Semaglutide versus placebo in individuals with poor weight loss after bariatric surgery: a double-blinded, randomized, placebo-controlled trial.
Stanley C, Mallik R, Hamid N, Jassil FC, Reuven V, Ruwona T, Dicken SJ, Qamar S, Norton B, Kingett H, Pucci A, Magee C, Boniface D, Batterham RL, Devalia K, Mok J, Elkalaawy M, Loy J, Jenkinson A, Markakis H, Adamo M, Brown A, Makaronidis J
- DOI
- 10.1038/s41591-026-04416-4
- Record issued
- 2026-08-15
- 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/0520c931-38e8-486d-b7e7-b11ea8212edb is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
- ClaimsOverstated claim−0.5★
- 01Conclusion reaches beyond the evidence
Semaglutide 2.4 mg may rival weight loss results from revisional metabolic surgery.
“Additionally, we provide important clinical evidence that 68 weeks of semaglutide 2.4 mg weekly may rival weight loss results from revisional metabolic surgery when undertaken for weight optimization after a suboptimal outcome to a primary surgical procedure”
Discussion ¶5Find in source
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 and transparently reported randomized controlled trial with rigorous design, clear ethical approvals, and comprehensive reporting. Minor internal inconsistencies in adverse event counts and a slightly overstated claim about rivaling revisional surgery are the main weaknesses.
Both reviewers classified the study as interventional, and this was adopted. The evaluation covered the full text, with verification of statistics (7 tests recomputed consistently), citations (no retracted or not-found references), and reproducibility (GitHub link live). The GRIM/GRIMMER checks were not applicable due to continuous outcomes.
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; 2 recomputed directly from the reported test statistics, 5 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = .001Recomputed RR 3.0 (95% CI 1.7–6.3), reported p<0.001
“RR 3.0 (95% CI 1.7 to 6.3); P < 0.001”
Taken as given: 1.7–6.3 is a two-sided 95% confidence interval for the RR of 3.0, not a range, an IQR, or a different interval level; the RR is a RATIO measure, so the interval is symmetric on the log scale; p<0.001 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(3, 1.7, 6.3, 1) - CONSISTENTreported p = .009 · recomputed p = .013Recomputed RR 3.8 (95% CI 1.5–12.2), reported p=0.009
“RR 3.8 (95% CI 1.5 to 12.2); P = 0.009”
Taken as given: 1.5–12.2 is a two-sided 95% confidence interval for the RR of 3.8, not a range, an IQR, or a different interval level; the RR is a RATIO measure, so the interval is symmetric on the log scale; p=0.009 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(3.8, 1.5, 12.2, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Primary outcome treatment difference p-value
“The mean adjusted treatment difference in percentage body weight change for semaglutide 2.4 mg versus placebo was −19.18 (95% confidence interval −23.4 to −14.8; P < 0.001).”
Taken as given: The estimate is the adjusted mean difference.; The CI is a 95% confidence interval.; The test is two-sided.Method: Recomputed p-value from the estimate and 95% CI using a normal approximation.How we recomputed it: pCI(-19.18, -23.4, -14.8, 0) - CONSISTENTreported p = .903 · recomputed p = .569Reviewers 1, 2RR for at least one AE
“The risk ratio (RR) for the proportion of participants experiencing at least one AE between the semaglutide and placebo arms was 1.03 (95% CI 0.93 to 1.14; reference = placebo; P = 0.903).”
Taken as given: The RR is the ratio of two proportions.; The CI is a 95% confidence interval.; The test is two-sided.Method: Recomputed p-value from the RR and 95% CI using a normal approximation on the log scale.How we recomputed it: pCI(1.03, 0.93, 1.14, 1) - CONSISTENTreported p < .001 · recomputed p = .001Reviewers 1, 2RR for nausea
“There was a statistically significant greater frequency of reporting nausea (RR 3.0 (95% CI 1.7 to 6.3); P < 0.001)”
Taken as given: The RR is the ratio of two proportions.; The CI is a 95% confidence interval.; The test is two-sided.Method: Recomputed p-value from the RR and 95% CI using a normal approximation on the log scale.How we recomputed it: pCI(3.0, 1.7, 6.3, 1) - CONSISTENTreported p = .009 · recomputed p = .013Reviewer 2RR for decreased appetite
“There was a statistically significant greater frequency of reporting nausea (RR 3.0 (95% CI 1.7 to 6.3); P < 0.001) and decreased appetite (RR 3.8 (95% CI 1.5 to 12.2); P = 0.009) in the semaglutide group compared with placebo.”
Taken as given: The RR is the ratio of proportions.; The 95% CI is two-sided.; The p-value is two-sided.Method: Recomputed p-value from RR and 95% CI using log-normal approximation.How we recomputed it: pCI(3.8, 1.5, 12.2, 1) - CONSISTENTreported p = .253 · recomputed p = .284Reviewer 2RR for at least one SAE
“The RR for the proportion of participants experiencing at least one SAE between semaglutide and placebo arms was 0.40 (95% CI 0.06 to 1.72; reference = placebo; P = 0.253).”
Taken as given: The RR is the ratio of proportions.; The 95% CI is two-sided.; The p-value is two-sided.Method: Recomputed p-value from RR and 95% CI using log-normal approximation.How we recomputed it: pCI(0.40, 0.06, 1.72, 1)
- lowinternal contradictionThe abstract states 'There were eight serious AEs' but the results section and Table 4 report 12 SAEs (11 SAEs and one SUSAR). This discrepancy may be due to different counting methods (e.g., events vs. patients) but is not explained.
There were eight serious AEs, one suspected unexpected serious adverse reaction and no treatment-related deaths. ... There were 12 serious AEs (SAEs) (11 SAEs and one SUSAR) in the trial (Table ).
Table 4reviewer’s wording - lowinternal contradictionThe abstract reports a treatment difference of −19.18% while the results section reports −19.1%; this is a minor rounding inconsistency.
Abstract: '−19.18 (95% confidence interval −23.4 to −14.8; P < 0.001)' vs Results: '−19.1% (95% CI −23.4 to −14.8)'
Abstractreviewer’s wording
Overstated conclusions
2 findings · worst mediumConclusions 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 overstated beyond the evidenceAssessed
- Conclusions only partially backed by the presented evidenceAssessed
6 major claims checked against the paper's own evidence: 1 not fully backed by the presented evidence (unsupported or overstated), 1 only partially supported (evidence backs part of the claim; gaps or caveats remain).
- overstatedReviewer 2Semaglutide 2.4 mg may rival weight loss results from revisional metabolic surgery.The claim is speculative as there are no head-to-head comparisons; the authors themselves note variability in revisional surgery outcomes, so this is an overstatement.Evidence: Discussion states 'may rival' but acknowledges no head-to-head studies.
“Additionally, we provide important clinical evidence that 68 weeks of semaglutide 2.4 mg weekly may rival weight loss results from revisional metabolic surgery when undertaken for weight optimization after a suboptimal outcome to a primary surgical procedure”
Discussion ¶5Find in source - partialReviewers 1, 2The weight loss response to semaglutide in BARI-STEP appears greater than in other semaglutide trials.The comparison with STEP 1 is indirect and the authors acknowledge potential confounders (baseline BMI, sex), so the claim is cautiously supported but not definitive.Evidence: Comparison of %WL and categorical weight loss with STEP 1.
“Interestingly, although no conclusions can be drawn, the weight loss response to semaglutide 2.4 mg in the BARI-STEP trial appears to be greater than that reported in other semaglutide trials.”
DiscussionFind in source - supportedReviewers 1, 2Semaglutide 2.4 mg results in substantial and clinically significant body weight reduction compared to placebo in patients with suboptimal response after bariatric surgery.The primary outcome shows a large, statistically significant difference in %WL, supported by the adjusted treatment difference and CI.Evidence: Primary outcome: adjusted mean difference −19.1% (95% CI −23.4 to −14.8; P < 0.001).
“BARI-STEP demonstrates that in people with a suboptimal clinical response after MBS, semaglutide results in substantial and clinically significant body weight reduction along with improvement in metabolic parameters and quality of life, compared to placebo.”
AbstractFind in source - supportedReviewers 1, 2Semaglutide 2.4 mg is safe and well tolerated in this population.Safety data show AEs consistent with known profile, no new safety concerns, and no treatment-related deaths.Evidence: AE table and SAE table; no treatment-related deaths.
“Adverse events (AEs) were consistent with the known safety and tolerability profile of semaglutide, with no new safety concerns for the post-bariatric population.”
DiscussionFind in source - supportedReviewer 1Semaglutide 2.4 mg improves metabolic parameters and quality of life.Secondary outcomes show significant improvements in HbA1c, lipids, and IWQOL-Lite scores, though some are exploratory.Evidence: Table 2 shows significant improvements in HbA1c, cholesterol, triglycerides, and IWQOL-Lite total score.
“Semaglutide 2.4 mg led to significant reductions in HbA1c both in patients with ( P = 0.011) and without ( P < 0.001) T2D.”
ResultsFind in source - supportedReviewer 2Semaglutide 2.4 mg improves metabolic parameters and quality of life compared to placebo.Secondary outcomes show significant improvements in HbA1c, lipids, and IWQOL-Lite scores, supporting the claim.Evidence: Table 2: adjusted differences for HbA1c −5.3 mmol/mol (P<0.001), total cholesterol −0.5 mmol/L (P=0.014), IWQOL-Lite total 14.3 (P<0.001).
“BARI-STEP demonstrates that in people with a suboptimal clinical response after MBS, semaglutide results in substantial and clinically significant body weight reduction along with improvement in metabolic parameters and quality of life, compared to placebo.”
AbstractFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoint is percentage weight loss, which is a direct measure of the intervention's effect on body weight, not a surrogate. The paper also reports improvements in metabolic parameters and quality of life, but the primary claim is based on weight loss itself.
“The primary outcome was percentage weight loss after 68-week treatment on an intention-to-treat analysis.”
- ADEQUATEEffect sizeThe treatment difference in percentage weight loss is -19.1% (95% CI -23.4 to -14.8), which is large and clinically meaningful. The paper explicitly states that the weight loss is 'substantial and clinically significant' and cites that >15% weight loss is clinically relevant.
“BARI-STEP demonstrates that in people with a suboptimal clinical response after MBS, semaglutide results in substantial and clinically significant body weight reduction”
Data authenticity concerns
None foundAn 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.
Checked — nothing surfaced.
Reporting gaps
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites prior studies (GRAVITAS, BARI-OPTIMISE, STEP program) and explains the rationale for testing semaglutide in this population. It acknowledges gaps, such as the lack of a randomized trial for semaglutide in post-bariatric patients with suboptimal response. Limitations of prior work are addressed by designing a trial with a longer duration and a more potent GLP-1RA.
“The GRAVITAS trial demonstrated that, in individuals with persistent or recurrent T2D after MBS, liraglutide 1.8 mg daily as an adjunct to lifestyle intervention, improved glycemic control with a mean weight difference of −4.2 kg (95% confidence interval (CI) −6.8 to −1.4) after 26 weeks.”
“Considering the promising results from the BARI-OPTIMISE trial and the STEP program, we hypothesized that once-weekly semaglutide 2.4 mg will lead to greater weight loss and health improvement compared to placebo in patients with a suboptimal clinical response after MBS.”
“To our knowledge, there is no randomized clinical trial evaluating the safety and efficacy of semaglutide 2.4 mg for a suboptimal clinical response after MBS.”
“we hypothesized that once-weekly semaglutide 2.4 mg will lead to greater weight loss and health improvement compared to placebo in patients with a suboptimal clinical response after MBS.”
“BARI-OPTIMISE demonstrated comparatively higher %WL to the SCALE trial; however, given the trial design, a relationship between the lower native GLP-1 levels in the trial population and the higher %WL could not be inferred.”
Randomization was computer-generated and stratified by surgery type, sex, and diabetes status. Blinding of participants and clinical personnel is described. A sample size calculation is provided with assumptions. Inclusion/exclusion criteria are detailed. The analysis population (ITT) is defined, and missing data handled via inverse probability weighting. The trial is registered and follows CONSORT.
“Randomization was carried out by a computer-generated randomization sequence (sealed envelope) stratified according to surgical procedure type (GB versus SG), sex and T2D status.”
“Written informed consent was obtained and all participants and clinical study personnel were blinded.”
“Assuming an s.d. for %WL of 4.0, a dropout rate of 10% and 1% critical significance level, 62 patients (31 per group) would provide at least 95% power to detect an estimated difference of 10%WL with a 95% CI no wider than 8.0 to 12.0.”
“Randomization was carried out by a computer-generated randomization sequence (sealed envelope) stratified according to surgical procedure type (GB versus SG), sex and T2D status.”
“Written informed consent was obtained and all participants and clinical study personnel were blinded.”
“Assuming an s.d. for %WL of 4.0, a dropout rate of 10% and 1% critical significance level, 62 patients (31 per group) would provide at least 95% power to detect an estimated difference of 10%WL with a 95% CI no wider than 8.0 to 12.0.”
Sex is reported for all participants, and both sexes are included, so no justification for single-sex is needed. Age, weight, BMI, and health status (diabetes) are reported. Demographics include ethnicity. Species/strain and housing are not applicable as this is a human trial.
“58 (82.9%) female and 12 (17.1%) male”
“Age (years) | 47.7 (10.3) | 46.9 (10.4) | 47.3 (10.3)”
“Seventy participants (mean (s.d.) age = 47.3 (10.3) years, 58 (82.9%) female and 12 (17.1%) male) were randomized”
The trial was approved by a named ethics committee (London Surrey Borders Research Ethics Committee) with a protocol number. Informed consent is mentioned. Compliance with the Declaration of Helsinki and GCP is stated. The trial is registered.
“The trial was approved by the London Surrey Borders Research Ethics Committee (no. 22/LO/0045)”
“Written informed consent was obtained”
“conducted in accordance with the Declaration of Helsinki, the principles of good clinical practice and all applicable regulatory requirements”
“The trial was approved by the London Surrey Borders Research Ethics Committee (no. 22/LO/0045)”
“Written informed consent was obtained”
“conducted in accordance with the Declaration of Helsinki, the principles of good clinical practice and all applicable regulatory requirements”
Semaglutide is identified as 'semaglutide 2.4 mg weekly (Novo Nordisk)' and placebo as 'saline solution'. The dose and regimen are described. Statistical software R version is identified. No antibodies, cell lines, or organisms are used, so those are not applicable.
“semaglutide 2.4 mg weekly (Novo Nordisk)”
“Merck Active GLP-1 ELISA Kit”
“Analyses were conducted in R (v.2024.12.1+563).”
“semaglutide 2.4 mg weekly (Novo Nordisk) ( n = 35) or placebo (saline solution) ( n = 35)”
“Analyses were conducted in R (v.2024.12.1+563).”
Tests are named (linear regression, chi-squared, Fisher's exact, log-binomial regression). Assumptions are checked visually. Exact p-values are reported (e.g., P < 0.001). Effect sizes with 95% CIs are provided. Software is identified. Data presentation includes individual data points (waterfall plot) and error bars defined. Mathematical plausibility checks were not possible for all values, but no obvious errors found.
“The primary outcome was analyzed using a linear regression model, adjusted for baseline weight, type of surgery, sex and diabetes status, according to the ITT principle.”
“The mean adjusted treatment difference for %WL in placebo versus semaglutide at 68 weeks was −19.1% (95% CI −23.4 to −14.8) ( P < 0.001)”
“The normality assumptions of models were visually checked where appropriate; no transformations or nonparametric methods were used.”
“The primary outcome was analyzed using a linear regression model, adjusted for baseline weight, type of surgery, sex and diabetes status, according to the ITT principle.”
“The mean adjusted treatment difference in percentage body weight change for semaglutide 2.4 mg versus placebo was −19.18 (95% confidence interval −23.4 to −14.8; P < 0.001).”
“The normality assumptions of models were visually checked where appropriate; no transformations or nonparametric methods were used.”
The data availability statement describes a managed access process with a timeframe. Code is available on GitHub. Repository deposit and accession numbers are not applicable for patient-level data.
“Individual participant data collected during the trial, after deidentification, the study protocol, statistical analysis plan and the patient information sheet (consent form) can be shared on request. Study data can be requested after publication through the corresponding author. An initial response to the request will be provided within 2 weeks; if additional approvals are required, the final response time may take up to 6 months.”
“Code for the analysis is publicly available and available without restriction via GitHub at https://github.com/SamuelJDicken/BARI-STEP”
“Individual participant data collected during the trial, after deidentification, the study protocol, statistical analysis plan and the patient information sheet (consent form) can be shared on request. Study data can be requested after publication through the corresponding author. An initial response to the request will be provided within 2 weeks; if additional approvals are required, the final response time may take up to 6 months.”
“Code for the analysis is publicly available and available without restriction via GitHub at https://github.com/SamuelJDicken/BARI-STEP”
The trial is registered (NCT05073835). CONSORT guideline is followed. All pre-specified outcomes are reported, including non-significant ones. Limitations are thoroughly discussed. Conclusions are proportional. Funding sources and competing interests are declared.
“ClinicalTrials.gov: NCT05073835”
“The study followed the CONSORT reporting guideline.”
“Limitations of our study include only recruiting patients after primary MBS surgery (with a small sample size of GB patients).”
“The trial was registered at ClinicalTrials.gov (registration: NCT05073835 (https://clinicaltrials.gov/ct2/show/NCT05073835) )”
“The study followed the CONSORT reporting guideline.”
“Limitations of our study include only recruiting patients after primary MBS surgery (with a small sample size of GB patients).”
Registered (2 IDs: ClinicalTrials.gov, EudraCT). Reporting guideline cited: CONSORT.
Broken references and links
None foundReferences 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.
Checked — nothing surfaced.
Checked 41 references by DOI: 40 verified — 1 no DOI (shown, not verified).
- NO DOIObesity: identification, assessment and management. Clinical Guideline CG189No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 data/code link checked; 1 live.
- codeGitHubLIVEHTTP 200https://github.com/SamuelJDicken/BARI-STEPResolves to GitHub (code repository).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly consistency, typo.
- MINORconsistencyAbstract“Eight serious AEs”→ Eight serious AEs (SAEs)Inconsistent use of 'serious AEs' vs 'SAEs' elsewhere.
- MINORtypoDiscussion“IQWOL-Lite”→ IWQOL-LiteTypo in the acronym.
- MINORconsistencyResults, Categorical weight loss“93.1% ( n = 27) of participants in the placebo group lost less than 10% weight loss”→ 93.1% (n = 27) of participants in the placebo group lost less than 10% of their body weightClarify wording.
- MINORtypoAbstract“−19.18 (95% confidence interval −23.4 to −14.8; P < 0.001)”→ Consider using consistent decimal places: −19.2 (95% CI −23.4 to −14.8; P < 0.001)Minor inconsistency in decimal places.
- MINORconsistencyResults, Primary outcome“−19.1% (95% CI −23.4 to −14.8)”→ Ensure consistency with abstract value of −19.18%.Minor numerical inconsistency between abstract and results.
The published work is robust and generally reliable, but readers should weigh the minor internal inconsistency in serious adverse event counts (abstract vs. results) and the slightly overstated claim about rivaling revisional surgery. These do not invalidate the primary findings but warrant clarification or a correction.
- 1.HIGHreportingReconcile the serious adverse event count: the abstract states 'eight serious AEs' while the results/Table 4 report 12 SAEs (11 SAEs and one SUSAR). Clarify whether this is events vs. patients or a counting error, and correct the abstract or results accordingly.An internal contradiction in a safety outcome is a validity concern that readers and regulators will flag.
- 2.HIGHrigorTemper the claim that semaglutide 2.4 mg 'may rival weight loss results from revisional metabolic surgery' in the Discussion, since no head-to-head comparison exists and the authors themselves note variability in revisional surgery outcomes.Overstating a conclusion beyond the evidence is a common reviewer objection and could mislead readers.
- 3.MEDIUMcopyeditFix the typo 'IQWOL-Lite' to 'IWQOL-Lite' in the Discussion.Correct terminology is essential for clarity and professionalism.
- 4.MEDIUMcopyeditStandardize the primary outcome value: use −19.2% (or −19.18%) consistently in the Abstract and Results, and ensure the decimal places match.Minor numerical inconsistency between abstract and results can confuse readers.
- 5.MEDIUMcopyeditClarify the wording in Results, Categorical weight loss: change 'lost less than 10% weight loss' to 'lost less than 10% of their body weight'.The current phrasing is grammatically awkward and could be misinterpreted.
- 6.MEDIUMcopyeditIn the Abstract, expand 'Eight serious AEs' to 'Eight serious adverse events (SAEs)' for consistency with the rest of the paper.Consistent use of terminology improves readability and avoids ambiguity.
- 7.LOWdata codeConsider providing a more detailed data-sharing mechanism, such as a named repository or data access committee, in the Data Availability statement.A more concrete access route enhances transparency and reproducibility.
- 8.LOWstatisticsReport exact p-values for all secondary outcomes, even exploratory ones, instead of threshold-only values where applicable.Exact p-values improve transparency and allow readers to assess evidence strength.
- 9.LOWreportingAdd a statement on whether any analyses were pre-registered beyond the trial registration, and provide a link to the full statistical analysis plan in the Data Availability statement.This strengthens transparency and helps readers verify that analyses were pre-specified.
- 10.LOWstatisticsReport the results of normality checks in more detail, including any violations and how they were handled, and provide model diagnostics for secondary outcomes.Detailed assumption checks and diagnostics increase confidence in the statistical methods.
- 11.LOWstatisticsAdd a sensitivity analysis excluding participants with protocol deviations to assess robustness of the primary findings.Sensitivity analyses strengthen the credibility of the results.
- 12.LOWreportingReport the number of participants who discontinued due to adverse events separately for each arm, and consider a CONSORT-style flow diagram for secondary outcomes.Detailed safety reporting and flow diagrams improve completeness and transparency.
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.