{ "slug": "ccsvi-ms-liberation", "authoring_model": "Sol 6.1", "adversarial_reviewer": "Claude Opus 5.5 (Anthropic)", "review_date": "2026-10-14", "canonical_path": "hsef_v53_harm_related_three_2026-10-12\\ccsvi-ms-liberation\\ccsvi-ms-liberation.review.md", "level_before": "Β", "level_after": "Β", "descriptor_after": "Associational", "confidence_before": "Moderate", "confidence_after": "Moderate", "causal_role_before": "epiphenomenon", "causal_role_after": "epiphenomenon", "msd_changed": false, "pmids_checked": 71, "pmids_failed": 0, "retractions_or_eoc": 0, "claims_spot_checked": 16, "corrections_applied": 8, "counts_by_severity": { "Critical": 0, "Major": 1, "Minor": 5, "Editorial": 2 }, "validator": { "tool": "framework/hsef-hypothesis-review-v5/tools/validate_review.py", "result": "FAIL", "fail_checks": [ "v5.lay_summary" ], "passes": 31, "pre_edit_result_identical": true, "note": "The only failure is the v5.4 lay_summary requirement, which post-dates this v5.3 review; the identical failure is produced by the pre-edit file, so it is not a regression." }, "findings": [ { "id": "F1", "severity": "Major", "type": "harm_understatement", "location": [ "Evidence and causal interpretation (Canadian trial)", "decisive-perturbation item [2]", "evidence card [2]" ], "original": "Its blinded physician classified six SAEs as unrelated / Six serious events were judged unrelated by the blinded physician; do not count all as procedural injury.", "corrected": "Serious adverse events occurred in 5 of 49 venoplasty and 1 of 55 sham participants; the blinded physician judged all six unrelated to the procedure, so they are not counted as procedural injury, but the arm imbalance is recorded.", "evidence": "PMID 30266886 abstract: one sham (2%) and 5 (10%) venoplasty participants had a serious adverse event; 55 sham and 49 venoplasty randomized.", "level_impact": "none" }, { "id": "F2", "severity": "Minor", "type": "misstatement", "location": [ "decisive-perturbation item [22]", "evidence card [22]" ], "original": "surgical removal of a damaged angioplasty balloon", "corrected": "surgical femoral venotomy to retrieve an angioplasty balloon", "evidence": "PMID 21107001 abstract: surgical opening of femoral vein to remove angioplastic balloon in one case (0.3% of procedures); the balloon is not described as damaged.", "level_impact": "none" }, { "id": "F3", "severity": "Minor", "type": "misstatement", "location": [ "decisive-perturbation item [30]", "evidence card [30]" ], "original": "similar low CCSVI prevalence in MS, other neurological disorders and healthy controls", "corrected": "similar low CCSVI prevalence in MS, other neurodegenerative diseases and healthy controls", "evidence": "PMID 24014572 (CoSMo) abstract names the comparator cohort as patients with other neurodegenerative diseases (ONDs).", "level_impact": "none" }, { "id": "F4", "severity": "Minor", "type": "rendering_defect", "location": [ "evidence card [1]", "evidence card [7]", "evidence card [19]" ], "original": "ErratumIn: PMIDnot supplied.", "corrected": "ErratumIn: PMID not supplied in the PubMed record.", "evidence": "Known reference-builder defect. A PubMed published-erratum search on this topic returns only 27110450 and 28072565, neither matching these three records, so no identifier could be substituted.", "level_impact": "none" }, { "id": "F5", "severity": "Minor", "type": "rendering_defect", "location": [ "REFERENCES 12, 13, 27, 28, 51, 53, 54, 55, 65" ], "original": "DOI: .", "corrected": "DOI: none in PubMed record.", "evidence": "PubMed metadata confirms no DOI for PMIDs 21876515, 22155803, 20018140, 20351674, 20351669, 20351666, 20351668, 22364941, 20351672.", "level_impact": "none" }, { "id": "F6", "severity": "Editorial", "type": "rendering_defect", "location": [ "decisive-perturbation item [4]", "evidence card [4]" ], "original": "RR1.42, 95%CI1.00–2.01, p.032", "corrected": "RR 1.42, 95% CI 1.00–2.01, p = 0.032", "evidence": "PMID 31735108 abstract: RR 1.42, 95% CI 1.00 to 2.01, p=0.032.", "level_impact": "none" }, { "id": "F7", "severity": "Editorial", "type": "rendering_defect", "location": [ "REVIEWER NOTES", "Consensus and original-source reconciliation" ], "original": "blocked403 / amended target300", "corrected": "blocked (HTTP 403) / amended target of 300", "evidence": "Missing-space rendering defects.", "level_impact": "none" }, { "id": "F8", "severity": "Minor", "type": "confidence_rationale_from_access_depth", "location": [ "scorecard final_score.confidence_rationale" ], "original": "69 primary reports with 8 selected partial-body examinations; density, heterogeneous controls and incomplete access set the confidence, not correctness.", "corrected": "Moderate reflects evidence density and reproducibility, not the reviewer's access depth and not any belief about correctness: three independent sham-controlled trials and several blinded diagnostic programmes replicate the clinical null, but every trial was small or under-recruited with incomplete target engagement, and the diagnostic literature remains operator- and protocol-dependent. High was considered and declined on that heterogeneity; Low was declined because the work is peer reviewed and replicated across centres.", "evidence": "hsef-framework.md ties the confidence band to evidence density and reproducibility and does not list the reviewer's access depth, so 'incomplete access' was not a legitimate determinant. The band itself was re-derived and retained.", "level_impact": "none; confidence band unchanged" } ], "rejected_findings": [ { "item": "headline letter and aggregation rule", "reason": "Primary-subclaim rule pre-registered with SC3 named; alternatives correctly reported as weakest-link Α and scope-matched Β; the Ε on SC4 is scope-restricted to procedural injury and correctly does not set the headline. No index entry prescribes a different rule." }, { "item": "causal role epiphenomenon", "reason": "Step 1 downstream-consequence reading considered; PMID 23301144 found no relation between MS duration and venous-lesion severity, so no dependency is demonstrated in either direction and step 2 governs." }, { "item": "momentum 'Little new work'", "reason": "Required artifact present: completed PREMiSe, Canadian, Brave Dreams and wait-list programmes are named with the search bound. Not the 'Decisive work underway' boilerplate." }, { "item": "divergence 'Contested'", "reason": "Permitted without an arithmetic gap where no single implied level applies; the review says so explicitly." }, { "item": "collective-author rendering", "reason": "REFERENCE 21 already restores 'Imperial College CCSVI Investigation Group' for PMID 21803799; no other reference has a collective first author." }, { "item": "citation years", "reason": "PMIDs 21839654, 31445091, 31735108, 22088659, 19060024, 24243930, 24531803 print the issue year against an earlier e-pub year; standing convention rejects these." }, { "item": "compact numeric tokens", "reason": "'phase2', 'rubric01', 'Step11', 'MSQOL54', 'online2019' are house style throughout the document." }, { "item": "PMID 28430292 rat ligation attribution", "reason": "The PubMed record carries no abstract, but the publisher abstract confirms eight-month follow-up, increased white-matter iron, increased astrocyte size/number with raised GFAP, and no demyelination, inflammation, axonopathy, weakness or paralysis. Content verified." }, { "item": "scorecard review_metadata.reviewer = 'Codex'", "reason": "Pre-existing provenance field common to this corpus; the adversarial banner and approval section state the authoring model (Sol 6.1) explicitly. Flagged to the lead rather than silently rewritten." }, { "item": "renderer bug: 'Best-supported rival' printing the nominal model", "reason": "Absent. The score card marks the rank-1 model is_nominal=false and the nominal model at its own lower rank; the AT A GLANCE prose names the rank-1 rival." }, { "item": "false-approval boilerplate and foreign-provenance clauses", "reason": "Absent. This case is not one of the 9 October revised editions, carries no 'earlier adversarial approval' sentence, and its Selection and methods provenance statements apply to this case. REVIEWER NOTES and the Consensus section do not contradict each other." }, { "item": "confidence band Moderate -> High", "reason": "Re-derived against the framework's band texts and retained at Moderate; see the finding above for the specific heterogeneity and independence limits that keep it out of the High band." } ], "residual_limitations": [ "Corrigendum body for PMID 23380649 remains inaccessible; corrected adverse-event numerators for the 462-patient Italian series are unverified.", "Target-engagement percentages, power calculations and the amended Brave Dreams recruitment target rest on partial-body passages not independently re-retrieved in this pass.", "The harm corpus is case reports and uncontrolled series, supporting a can-cause claim rather than an incidence." ], "package_level_strings_for_lead_to_propagate": [ { "file_hint": "READER.html", "old": "Six serious events were judged unrelated by the blinded physician; do not count all as procedural injury.", "new": "Serious adverse events occurred in 5 of 49 venoplasty and 1 of 55 sham participants; the blinded physician judged all six unrelated to the procedure, so they are not counted as procedural injury, but the arm imbalance is recorded." }, { "file_hint": "READER.html", "old": "surgical removal of a damaged angioplasty balloon", "new": "surgical femoral venotomy to retrieve an angioplasty balloon" }, { "file_hint": "READER.html", "old": "similar low CCSVI prevalence in MS, other neurological disorders and healthy controls", "new": "similar low CCSVI prevalence in MS, other neurodegenerative diseases and healthy controls" }, { "file_hint": "READER.html", "old": "ErratumIn: PMIDnot supplied.", "new": "ErratumIn: PMID not supplied in the PubMed record." }, { "file_hint": "READER.html", "old": "RR1.42, 95%CI1.00–2.01, p.032", "new": "RR 1.42, 95% CI 1.00–2.01, p = 0.032" }, { "file_hint": "READER.html", "old": "blocked403", "new": "blocked (HTTP 403)" }, { "file_hint": "READER.html", "old": "amended target300", "new": "amended target of 300" }, { "file_hint": "MANIFEST_SHA256.json", "old": "pre-review hashes for ccsvi-ms-liberation.review.md and ccsvi-ms-liberation.scorecard.json", "new": "see post_edit_hashes in this file" } ], "approval": "Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-14.", "confidence_band_rederived": "Moderate retained; rationale rescoped off access depth", "lead_policy_applied": { "confidence_two_step_test": "Lead policy, round 2: (1) strike access depth as a confidence determinant everywhere; (2) re-derive the band from what remains, raising it only if access depth was the sole thing holding it down and the evidence meets the High text. Applied to this case; see the confidence finding for the outcome.", "momentum_decisive_work": "Lead policy, round 2: 'Decisive work underway' requires work actually being run that could cross the next threshold; a completed publication or trial never satisfies it, even when the artifact names a real study.", "false_approval_boilerplate": "Lead policy, round 2: the 2026-10-09 splice inserted 'The earlier adversarial approval refers to its original edition' unconditionally. Verify against prepublication-enrichment-2026-10-05/ADVERSARIAL_REVIEW.md before changing; rescope only where no earlier approval exists.", "renderer_best_supported_rival": "Lead policy, round 2: check the score card's is_nominal / rank_in_tournament fields; fix the prose, not the data, if the nominal is printed as the rival." } }