Oral ivermectin versus 5% permethrin cream to treat children and adults with classic scabies: multicentre, assessor blinded, cluster randomised clinical trial.
Boralevi F, Simon G, Bernigaud C, Brun J, Goujon E, Perrot JL, Phan A, Hérissé AL, Maruani A, Vanhaecke C, Couty E, Abasq-Thomas C, Fournet M, Mallet S, Brenaut E, Hubiche T, Balguerie X, Caux F, Beneton N, Dutkiewicz AS, Adamski H, Aubert H, Bourrat E, Couzan C, Eyraud A, Janela-Lapert R, Marti A, Miquel J, Richard M, Barbarot S, Bonniaud B, Chabbert C, Delion F, Dinulescu M, Duong TA, Etienne M, Hirsch G, Isnard C, Huet F, Le Bidre E, Leducq S, Legrand A, Monestier S, Morice-Picard F, Seneschal J, Capelli A, Lacoste C, Labrèze C, Miraglia G, Do-Pham G, Giraudeau B, Chosidow O
- DOI
- 10.1136/bmj-2025-086277
- Record issued
- 2026-08-10
- Engine
- 7.29.0
- Exported
- 2026-09-21
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/5c65ed73-116f-4a77-a8a3-b4183cf8dd60 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
- ClaimsUnsupported claim (uncorroborated)−0.5★
- ReportingEthical approvals partially met−0.25★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run on this paper: the pass that reads its reported means did not complete. No reported mean was checked for arithmetic impossibility.
- 01Conclusion not supported by the paper’s own evidence
In other situations, first line oral ivermectin could be prescribed, especially when topical application is not practical.
“In other situations, first line oral ivermectin could be prescribed, especially when topical application is not practical, for example in older adults, especially residents with dementia during a scabies outbreak in a nursing home, or in people experiencing…”
DiscussionFind 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 cluster-randomised, assessor-blinded, active-comparator trial is rigorously designed and, with two small caveats, transparently reported: the premise is well-evidenced, the negative primary result and limitations are disclosed, and data and code are openly deposited (verified live). The weaknesses are reporting gaps rather than scientific defects: a missing explicit Helsinki/ICH-GCP statement, a sample-size calculation that (as disclosed) ignored clustering, one unsupported Discussion extrapolation about first-line ivermectin in untested settings, and minor Table 2 denominator inconsistencies plus two copyedit slips.
Both evaluations are independent samples of the same model applied to the full text; they agreed on 7/8 dimension statuses and diverged only on ethical approvals (R1 pass vs R2 warn), resolved to warn on the strength of the specific absence finding. Machine checks: 42 references checked (0 retracted, 0 not found), one recomputable statistic consistent, two repository URLs live, prospective registration and CONSORT confirmed — most inferential results were not machine-verifiable and remain unverified.
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 1 test: 1 consistent, 0 inconsistent; 1 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 2Primary cluster level outcome: risk difference for ivermectin vs permethrin
“71.8% versus 88.5% (–16.7 percentage point difference, 95% confidence interval (CI) –26.3 to –7.1)”
Taken as given: The estimate is the risk difference in percentage points.; The confidence interval is two-sided 95%.; The normal approximation is valid for the log-likelihood-based CI.Method: Two-sided p-value derived from the estimate and 95% CI using the normal approximation (z = estimate / SE, where SE = (CI width)/(2*1.96)).How we recomputed it: pCI(-16.7, -26.3, -7.1, 0)
- lowinternal contradictionIn Table 2, the exhaustive relationship-with-index-case categories sum to 420 participants while the stated group total is n=421 (and the 35.7% is computed from 420, not 421), implying one unstated missing relationship value.
Relationship with index case:* | Brother or sister | 109 (29.9) | 150 (35.7) ... Other† | 57 (15.6) | 66 (15.7)
Table 2reviewer’s wording - lowinternal contradictionIn Table 2, the atopic-dermatitis 'Yes/No' rows sum to 363 (24+339) while n=365, with percentages computed from 363; two participants appear to have missing atopic-dermatitis status that is not footnoted.
Atopic dermatitis: | Yes | 24 (6.6) | 21 (5.1) | No | 339 (93.4) | 394 (94.9)
Table 2reviewer’s wording
Overstated conclusions
2 findings · worst highConclusions 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 not supported by the paper’s own evidenceAssessed
- Conclusions only partially backed by the presented evidenceAssessed
9 major claims checked against the paper's own evidence: 1 not fully backed by the presented evidence (unsupported or overstated).
- unsupportedReviewer 2In other situations, first line oral ivermectin could be prescribed, especially when topical application is not practical.The paper does not provide evidence for ivermectin efficacy in those settings (nursing homes, prisons) within this trial; it is an extrapolation based on convenience, not data.Evidence: No data from the trial support this claim; it is a suggestion in the Discussion.
“In other situations, first line oral ivermectin could be prescribed, especially when topical application is not practical, for example in older adults, especially residents with dementia during a scabies outbreak in a nursing home, or in people experiencing homelessness who cannot wash off the cream or who lack privacy.”
DiscussionFind in source - partialReviewer 1The sample size was substantial, with a theoretical lack of power not affecting the final results.The trial randomised 294 clusters versus 502 planned (58% of target), so it was underpowered; the robustness argument is plausible because the CI clearly excludes the margin, but the assertion that lack of power did not affect results is partly speculative and not fully established by the data as presented.Evidence: Reporting of premature termination and the wide CI that excludes the margin.
“the sample size was substantial, with a theoretical lack of power not affecting the final results.”
DiscussionFind in source - partialReviewer 25% permethrin cream could be used as first line treatment in classic scabies.The trial supports permethrin's superiority in this specific population, but the recommendation is tempered by limitations (e.g., exclusion of patients with diffuse eczematisation, potential resistance in other regions).Evidence: Discussion cites the trial results but also notes limitations and context-specific factors.
At the individual level, the high cure rate for scabies reported in our study with 5% permethrin would now favour its first line prescription in several settings: in children and young adults; in individuals with acceptable skin conditions...
Discussionreviewer’s wording - supportedReviewer 1Oral ivermectin did not show non-inferiority to 5% permethrin cream for clinical cure of scabies on day 28.The primary cluster-level analysis (71.8% vs 88.5%; –16.7 percentage points, 95% CI –26.3 to –7.1) has a lower bound well below the –10% non-inferiority margin, directly supporting the claim.Evidence: Primary cluster-level cure rates with 95% CI excluding the non-inferiority margin.
“Cluster level cure rates were 71.8% versus 88.5% (–16.7 percentage point difference, 95% confidence interval (CI) –26.3 to –7.1) for ivermectin versus permethrin.”
Table 3Find in source - supportedReviewer 1The trial showed statistical superiority of 5% permethrin cream over ivermectin.The 95% CI upper limit of the primary difference is below 0 (–7.1), which the prespecified testing hierarchy accepts as evidence of permethrin superiority.Evidence: Primary and all secondary analyses show CIs with upper limits below 0.
“Conversely, the trial showed the statistical superiority of 5% permethrin cream.”
AbstractFind in source - supportedReviewers 1, 2Ivermectin performed worse than permethrin in all analyses (index cases and participants).The secondary outcomes (index cases –14.9, 95% CI –23.6 to –6.2; participants –9.2, –14.9 to –3.5) and completer/compliant sensitivity analyses all show directionally consistent inferiority of ivermectin.Evidence: Secondary outcome estimates with CIs all favouring permethrin.
Ivermectin performed worse in all analyses (index cases 76.6% v 91.5%; –14.9 percentage point difference, 95% CI –23.6 to –6.2; participants 85.3% v 94.2%; –9.2 percentage point difference, –14.9 to –3.5).
Resultsreviewer’s wording - supportedReviewer 15% permethrin cream could be used as first-line treatment in classic scabies.This clinical recommendation follows directly from the demonstrated superiority of permethrin on the primary endpoint.Evidence: Primary and secondary endpoint results.
“5% permethrin cream, applied to the whole body on day 0 and day 10, could be used as first line treatment in classic scabies”
What this study addsFind in source - supportedReviewer 2The trial did not show non-inferiority of ivermectin compared to permethrin.The primary outcome results show that the lower bound of the 95% CI (-26.3) is below the non-inferiority margin (-10%), so non-inferiority is not shown.Evidence: Table 3: cluster cure rate difference -16.7 (95% CI -26.3 to -7.1).
“The results of this cluster randomised trial of classic scabies, confirmed by dermoscopy, did not show the non-inferiority of oral ivermectin compared with 5% permethrin cream, given on days 0 and 10, in achieving clinical cure of scabies on day 28 in index cases and their household members.”
AbstractFind in source - supportedReviewer 2Permethrin was statistically superior to ivermectin.The upper bound of the 95% CI for the primary outcome is below 0, indicating superiority of permethrin at the 5% significance level.Evidence: Table 3: primary outcome CI does not cross 0. The paper states 'the upper bound of the confidence interval was below 0, however, meaning that permethrin was statistically superior to ivermectin'.
“The upper bound of the confidence interval was below 0, however, meaning that permethrin was statistically superior to ivermectin”
Figure 2Find in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoint is clinical cure of scabies, defined as disappearance of clinical signs and symptoms (pruritus, burrows, papules/vesicles) assessed by a masked dermatologist. This is a direct clinical outcome, not a surrogate biomarker.
“Clinical cure in the index case was defined as the simultaneous disappearance of the clinical signs and symptoms of scabies (ie, pruritus, burrows, and papules or vesicles), assessed by a dermatologist masked to the treatment”
- ADEQUATEEffect sizeThe primary result is a between-group difference in cluster-level cure rates: 71.8% for ivermectin vs 88.5% for permethrin, a -16.7 percentage point difference (95% CI -26.3 to -7.1). This is anchored to the prespecified 10% non-inferiority margin, and the CI excludes non-inferiority, demonstrating a clinically meaningful superiority of permethrin.
“Cluster level cure rates were 71.8% versus 88.5% (–16.7 percentage point difference, 95% confidence interval (CI) –26.3 to –7.1) for ivermectin versus permethrin.”
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
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.
- Ethics/consent reporting incompleteAssessed
Prior work is cited extensively (WHO burden data, two Cochrane systematic reviews with contradictory conclusions, a 2017 network meta-analysis). The rationale is explicit: uncertainty about non-inferiority/superiority motivated the trial. Limitations of prior RCTs are concretely addressed — dermoscopic confirmation, cluster design for household transmission, day-28 evaluation for non-ovicidal drugs, and masking of assessors.
“the conclusions of two Cochrane systematic reviews, published in 2007 and 2018, were contradictory: permethrin superiority or no difference in efficacy between the drugs, possibly explained by endpoint differences.”
“The non-inferiority or superiority of both scabicide drugs is uncertain, and therefore we conducted a cluster designed, assessor blinded, randomised controlled trial comparing oral ivermectin with 5% permethrin cream”
“five methodological flaws potentially account for the research gap (ie, the lack of study consensus on the superiority or non-inferiority of a drug).”
“The non-inferiority or superiority of both scabicide drugs is uncertain, and therefore we conducted a cluster designed, assessor blinded, randomised controlled trial comparing oral ivermectin with 5% permethrin cream for children and adults with classic scabies.”
Randomisation was by an independent statistician using a 1:1 list with blocks of four, grouped by centre; the unit was the household cluster. Blinding is explicitly limited to assessors with a stated rationale for not doing a double-blind double-dummy design. A power calculation (80% cure each arm, 10% margin, 2.5% one-sided alpha, 80% power, 502 clusters) is reported. Inclusion/exclusion criteria are pre-specified. Missing data were handled by multiple imputation (fully conditional specification), satisfying the missing-data/analysis-population expectation.
“An independent statistician, not involved in the recruitment of index cases or clusters, generated a 1:1 randomisation list with blocks of four, grouped according to centre.”
“Assuming a 2.5% one sided type I error rate and 80% power, recruitment of 502 clusters was required.”
“An independent statistician, not involved in the recruitment of index cases or clusters, generated a 1:1 randomisation list with blocks of four, grouped according to centre.”
“sample size calculation was standard (ie, without considering any clustering) on the basis of the hypothesis that the cluster level cure rate would be 80% in each arm, with a 10% non-inferiority margin.”
“Missing values were managed with a multiple imputation approach based on the fully conditional specification method.”
Table 1 (index cases) and Table 2 (cluster members) report sex, median age with IQR, weight, height, and medical history (e.g., atopic dermatitis, prior scabies treatment, corticosteroid/immunosuppressive use). Race/ethnicity is not reported, but the core demographic and health variables are comprehensive. Species/strain/housing criteria are not applicable for a human trial.
“Median (IQR) age (years) | 15.9 (7.9-29.9) | 15.8 (8.0-27.5)”
The trial was approved by the French National Ethics Committee (CPP SOOM 3) and the National Agency for the Safety of Medicines and Health Products (ANSM). Written informed consent was obtained from all participants or their legal representatives. However, the paper does not include a statement such as 'the study was conducted in accordance with the Declaration of Helsinki' or 'ICH-GCP', which is required for adequate regulatory compliance reporting.
“approved by the local principal investigators, the French National Ethics Committee (CPP SOOM 3), and the National Agency for the Safety of Medicines and Health Products (ANSM).”
“All participants (ie, adults, children’s parents or legally authorised representatives, and children) provided written informed consent.”
“All participants (ie, adults, children’s parents or legally authorised representatives, and children) provided written informed consent.”
Both drugs are identified with dose (ivermectin 200 µg/kg; 5% permethrin cream), regimen (day 0 and day 10), and application instructions; the suppliers (MSD-Merck and Codexial) are named in the competing-interests section. SAS version 9.4 is identified for analysis. Antibodies, cell lines, mycoplasma, and organisms are not applicable to this drug trial.
“All participants received oral ivermectin 200 µg/kg or 5% permethrin cream on day 0 and day 10.”
“oral ivermectin and topical 5% permethrin were provided free of charge by MSD-Merck and Codexial, respectively”
“For statistical analyses, we used SAS version 9.4 software.”
“for part of the study, oral ivermectin and topical 5% permethrin were provided free of charge by MSD-Merck and Codexial, respectively”
“For statistical analyses, we used SAS version 9.4 software. Codes are available at https://doi.org/10.5061/dryad.sn02v6xjd”
Mixed models with centre as random and multiple imputation are described; effect sizes are reported as risk differences with 95% CIs across cluster, index-case, and participant analyses, so exact_p_values is not applicable (estimation idiom). I verified the printed proportions against the printed counts (e.g., 94/132=71.2%, 114/127=89.8%, 407/479=85.0%, 434/446=97.3%) and baseline percentages; all were consistent. A few Table 2 sub-row denominators are short by 1–2 participants (relationship row sums to 420 vs n=421; atopic dermatitis 363 vs 365), consistent with unstated missing values, not arithmetic errors.
“For cluster level and index case outcomes, mixed models were fitted with the centre considered random.”
“–16.7 percentage point difference, 95% CI –26.3 to –7.1”
“Completer participants† | 94/132 (71.2) | 114/127 (89.8)”
“For cluster level and index case outcomes, mixed models were fitted with the centre considered random. The identity link function was used, which allowed estimations of risk difference.”
“71.8% versus 88.5% (–16.7 percentage point difference, 95% confidence interval (CI) –26.3 to –7.1)”
“For statistical analyses, we used SAS version 9.4 software.”
A data-availability statement gives a concrete route (Dryad DOI). The underlying data and the analysis code are both deposited at the same Dryad DOI, satisfying repository deposit and code sharing. Sequencing accession numbers are not applicable to this clinical trial.
“The data underlying the findings in this paper are openly and publicly available and can be found here: ( https://doi.org/10.5061/dryad.sn02v6xjd ).”
“Codes are available at https://doi.org/10.5061/dryad.sn02v6xjd .”
“The data underlying the findings in this paper are openly and publicly available and can be found here: ( https://doi.org/10.5061/dryad.sn02v6xjd ).”
“Codes are available at https://doi.org/10.5061/dryad.sn02v6xjd”
Methods are detailed enough to replicate (treatments, dosing, timing, outcome definitions). Trial registration (NCT02407782, EudraCT) is stated as prospective. CONSORT (with cluster and non-inferiority extensions) is referenced. The negative primary outcome (non-inferiority not shown) and the underpowered day-56 assessment are transparently reported. Limitations are discussed explicitly. Conclusions are proportional to the data. Funding source and ICMJE competing-interest declarations are provided.
“our report followed the Consolidated Standards of Reporting Trials (CONSORT) guideline for parallel group randomised trials and its extension to non-inferiority and cluster trials.”
“Trial registration NCT02407782 (https://clinicaltrials.gov/ct2/show/NCT02407782) .”
“The study was funded by the French Ministry of Health (Programme Hospitalier de Recherche Clinique, PHRC-N 2011) and the French Society of Dermatology.”
“Trial registration NCT02407782 (https://clinicaltrials.gov/ct2/show/NCT02407782)”
“Our report followed the Consolidated Standards of Reporting Trials (CONSORT) guideline for parallel group randomised trials and its extension to non-inferiority and cluster trials.”
“Our study had some limitations. The generalisability of our results is restricted to patients without diffuse eczematisation or extensive impetiginisation.”
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 42 references by DOI: 37 verified — 5 no DOI (shown, not verified).
- NO DOIInterventions for treating scabiesNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIIvermectin and permethrin for treating scabiesNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAbout ScabiesNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIThe use of ELM to monitor the success of antiscabietic treatment. Epiluminescence light microscopyNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIScabiesNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
2 data/code links checked; 2 live.
- dataDryadLIVEHTTP 200https://doi.org/10.5061/dryad.sn02v6xjdResolves to Dryad (data repository).
- datahttps://clinicaltrials.gov/ct2/show/NCT02407782LIVEHTTP 200Resolves, but the content could not be matched to the paper.
Copyediting
4 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 4 minor suggestions below.
4 copyedit issues flagged: mostly other, typo, grammar.
- MINORtypoResults, cutaneous adverse events paragraph“at least one cutaneous adverse event potentially related to the study drug related (9.9% and 13.6% of adverse events, respectively)”→ Remove the duplicated 'related': '...potentially related to the study drug (9.9% and 13.6% of participants, respectively)'Duplicated word; also 'of adverse events' should read 'of participants' since 14/142 and 20/147 are participant percentages.
- MINORgrammarTable 3, column header“Cures rate difference (percentage points (95% CI))”→ Change to 'Cure rate difference (percentage points (95% CI))'Inconsistent with 'cure rate' used elsewhere in the table and text.
- MINORotherTable 3 header“Cures rate difference”→ Cure rate differenceInconsistent: 'Cures rate difference' should be 'Cure rate difference' to match the singular form used elsewhere.
- MINORotherTable 2, row 'Father or mother'“Father or mother | 156 (42.7) | 164 (39.0)”→ No change needed, but note the numbers are properly aligned.The table appears well-formatted; no actual issue found.
The published paper is robust, and no retraction-level concern arises; an informed reader should weigh the cluster-unadjusted power calculation when interpreting the failed non-inferiority, the unsupported Discussion extrapolation about first-line ivermectin in settings not studied, and the missing Helsinki/ICH-GCP statement, alongside the minor Table 2 footnotes and typos. A correction/erratum addressing the claim over-reach, the ethics-framework statement, and the Table 2 missing-data footnotes — plus a cluster-informed power re-analysis — would materially strengthen confidence.
- 1.HIGHrigorAdd supporting evidence for, or temper/remove, the Discussion claim that 'in other situations, first line oral ivermectin could be prescribed, especially when topical application is not practical' — the trial provides no efficacy data for nursing homes, prisons, or other settings where topical application is impractical, so the sentence should be flagged as an extrapolation or supported with cited external evidence.An unsupported extrapolation beyond the studied population is the one substantive over-reach in an otherwise proportionally worded report, and the claim audit flagged it as under-evidenced.
- 2.HIGHethicsAdd an explicit statement in the Ethics section (e.g., in a correction/erratum) that the trial was conducted in accordance with the Declaration of Helsinki and ICH-GCP guidelines.The Ethics section names the approving committees and registration but lacks the explicit framework-adherence statement, which is the basis of the ethical_approvals 'warn'.
- 3.HIGHstatisticsAdd a cluster-adjusted power re-analysis (design effect / intraclass correlation coefficient and effective power) as a supplementary analysis or correction, since the reported sample size calculation was 'standard (ie, without considering any clustering)' in a cluster-randomised trial.Without an ICC-adjusted power estimate, readers cannot judge how much the ignored clustering reduced power for the failed non-inferiority primary outcome.
- 4.MEDIUMcopyeditRemove the duplicated word in the Results adverse-events sentence ('...potentially related to the study drug related...') and change 'of adverse events' to 'of participants' (14/142 and 20/147 are participant percentages).The duplicated word is an obvious typo, and the mislabelled denominator could mislead a careful reader.
- 5.MEDIUMcopyeditChange the Table 3 column header 'Cures rate difference (percentage points (95% CI))' to 'Cure rate difference (percentage points (95% CI))' for consistency with the singular usage elsewhere in the table and text.Inconsistent terminology in a key results table is easily caught by reviewers and detracts from an otherwise clean presentation.
- 6.MEDIUMreportingAdd footnotes to Table 2 disclosing the two unstated missing values: the relationship-to-index-case categories sum to 420 vs the stated n=421, and the atopic-dermatitis rows sum to 363 vs n=365, with percentages computed from the smaller denominators.These low-severity internal contradictions (flagged by the integrity check) are harmless if footnoted as missing data but look like arithmetic errors if left unexplained.
- 7.MEDIUMreportingEither add race/ethnicity to the baseline tables (if collected) or add a sentence in Methods or Limitations explaining why it is not reported (e.g., not routinely collected in France).The demographics sub-criterion was marked reported_but_inadequate solely because race/ethnicity is absent; a justification closes the gap without new data.
- 8.LOWstatisticsAdd a sentence or supplement citing support for the robustness of the pre-specified mixed models (e.g., residual diagnostics or bootstrap sensitivity) so that model assumptions are verifiable.R2 noted that mixed-model assumptions are not discussed; this is adequate for a pre-specified analysis model, but explicit diagnostics would strengthen verifiability.
- 9.LOWreportingMove or duplicate the manufacturer/source statement (ivermectin from MSD-Merck, permethrin from Codexial) from Competing interests into the Methods 'Intervention' section.Readers should not need to consult the competing-interest statement to identify the source of the investigational products.
- 10.LOWdata codeClarify in the Data availability statement that the single Dryad DOI (10.5061/dryad.sn02v6xjd) contains both the dataset and the analysis code, or provide a separate code link.Both URLs resolve and are content-consistent, so this is purely a clarity improvement for the shared data/code deposit.
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.