely satisfied by frameworks of this complexity.”
InfluentialCao, Z., Zheng, P., Qi, Z., Yao, Y., Xie, D., He, Y., Zong, S., Zhang, Z., & Shi, F. (2026) · EMBO Molecular Medicine · doi:10.1038/s44321-026-00527-0
AI Model Reporting Standards in Health Care: Systematic Review and Consolidation
Akhmad, E. S., Slob, D., Gomes, A. L., Mingels, S., Dekker, A. L. A. J., & van Soest, J. (2026) · JMIR AI · doi:10.2196/92329
“The general topics, such as methods (n=17 ) [17,18,22-27,31-34,36, 37, 39-41 39-41”
Dongo, K. Y. S., Teti, F. S. L., Keke, K. J.-B., Fionko, Y. B., & Haïdara, A. (2026) · Neuroscience & Medicine · doi:10.4236/nm.2026.174014
Ardila, C. M., Pineda‐Vélez, E., & Vivares‐Builes, A. M. (2026) · Medical Sciences · doi:10.3390/medsci14060677
Jin, Y.-Z., Li", Y., Shen, Y., Zhang, J., Wu, Z., Gao, Y., Su, D., Li, L., Si, Y., Su, Y., et al. (2026) · medRxiv · doi:10.64898/2026.10.14.26365588
Byrne, A. R., & Sarpong, N. O. (2026) · JBJS Reviews · doi:10.2106/jbjs.rvw.26.00182
Allahwerdi, I., Soomro, G. M., Krayem, S., & Chramcov, B. (2026) · Information · doi:10.3390/info17101045
Peng, D., Liu, W., & Zhang, B. (2026) · Frontiers in Oncology · doi:10.3389/fonc.2026.1917622
Shahid, S. (2025) · International journal of medical and health research. · doi:10.61424/ijmhr.v3i3.1107
Riegelnegg, M., Giess, D., Henschke, C., et al. (2026) · BMJ Digital Health & AI · doi:10.1136/bmjdh-2026-000097
Федорова, А. А., Gavrishchaka, V., Makeyev, A., Miao, R., Yang, Z.-Y., & Orlova, M. (2026) · Proceedings of the 9th International Conference on Machine Learning and Machine Intelligence · doi:10.1145/3841358.3841364
“…Prognosis Or Diagnosis, AI extension (TRIPOD+AI) [20], the Minimum Information about Clinical Artificial Intelligence Modeling checklist (MI-CLAIM) [21], the Strengthening the Reporting of Observational Studies in Epidemiology statement (STROBE) [22], and — where health-economic outcomes enter the…”
A scoping review of artificial intelligence-enabled wearables for medication adherence
Chen, Y., Zhao, H., Xu, R., Li, Y., Cai, S., & Hu, Y. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-03246-5
Metadata
- Authors
- Norgeot, B., Quer, G., Beaulieu-Jones, B. K., Torkamani, A., Dias, R., Gianfrancesco, M., Arnaout, R., Kohane, I. S., Saria, S., Topol, E., Obermeyer, Z., Yu, B., & Butte, A. J.
- Affiliations
- University of California, San Francisco (US)
- Scripps Research Institute (US)
- Johns Hopkins University (US)
- University of California, Berkeley (US)
- Corresponding author
- Butte, Atul J.
- ORCID
- Venue
- Nature Medicine
- Published
- 2020-09
- Volume
- 26
- Issue
- 9
- Pages
- 1320-1324
- ISSN
- 1078-8956 (ISSN-L)
- 1546-170X
- PMC
- PMC7538196
- OpenAlex
- W3083804794
- Semantic Scholar
- ac9bfb4ef6a1b3db3143ccf5394edc1171725234
- Cited by
- 665
- References
- 23
- Citations by year
- 2026: 165 · 2025: 112 · 2024: 128 · 2023: 107 · 2022: 89 · 2021: 58 · 2020: 6
- FWCI
- 19.24
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Nature Medicine (publishedVersion)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- Funders
- National Heart, Lung, and Blood Institute (K08HL125945, R01HL150394)
- National Institute of Arthritis and Musculoskeletal and Skin Diseases (K01AR075085)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
- Keywords
- clinical AI, reporting guidelines, algorithmic transparency, reporting quality
- MeSH terms
- Machine Learning, Arthritis, Rheumatoid, Artificial Intelligence, Diabetic Retinopathy, Humans, Drug-Related Side Effects and Adverse Reactions
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.
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
