Colchicine in patients with acute ischaemic stroke or transient ischaemic attack (CHANCE-3): multicentre, double blind, randomised, placebo controlled trial.
Li J, Meng X, Shi FD, Jing J, Gu HQ, Jin A, Jiang Y, Li H, Johnston SC, Hankey GJ, Easton JD, Chang L, Shi P, Wang L, Zhuang X, Li H, Zang Y, Zhang J, Sun Z, Liu D, Li Y, Yang H, Zhao J, Yu W, Wang A, Pan Y, Lin J, Xie X, Jin WN, Li S, Niu S, Wang Y, Zhao X, Li Z, Liu L, Zheng H, Wang Y, CHANCE-3 Investigators
- DOI
- 10.1136/bmj-2023-079061
- Record issued
- 2026-08-16
- Engine
- 7.39.0
- Exported
- 2026-09-20
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/06f2d18f-ff42-43f7-87e6-58cc61a37831 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
- ReportingData & code availability partially met−0.25★
- No data or code availability links were detected to verify.
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted, rigorously reported randomized controlled trial with clear scientific premise, robust design, and transparent reporting. The main weakness is the vague data availability statement, which lacks a concrete access mechanism. Minor reporting gaps include lack of explicit CONSORT adherence and regulatory framework statement.
Both reviewers agreed on all dimensions; no divergence in study type or ratings. The statistics verification component checked 7 tests, all consistent, but coverage is limited to tests with test statistics/df or effect estimates with CIs; other tests were not machine-verified. The citation check found no retracted or non-existent references.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 7 tests: 7 consistent, 0 inconsistent; 5 recomputed directly from the reported test statistics, 2 via agent-written checks.
- CONSISTENTreported p = .790 · recomputed p = .813Recomputed hazard ratio 0.98 (95% CI 0.83–1.16), reported p=0.79
“hazard ratio 0.98 (95% confidence interval 0.83 to 1.16); P=0.79”
Taken as given: 0.83–1.16 is a two-sided 95% confidence interval for the hazard ratio of 0.98, not a range, an IQR, or a different interval level; the hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.79 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.98, 0.83, 1.16, 1) - CONSISTENTreported p = .970 · recomputed p = 1.000Recomputed subdistribution hazard ratio 1.00 (95% CI 0.84–1.19), reported p=0.97
“subdistribution hazard ratio 1.00 (95% CI 0.84 to 1.19); P=0.97”
Taken as given: 0.84–1.19 is a two-sided 95% confidence interval for the subdistribution hazard ratio of 1.00, not a range, an IQR, or a different interval level; the subdistribution hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.97 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(1, 0.84, 1.19, 1) - CONSISTENTreported p = .640 · recomputed p = .618Recomputed hazard ratio 0.96 (95% CI 0.82–1.13), reported p=0.64
“hazard ratio 0.96 (95% CI 0.82 to 1.13); P=0.64”
Taken as given: 0.82–1.13 is a two-sided 95% confidence interval for the hazard ratio of 0.96, not a range, an IQR, or a different interval level; the hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.64 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.96, 0.82, 1.13, 1) - CONSISTENTreported p = .860 · recomputed p = .889Recomputed odds ratio 0.99 (95% CI 0.86–1.14), reported p=0.86
“odds ratio 0.99 (95% CI 0.86 to 1.14); P=0.86”
Taken as given: 0.86–1.14 is a two-sided 95% confidence interval for the odds ratio of 0.99, not a range, an IQR, or a different interval level; the odds ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.86 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.99, 0.86, 1.14, 1) - CONSISTENTreported p = .740 · recomputed p = .716Recomputed odds ratio 1.03 (95% CI 0.88–1.21), reported p=0.74
“odds ratio 1.03 (95% CI 0.88 to 1.21), P=0.74”
Taken as given: 0.88–1.21 is a two-sided 95% confidence interval for the odds ratio of 1.03, not a range, an IQR, or a different interval level; the odds ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.74 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(1.03, 0.88, 1.21, 1) - CONSISTENTreported p = .790 · recomputed p = .813Reviewers 1, 2Primary outcome hazard ratio p-value
“hazard ratio 0.98 (95% CI 0.83 to 1.16); P=0.79”
Taken as given: The hazard ratio is 0.98 with 95% CI 0.83 to 1.16.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from hazard ratio and 95% CI using the pCI function with log=1.How we recomputed it: pCI(0.98, 0.83, 1.16, 1) - CONSISTENTreported p = .830 · recomputed p = .832Reviewers 1, 2Primary safety outcome p-value
“Any serious adverse event was observed in 91 (2.2%) patients in the colchicine group and 88 (2.1%) in the placebo group (P=0.83).”
Taken as given: The numbers 91 and 88 are the event counts in each group.; The group totals are 4176 and 4167.; The test used is chi-square (or Fisher's exact, but chi-square is a reasonable approximation).Method: Recomputed p-value using Pearson chi-square test for 2x2 table.How we recomputed it: pChi2x2(91, 4176-91, 88, 4167-88)
- lowinternal contradictionThe abstract states '8343 patients aged 40 years of age or older' while the results section states '8343 patients (57.7%) were enrolled'. The phrase 'aged 40 years of age or older' is redundant but not contradictory.
8343 patients aged 40 years of age or older were enrolled. ... 8343 patients (57.7%) were enrolled
Abstractreviewer’s wording
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
5 major claims checked against the paper's own evidence: 2 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewer 1The results differ from COLCOT, possibly due to differences in treatment duration and patient population.The paper offers plausible explanations but does not provide direct comparative evidence; it is a hypothesis.Evidence: Discussion compares CHANCE-3 with COLCOT, noting differences in treatment duration and pathophysiology.
“Our results differ from that of the COLCOT trial and various reasons might explain these differential findings.”
DiscussionFind in source - partialReviewer 1Subgroup analyses suggest potential treatment effect heterogeneity by age and symptomatic intracranial artery stenosis.The subgroup findings are exploratory and not adjusted for multiple comparisons, so they are hypothesis-generating.Evidence: Crude interaction P=0.02 for age and P=0.03 for symptomatic intracranial artery stenosis.
“treatment effects might be different between the two trial groups according to ages (crude interaction P=0.02 when age is dichotomised, and P=0.018 when age is treated as a continuous variable)”
ResultsFind in source - supportedReviewers 1, 2Low-dose colchicine did not reduce the risk of subsequent stroke within 90 days compared with placebo.The primary outcome result (HR 0.98, 95% CI 0.83-1.16, P=0.79) directly supports this claim.Evidence: Primary outcome: stroke occurred in 264 (6.3%) colchicine vs 270 (6.5%) placebo, HR 0.98 (0.83-1.16), P=0.79.
“Stroke occurred within 90 days in 264 patients (6.3%) in the colchicine group and 270 patients (6.5%) in the placebo group (hazard ratio 0.98 (95% confidence interval 0.83 to 1.16); P=0.79).”
AbstractFind in source - supportedReviewers 1, 2Colchicine did not increase the risk of serious adverse events compared with placebo.The primary safety outcome showed no significant difference (P=0.83), supporting the claim.Evidence: Any serious adverse event occurred in 91 (2.2%) colchicine vs 88 (2.1%) placebo, P=0.83.
“Any serious adverse event was observed in 91 (2.2%) patients in the colchicine group and 88 (2.1%) in the placebo group (P=0.83).”
AbstractFind in source - supportedReviewers 1, 2The study did not provide evidence that low-dose colchicine could reduce the risk of subsequent stroke within 90 days.The conclusion accurately reflects the null primary outcome.Evidence: Primary outcome HR 0.98 (0.83-1.16), P=0.79.
“The study did not provide evidence that low-dose colchicine could reduce the risk of subsequent stroke within 90 days as compared with placebo”
ConclusionFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary efficacy outcome is a hard clinical outcome: any new stroke within 90 days. No surrogate endpoint is used for the primary efficacy claim.
“The primary efficacy outcome was any new stroke within 90 days after randomisation.”
- ADEQUATEEffect sizeThe primary outcome shows no significant difference between groups (HR 0.98, 95% CI 0.83 to 1.16, P=0.79). The effect size is reported with confidence intervals and is interpreted as no evidence of benefit, which is appropriate for a null result.
“hazard ratio 0.98 (95% confidence interval 0.83 to 1.16); P=0.79”
Data authenticity concerns
1 finding · worst lowAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
- Other integrity concernAssessed
2 integrity concerns flagged (0 high).
- lowotherThe paper reports a subgroup interaction p-value of 0.02 for age, but this is not adjusted for multiple comparisons, which the authors acknowledge. This is not a validity threat but a caution.
“treatment effects might be different between the two trial groups according to ages (crude interaction P=0.02 when age is dichotomised, and P=0.018 when age is treated as a continuous variable)”
ResultsFind in source
Reporting gaps
1 finding · worst mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data/code availability incompleteAssessed
The paper cites prior work (COLCOT, meta-analyses) showing colchicine's benefit in coronary artery disease and the association of inflammation with stroke risk. It identifies the gap: whether colchicine can prevent early recurrent stroke in acute stroke/TIA patients. The rationale is logical, linking the anti-inflammatory effects of colchicine to the high-risk period. Limitations of prior work are implicitly addressed by focusing on the acute phase and a specific high-risk population.
“The Colchicine Cardiovascular Outcomes Trial (COLCOT) found that colchicine reduced the risk of cardiovascular events and stroke as one component of primary endpoints in patients with acute coronary artery disease”
“However, whether colchicine can prevent early subsequent stroke in patients with acute stroke or transient ischaemic attack is still unknown.”
“However, whether colchicine can prevent early subsequent stroke in patients with acute stroke or transient ischaemic attack is still unknown.”
“The Colchicine Cardiovascular Outcomes Trial (COLCOT) found that colchicine reduced the risk of cardiovascular events and stroke as one component of primary endpoints in patients with acute coronary artery disease”
“However, whether colchicine can prevent early subsequent stroke in patients with acute stroke or transient ischaemic attack is still unknown.”
“However, whether colchicine can prevent early subsequent stroke in patients with acute stroke or transient ischaemic attack is still unknown.”
Randomization method (central, block length of four) and unit (patient) are described. Blinding is comprehensive (patients, caregivers, outcome assessors). Power analysis is provided (90% power to detect 25% reduction). Inclusion/exclusion criteria are detailed. Outlier handling is addressed via ITT and per-protocol analyses. Controls are inherent (placebo). Independent replication is not applicable for a single pivotal trial.
“Eligible participants were randomly assigned in a 1:1 ratio with a block length of four to receive colchicine or placebo colchicine.”
“Patients, care givers, and those assessing outcomes were masked to allocation.”
“We determined that a total of 8238 patients would provide 90% statistical power to detect a 25% reduction in the risk of any new stroke, with a significance level (α) of 0.05”
“Eligible participants were randomly assigned in a 1:1 ratio with a block length of four to receive colchicine or placebo colchicine.”
“Patients, care givers, and those assessing outcomes were masked to allocation.”
“We determined that a total of 8238 patients would provide 90% statistical power to detect a 25% reduction in the risk of any new stroke, with a significance level (α) of 0.05”
Sex is reported (37.6% women overall; per-group in Table 1). Age is reported (median 66.3 years). Demographics include ethnicity, comorbidities, and other baseline characteristics. Since both sexes are enrolled, sex_justified is not applicable. Species/strain and housing are not applicable for a human trial.
“Male | 2632 (63.0) | 2578 (61.9)”
“Age, years | 66.3 (58.2-73.1) | 66.4 (58.3-73.2)”
“Han Chinese ethnic group† | 4056 (97.1) | 4047 (97.1)”
“The median age of the patients was 66.3 years, and 37.6% were women.”
“Hypertension | 3175 (76.0) | 3223 (77.3)”
The study was approved by the ethics committee at Beijing Tiantan Hospital (No. KY2022-093-02) and each participating site. Written informed consent was obtained from all patients or their representatives. Regulatory compliance is implied by adherence to ethical standards, though not explicitly named.
“The study was approved by the ethics committee at Beijing Tiantan Hospital (No. KY2022-093-02) and each participating site.”
“Written informed consent was provided by all patients or their representatives before enrolment.”
“The study was approved by the ethics committee at Beijing Tiantan Hospital (No. KY2022-093-02) and each participating site.”
“Written informed consent was provided by all patients or their representatives before enrolment.”
Colchicine is identified as the investigational drug, with manufacturer (Kunming Pharmaceuticals) and dose regimen specified. Statistical software (SAS 9.4) is identified. No other biological/chemical resources are used, so other criteria are not applicable.
“Patients randomly received colchicine at a dose of 0.5 mg twice daily or placebo on days 1-3, followed by 0.5 mg per day on days 4 to 90”
“Statistical analyses were done using SAS software, version 9.4 (SAS Institute).”
“Patients randomly received colchicine at a dose of 0.5 mg twice daily or placebo on days 1-3, followed by 0.5 mg per day on days 4 to 90”
“Statistical analyses were done using SAS software, version 9.4 (SAS Institute).”
Tests are named (shared frailty model, Fine-Gray, generalized linear mixed models, ordinal logistic regression, chi-square/Fisher). Assumptions are assessed (proportional hazards). Exact p-values are reported (e.g., P=0.79). Effect sizes with 95% CIs are provided. Software is identified. Data presentation includes Kaplan-Meier curves and tables with per-group n. Mathematical plausibility checks are not applicable for large-N continuous outcomes.
“We compared the difference in event risk between the two trial groups by a shared frailty model”
“hazard ratio 0.98 (95% CI 0.83 to 1.16); P=0.79”
“hazard ratio 0.98 (95% CI 0.83 to 1.16)”
“We compared the difference in event risk between the two trial groups by a shared frailty model”
“hazard ratio 0.98 (95% CI 0.83 to 1.16); P=0.79”
“hazard ratio 0.98 (95% CI 0.83 to 1.16)”
The data availability statement says data can be made available on reasonable request after signing data sharing agreements, but does not specify a platform, conditions, or timeframe. This is reported_but_inadequate. No repository deposit or accession numbers are provided, which is acceptable for patient-level data. No custom code is mentioned.
“Data collected for the study, including de-identified individual participant data and a data dictionary defining each field in the set, can be made available to other researchers on reasonable request and after signing appropriate data sharing agreements.”
“Data collected for the study, including de-identified individual participant data and a data dictionary defining each field in the set, can be made available to other researchers on reasonable request and after signing appropriate data sharing agreements.”
Trial registration is provided (NCT05439356). Methods are detailed. Limitations are explicitly discussed. Conclusions are proportional to the null result. Funding sources and conflicts of interest are declared. Reporting guideline adherence is not explicitly mentioned but the paper follows CONSORT-like structure.
“Trial registration ClinicalTrials.gov, NCT05439356”
“The CHANCE-3 trial has some limitations.”
“Funding: National Key Research and Development Program of China, the National Natural Science Foundation of China, the Capital's Funds for Health Improvement and Research, and Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences.”
“Trial registration ClinicalTrials.gov, NCT05439356”
“The CHANCE-3 trial has some limitations.”
“Funding: National Key Research and Development Program of China, the National Natural Science Foundation of China, the Capital's Funds for Health Improvement and Research, and Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences.”
Registered (1 ID: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None found · partly checkedReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Nothing surfaced — but not everything feeding this category ran (missing: data/code link verification), so read this as a partial clean bill.
Checked 46 references by DOI: 44 verified — 2 no DOI (shown, not verified).
- NO DOIA standardized method for measuring intracranial arterial stenosisNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIGuideline for clinical management of ischaemic cerebrovascular disorders (excerpt) [Chinese]No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
Copyediting
7 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 7 minor suggestions below.
7 copyedit issues flagged: mostly consistency, grammar, typo.
- MINORtypoAbstract, Results“8343 patients aged 40 years of age or older”→ Remove redundancy: 'aged 40 years or older'Redundant phrasing.
- MINORconsistencyTable 1, footnote“TIA=transient ischaemic stroke”→ Change to 'transient ischaemic attack'Incorrect expansion of TIA.
- MINORgrammarDiscussion, Comparison with other studies“The measurement of high sensitivity C-reactive protein is common in clinics and its concentration can usually be obtained rapidly.”→ Consider rephrasing for clarity: 'Measurement of high-sensitivity C-reactive protein is common in clinics, and results are usually available rapidly.'Awkward phrasing.
- MINORconsistencyResults, Patient“4176 randomly assigned to receive colchicine and 4167 receive placebo”→ Change 'receive' to 'received' for grammatical consistency.Verb tense inconsistency.
- MINORclarityMethods, Statistical analysis“Absolute event risks and subdistribution hazard ratios from the Fine-Gray mode”→ Change 'mode' to 'model'Typo.
- MINORconsistencyTable 1, footnote“Hs-CRP=high sensitivity C-reactive protein”→ Use consistent abbreviation (hs-CRP) throughout.Inconsistent capitalization of hs-CRP.
- MINORgrammarDiscussion, Comparison with other studies“The measurement of high sensitivity C-reactive protein is common in clinics and its concentration can usually be obtained rapidly.”→ Consider rephrasing for clarity.Minor grammatical flow.
The published work is robust and reliable; the null result is well-supported. An informed reader should weigh the minor reporting gaps (data availability vagueness, lack of explicit CONSORT statement) as transparency issues, but they do not undermine the validity of the findings. No erratum or re-analysis is warranted based on the evidence reviewed.
- 1.HIGHdata codeIn the Data availability statement, specify a concrete access mechanism (e.g., a named data access committee or platform like Vivli) and a timeframe for responding to requests.The current 'reasonable request' statement is vague and does not meet best practices for data sharing transparency.
- 2.HIGHreportingIn the Methods or Acknowledgments, explicitly state adherence to CONSORT reporting guidelines and mention that the CONSORT checklist is provided as supplementary material.Explicit reporting guideline adherence improves transparency and reproducibility.
- 3.MEDIUMethicsIn the Ethics section, explicitly state that the study was conducted in accordance with the Declaration of Helsinki or ICH-GCP guidelines.This strengthens the regulatory compliance statement, which is currently implied but not explicit.
- 4.MEDIUMdata codeConsider depositing de-identified aggregate data or statistical analysis code in a public repository (e.g., Figshare) with a DOI, if ethically permissible.This would enhance transparency and allow independent verification of the analyses.
- 5.MEDIUMreportingIn the Methods, clarify the role of the contract research organization in randomization code generation to ensure independence.This addresses a potential concern about the independence of the randomization process.
- 6.MEDIUMreportingIn the Methods, provide more detail on the blinding of the clinical event adjudication committee and data safety monitoring committee.This would fully describe the blinding procedures for all relevant parties.
- 7.LOWcopyeditIn the Abstract, Results, change 'aged 40 years of age or older' to 'aged 40 years or older'.Removes redundant phrasing.
- 8.LOWcopyeditIn Table 1 footnote, change 'TIA=transient ischaemic stroke' to 'TIA=transient ischaemic attack'.Corrects the expansion of the abbreviation.
- 9.LOWcopyeditIn Results, Patient, change '4167 receive placebo' to '4167 received placebo'.Fixes verb tense inconsistency.
- 10.LOWcopyeditIn Methods, Statistical analysis, change 'Fine-Gray mode' to 'Fine-Gray model'.Fixes a typo.
- 11.LOWcopyeditIn Table 1 footnote, standardize the abbreviation for high-sensitivity C-reactive protein to 'hs-CRP' throughout.Ensures consistent terminology.
- 12.LOWcopyeditIn Discussion, Comparison with other studies, rephrase 'The measurement of high sensitivity C-reactive protein is common in clinics and its concentration can usually be obtained rapidly.' for clarity.Improves readability.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.