Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI
Open accessThe DOI matches its Crossref record; 54 of the 56 references it cites were found in a registry or an index, and 2 are not checked yet; 6 of the 19 sources its claims cite were checked for support.
Abstract
A growing number of artificial intelligence (AI)-based clinical decision support systems are showing promising performance in preclinical, in silico, evaluation, but few have yet demonstrated real benefit to patient care. Early stage clinical evaluation is important to assess an AI system’s actual clinical performance at small scale, ensure its safety, evaluate the human factors surrounding its use, and pave the way to further large scale trials. However, the reporting of these early studies remains inadequate. The present statement provides a multistakeholder, consensus-based reporting guideline for the Developmental and Exploratory Clinical Investigations of DEcision support systems driven by Artificial Intelligence (DECIDE-AI). We conducted a two round, modified Delphi process to collect and analyse expert opinion on the reporting of early clinical evaluation of AI systems. Experts were recruited from 20 predefined stakeholder categories. The final composition and wording of the guideline was determined at a virtual consensus meeting. The checklist and the Explanation & Elaboration (E&E) sections were refined based on feedback from a qualitative evaluation process. 123 experts participated in the first round of Delphi, 138 in the second, 16 in the consensus meeting, and 16 in the qualitative evaluation. The DECIDE-AI reporting guideline comprises 17 AI specific reporting items (made of 28 subitems) and 10 generic reporting items, with an E&E paragraph provided for each. Through consultation and consensus with a range of stakeholders, we have developed a guideline comprising key items that should be reported in early stage clinical studies of AI-based decision support systems in healthcare. By providing an actionable checklist of minimal reporting items, the DECIDE-AI guideline will facilitate the appraisal of these studies and replicability of their findings.
Verified 2026-11-17 · 2 references not checked yet
Overview
CoChat Summary
TL;DR
DECIDE-AI is a consensus-based reporting guideline for early-stage, small-scale live clinical evaluation of AI-based decision support systems. It provides a 27-item minimum reporting checklist focused on clinical utility, safety, human factors, and preparation for larger summative trials.
Key findings
- The DECIDE-AI guideline comprises 17 AI-specific reporting items (28 subitems) and 10 generic reporting items, with an Explanation & Elaboration paragraph for each.
- It was developed through a consensus process involving 151 experts from 18 countries and 20 stakeholder groups.
- The guideline targets early live clinical evaluation, where supported decisions have an actual impact on patient care, and is intended to be independent of study design and AI system modality.
- The consensus group deliberately excluded items on interpretability and user trust because there is currently no generally accepted way to quantify or evaluate them.
Methods
- A two-round modified Delphi process was conducted, with expert recruitment across 20 predefined stakeholder categories; 123 experts completed round 1, 138 completed round 2, 16 participated in the consensus meeting, and 16 in qualitative evaluation.
- The initial item list contained 54 items; round 1 yielded 43,986 words of free text, 6,419 item scores, 228 comments, 64 new item proposals, and 109 themes, producing a revised list of 53 items. Round 2 yielded 7,101 item scores and 923 comments.
- A virtual consensus meeting used an 80% inclusion threshold and selected the final 17 AI-specific and 10 generic items; drafts were refined through qualitative evaluation by 16 independent experts.
Limitations
- The paper states that consensus processes carry potential biases, including anchoring and participant selection biases, which were mitigated but likely not eliminated.
- Representation was skewed toward Europe and the United Kingdom, partly due to selection bias in expert recommendations and higher interest from European/UK scientists.
- Few examples of early-stage clinical evaluation of AI systems were available when development began, which may have affected the exhaustiveness of the initial item list.
- The paper also notes that the guideline’s impact on reporting standards will need to be assessed once the community uses it, and that interpretability and trust were excluded due to lack of accepted measurement methods.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures

Verification
Claims the paper makes
4 Valid · 2 Not valid · 6 Not checked · 55 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“The DECIDE-AI guideline was developed through an international expert consensus process and in accordance with the EQUATOR Network's recommendations for guideline development.27”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > methods
What the sources say
- Supportedcites [27] source
Our copy is exactly the thing the citing sentence identifies: the Abstract states that 'David Moher and colleagues from the EQUATOR network offer guidance and recommended steps for developing health research reporting guidelines', and the Introduction (and Discussion) sets out the 18-step, five-phase strategy that constitutes those recommendations. The source also endorses the consensus element the sentence pairs with it — the Introduction states 'A consensus process [10] should be a crucial characteristic of developing a reporting guideline', with a Delphi exercise (step 5) and face-to-face consensus meeting (step 8) prescribed. No part of what is attributed to this source is unsupported; the specifics of DECIDE-AI's own process are the citing authors' statement, not this paper's. (LLM, conf=0.88)
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.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Finally, as also noted by the authors of the SPIRIT-AI and CONSORT-AI guidelines, 25 26 few examples of studies reporting on the early stage clinical evaluation of AI systems were available at the time we started developing the DECIDE-AI guideline.”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > User
What the sources say
- Supportedcites [25] source
The Discussion of our copy states directly that 'at the time of SPIRIT-AI development, there were only seven published trials and no published trial protocols in the field of AI for healthcare', so the decisions were 'not always supported by existing real-world examples'. The Literature review section quantifies this (316 registered trials on ClinicalTrials.gov as of 13 May 2019, only 62 completed and 7 with published results), and the Main text notes that 'the majority of published evidence has consisted of in silico, early-phase validation'. Together these support the attributed point that few studies reporting early-stage clinical evaluation of AI existed when the guideline was being developed; the reference to DECIDE-AI's own timing is the citing authors' framing, not a claim about this source. (LLM, conf=0.85)
- Supportedcites [26] source
The Discussion of CONSORT-AI states plainly that 'there are relatively few published interventional trials in the field of AI for healthcare; therefore, the discussions and decisions made during this study were not always supported by existing examples of completed trials', which is exactly the point the citing sentence attributes to the SPIRIT-AI/CONSORT-AI authors. The Main/introduction reinforces the 'early stage' element ('published evidence has consisted of in silico, early-phase validation'), and the Methods ('Literature review and candidate item generation') quantify it: of 316 registered AI-intervention trials found in a May 2019 ClinicalTrials.gov search, only 62 were completed and 7 had published results. The 'at the time we started developing the DECIDE-AI guideline' qualifier is the citing authors' own context, correctly paralleled with the source's own development period. (LLM, conf=0.85)
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.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“However, the reporting of these early studies remains inadequate.”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > Abstract
What the sources say
- Supportedcites [3] source
Our copy reports exactly this shortfall for the early clinical evaluations of ML-based AI systems it included: the Results (Risk of bias/funding paragraph) record that only 2 studies (5%) referenced a study protocol, 12 (32%) gave no or unclear funding information, 3 (8%) did not report clearly on test set independence, and none were low risk across all QUADAS-2 domains; the Discussion adds the absence of a protocol and of any power calculation as drivers of high risk of bias; and the Conclusions call for changes 'guided by accepted principles of trial conduct and reporting to avoid repetition of errors noted in the current literature.' The paper's own subject is early-stage (mostly non-randomized, median 4 clinicians) clinical studies of ML-based CDSSs, matching the claim's scope; the only caveat is that the source frames the deficit chiefly as risk of bias/study design, with reporting gaps as components of it. (LLM, conf=0.80)
- Partly supportedcites [22] sourceExperimental — accuracy not yet measured
The reporting-inadequacy part is squarely stated: the Abstract (Results) reports that 'adherence to reporting standards was suboptimal (<50% adherence for 12 of 29 TRIPOD items)' with full access to datasets/code unavailable in 95%/93% of studies, and Results > Non-randomised studies ('Adherence to reporting standards', median 62% adherence, flow diagram in 31%, hardware in 36%) plus Discussion (p. 1, 'improve the design, reporting, transparency') confirm poor reporting in these mostly retrospective, in silico studies (only 9/81 prospective, 6 in a real clinical setting). What the source does not support is the breadth of the paraphrase: it assessed only deep learning diagnostic medical imaging studies comparing AI with clinicians, and its Study limitations states 'we cannot definitively make this claim from our findings because we only assessed medical imaging studies' for other types of AI; the two published randomised trials it found actually had high adherence to reporting standards. (LLM, conf=0.75)
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.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“there is little high quality evidence for improved clinician performance or patient outcomes in clinical studies.3 4”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > Abstract
What the sources say
- Partly supportedcites [3] sourceExperimental — accuracy not yet measured
The Abstract Findings and Conclusions and the Discussion (p. unnumbered sections) state there was 'no robust evidence' that ML-based CDSS use was associated with improved clinician diagnostic performance, and that 'the evidence for any conclusion was weak because of a high risk of bias' — the Results report 28 of 37 studies (76%) at high risk of bias in at least one QUADAS-2 domain with none low across all four, which supports the 'little high quality evidence ... improved clinician performance' part. However, this review's outcomes were metrics of clinician performance only; no passage in our copy addresses evidence about patient outcomes, so that limb of the claim is not supported by this source (the citing sentence carries a second citation alongside it). (LLM, conf=0.80)
- Not checked yetcites [4] source
Checked before — its earlier verdict is held back until a new check runs.
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
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
“Most studies retrieved in the literature search described a comparator group (commonly the same group of clinicians without AI support).”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > User
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.
9 more claims
Create a free account to see all 67 claims and to verify the 11 references still open.
- Probably not a substantial claim. 45
- It may state two separate claims as one. 23
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 source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The same goes for the DECIDE-AI (Developmental and Exploratory Clinical Investigations of Decision Support Systems Driven by Artificial Intelligence) checklist, which is focused on reporting studies that involve the evaluation of AI systems during their implementation phase in the clinical setting [36].”
From Checklist Approach to Developing and Implementing AI in Clinical Settings: Instrument Development Study (2025) · cites this paper as [36]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Prospective evaluation provides a more accurate assessment by capturing workflow integration, clinician response, alert fatigue, and unintended consequences [44,45].”
From Artificial Intelligence in Clinical Decision-Making: A Comprehensive Review of Diagnostic, Prognostic, and Therapeutic Applications, Validation Gaps, Bias, and Deployment Challenges (2026) · cites this paper as [45]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Established through a multistakeholder Delphi process involving 123–138 experts, this 27-item checklist provides standards for early-stage clinical evaluation of AI-driven decision support systems, addressing key elements including interpretability, human-factor integration, and preliminary safety assessment.”
From Multimodal neuroimaging and AI integration in cognitive disorders: advances, challenges, and future directions for precision medicine (2026) · cites this paper as Vasey et al., 2022
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Reporting, evaluation, and appraisal frameworks including SPIRIT-AI, CONSORT-AI, DECIDE-AI, TRIPOD+AI, and PROBAST+AI have substantially strengthened the conceptual and reporting architecture for trustworthy clinical AI [10,11,12,13,14].”
From Beyond Model Development in Healthcare AI: Post-Development Robustness, Post-Deployment Monitoring, and Lifecycle Governance—A Scoping Review of Reviews (2026) · cites this paper as [12]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“international consensus initiatives have developed structured reporting frameworks such as the DECIDE-AI (Developmental and Exploratory Clinical Investigations of DEcision support systems driven by Artificial Intelligence) guideline (Vasey et al., 2022).”
From Multimodal neuroimaging and AI integration in cognitive disorders: advances, challenges, and future directions for precision medicine (2026) · cites this paper as Vasey et al., 2022
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
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
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > Reporting item checklist
“Depending on the chosen study design and if available, authors may also wish to complete the reporting according to study type specific guideline (eg, STROBE for cohort studies).47”
§ Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI > References
“47. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews.”
Citations
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283
Harris, P. A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., & Conde, J. G. (2009) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2008.08.010 · Cited by 51,384
The REDCap consortium: building an international community of software platform partners
Harris, P. A., Taylor, R., Minor, B. L., Elliott, V., Fernandez, M., O'Neal, L., McLeod, L., Delacqua, G., Delacqua, F., Kirby, J., et al. (2019) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2019.103208 · Cited by 24,013
S., N. L. (2017) · doi:10.1177/1609406917733847 · Cited by 14,051
International Organization for Standardization
(2011) · SpringerReference · doi:10.1007/springerreference_16847 · Cited by 12,534
Elm, E. V., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., & Vandenbroucke, J. P. (2007) · BMJ · doi:10.1136/bmj.39335.541782.ad · Cited by 10,766
An Experimental Application of the DELPHI Method to the Use of Experts
Dalkey, N., & Helmer, O. (1963) · Management Science · doi:10.1287/mnsc.9.3.458 · Cited by 6,061
Skivington, K., Matthews, L., Simpson, S. A., Craig, P., Baird, J., Blazeby, J. M., Boyd, K. A., Craig, N., French, D. P., McIntosh, E., et al. (2021) · BMJ · doi:10.1136/bmj.n2061 · Cited by 5,808
Lipton, Z. C. (2018) · Queue · doi:10.1145/3236386.3241340 · Cited by 5,427
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2 · Cited by 5,177
Research works printed but not linked: 1
- European Commission, Directorate-General for Communications Networks, Content and Technology. Ethics guidelines for trustworthy AI, Publications Office, 2019. https://data.europa.eu/doi/10.2759/346720 · doi:10.2759/346720
Cited by
Large language models in medicine
Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F., & Ting, D. S. W. (2023) · Nature Medicine · doi:10.1038/s41591-023-02448-8 · Cited by 3,944
Self-citationLekadir, K., Frangi, A. F., Porras, A. R., Glocker, B., Cintas, C., Langlotz, C. P., Weicken, E., Asselbergs, F. W., Prior, F., Collins, G. S., et al. (2025) · BMJ · doi:10.1136/bmj-2024-081554 · Cited by 957
“Compared with chat assistants and synthetic image generators that receive increased public interaction, healthcare is a more sensitive domain where errors can have major consequences.”
The TRIPOD-LLM reporting guideline for studies using large language models
Gallifant, J., Afshar, M., Ameen, S., Aphinyanaphongs, Y., Chen, S., Cacciamani, G., Demner-Fushman, D., Dligach, D., Daneshjou, R., Fernandes, C., et al. (2025) · Nature Medicine · doi:10.1038/s41591-024-03425-5 · Cited by 615
Ethical and Bias Considerations in Artificial Intelligence/Machine Learning
Hanna, M. G., Pantanowitz, L., Jackson, B., Palmer, O., Visweswaran, S., Pantanowitz, J., Deebajah, M., & Rashidi, H. H. (2025) · Modern Pathology · doi:10.1016/j.modpat.2024.100686 · Cited by 535
Metadata
- Authors
- Vasey, B., Nagendran, M., Campbell, B., Clifton, D. A., Collins, G. S., Denaxas, S., Denniston, A. K., Faes, L., Geerts, B., Ibrahim, M., Liu, X., Mateen, B. A., Mathur, P., McCradden, M. D., Morgan, L., Ordish, J., Rogers, C., Saria, S., Ting, D. S. W., Watkinson, P., Weber, W., Wheatstone, P., & McCulloch, P.
- Affiliations
- University of Oxford (GB)
- Imperial College London (GB)
- University of Exeter (GB)
- Centre for Statistics in Medicine, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UK
- UCL Biomedical Research Centre (GB)
- University Hospitals Birmingham NHS Foundation Trust (GB)
- Moorfields Eye Hospital NHS Foundation Trust (GB)
- Healthplus.ai-R&D, Amsterdam, Netherlands
- Maimonides Medical Center (US)
- The Alan Turing Institute (GB)
- Cleveland Clinic (US)
- University of Toronto (CA)
- Morgan Human Systems, Shrewsbury, UK
- Medicines and Healthcare products Regulatory Agency (GB)
- HeartFlow (United States) (US)
- Johns Hopkins University (US)
- National University of Singapore (SG)
- BMJ Group (United Kingdom) (GB)
- School of Medicine, University of Leeds, Leeds, UK
- Corresponding author
- Vasey, Baptiste
- ORCID
- Vasey, Baptiste
- Nagendran, Myura
- Campbell, Bruce
- Clifton, David A
- Denaxas, Spiros
- Denniston, Alastair K
- Faes, Livia
- Geerts, Bart
- Ibrahim, Mudathir
- Liu, Xiaoxuan
- Mateen, Bilal A
- Mathur, Piyush
- McCradden, Melissa D
- Morgan, Lauren
- Ordish, Johan
- Rogers, Campbell
- Saria, Suchi
- Ting, Daniel S W
- Watkinson, Peter
- Weber, Wim
- McCulloch, Peter
- Venue
- BMJ
- Publisher
- BMJ
- Published
- 2022-05-18
- Volume
- 377
- Pages
- e070904
- ISSN
- 1756-1833 (ISSN-L)
- PMC
- PMC9116198
- OpenAlex
- W4280488363
- Semantic Scholar
- 3a8c344f67d5081ead5f7dd5ebf0f760d69fc01d
- Cited by
- 665
- References
- 50
- Citations by year
- 2026: 418 · 2025: 99 · 2024: 73 · 2023: 62 · 2022: 13
- FWCI
- 18.60
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- BMJ (publishedVersion, cc-by-nc)
- PubMed Central (publishedVersion, cc-by-nc)
- PubMed (publishedVersion)
- Pure Amsterdam UMC (publishedVersion, other-oa)
- University of Birmingham Research Portal (University of Birmingham) (publishedVersion, cc-by-nc)
- White Rose Research Online (University of Leeds, The University of Sheffield, University of York) (acceptedVersion)
- Lirias (submittedVersion, cc-by-nc)
- Queen Mary Research Online (Queen Mary University of London) (submittedVersion, cc-by-nc)
- Data Archiving and Networked Services (DANS) (submittedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Open access LMU (Ludwid Maxmilian's Universitat Munchen) (submittedVersion, cc-by)
- Royal Devon and Exeter Research Repository (Royal Devon and Exeter) (submittedVersion, public-domain)
- Radboud Repository (Radboud University) (submittedVersion)
- Explore Bristol Research (submittedVersion, cc-by-nc)
- Spiral (Imperial College London) (submittedVersion, other-oa)
- OPUS (Augsburg University) (submittedVersion, cc-by-nc)
- Radboud Repository (Radboud University) (submittedVersion, other-oa)
- Funders
- National Science Foundation
- Alfred P. Sloan Foundation
- American Heart Association
- Pioneer Fund
- AstraZeneca
- GlaxoSmithKline
- Wellcome Trust (215010/Z/18/Z)
- Alan Turing Institute (N510129, TU/ASG/R-SPEH-105, TU/ASG/R-SPEH-113)
- UK Research and Innovation (P/S023283/1)
- Duke-NUS Medical School (05/FY2020/EX/15‐A58, Duke‐NUS/RSF/2021/0018, NUS/RSF/2021/0018)
- Cancer Research UK (27294, C49297/A27294, C49297)
- National Institute for Health and Care Research (NIHR302440, C49297/A27294, ACF-2009-14-007)
- Research Councils UK
- Academy of Medical Sciences (AMS-CSF4-Newcombe)
- University College London
- University of Oxford
- Agency for Science, Technology and Research (A20H4g2141, H20C6a0032)
- National Health Insurance Service
- Russian Science Foundation
- Sheffield Teaching Hospitals NHS Foundation Trust
- Medical Research Council (MR/K006584/1, MR/W01761X/1, C49297/A27294)
- Engineering and Physical Sciences Research Council (EP/P009824/1, N510129, EP/N020774/1)
- National Medical Research Council (MOH‐000655‐00, Duke-NUS/RSF/2021/0018, 05/FY2020/EX/15-A58)
- NIHR Oxford Biomedical Research Centre (C49297/ A27294)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Clinical practice guidelines implementation
- Keywords
- artificial intelligence, clinical decision support systems, reporting guidelines, AI safety, human factors, stakeholder consensus, reporting checklist
- MeSH terms
- Artificial Intelligence, Humans, Research Design, Decision Support Systems, Clinical, Research Report
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 2026-11-17, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-17 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash).
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: Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. https://app.cochat.ai/papers/10.1136/bmj-2022-070904
- The paper
- Vasey, B., Nagendran, M., Campbell, B., Clifton, D. A., Collins, G. S., Denaxas, S., Denniston, A. K., Faes, L., Geerts, B., Ibrahim, M., Liu, X., Mateen, B. A., Mathur, P., McCradden, M. D., Morgan, L., Ordish, J., Rogers, C., Saria, S., Ting, D. S. W., . . . McCulloch, P. (2022). Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. BMJ. https://doi.org/10.1136/bmj-2022-070904
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.