CoChat

Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI

Open access

The 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

Crossref · 10.1136/bmj-2022-070904

Cites

54 of 56 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig 1 Comparison of development pathways for drug therapies, artificial intelligence (AI) in healthcare, and surgical innovation. The coloured lines represent reporting guidelines, some of which are study design specific (TRIPOD-AI, STARD-AI, SPIRIT/CONSORT, SPIRIT/CONSORT-AI), others stage specific (DECIDE-AI, IDEAL). Depending on the context, more than one study design can be appropriate for each stage. *Only apply to AI in healthcare
Fig 1 Comparison of development pathways for drug therapies, artificial intelligence (AI) in healthcare, and surgical innovation. The coloured lines represent reporting guidelines, some of which are study design specific (TRIPOD-AI, STARD-AI, SPIRIT/CONSORT, SPIRIT/CONSORT-AI), others stage specific (DECIDE-AI, IDEAL). Depending on the context, more than one study design can be appropriate for each stage. *Only apply to AI in healthcareCC BY-NC · PubMed Central

Verification

Claims the paper makes

4 Valid · 2 Not valid · 6 Not checked · 55 Own results and interpretations

9 more claims

Create a free account to see all 67 claims and to verify the 11 references still open.

See all claims — free account
  • Probably not a substantial claim. 45
  • It may state two separate claims as one. 23

Create a free account to see these claims.

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?

2 more claims about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

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.

1 of 1

Citations

Cites

Show all 56 references

Research works printed but not linked: 1

  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

See all 923

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
Venue
BMJ
Publisher
BMJ
Published
2022-05-18
Volume
377
Pages
e070904
ISSN
1756-1833 (ISSN-L)
OpenAlex
W4280488363
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
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.