{ "slug": "h-pylori-peptic-ulcer", "case_path": "consensus_consistency_audit_2026-10-09\\revised\\h-pylori-peptic-ulcer", "round": 2, "group": 1, "review_date": "2026-10-14", "adversarial_reviewer": "Claude Opus 5.5 (Anthropic)", "authoring_model_of_record": "Sol 6.1", "authoring_model_as_declared_in_document": "Editions A and B graded by OpenAI Codex; red-team synthesis pass-01 by Claude (Anthropic), 2026-09-27", "status": "Approved as revised", "level_before": "Θ", "level_after": "Θ", "descriptor_after": "High predictive accuracy, minimal unresolved exceptions", "confidence_before": "High", "confidence_after": "High", "role_before": "primary driver", "role_after": "primary driver", "subclaims_before": { "SC1": "Θ", "SC2": "Θ", "SC3": "Η" }, "subclaims_after": { "SC1": "Θ", "SC2": "Θ", "SC3": "Η" }, "aggregation_rule": "primary-subclaim (SC1)", "aggregation_alternatives": { "weakest_link": "Η", "primary_subclaim": "Θ", "scope_matched": "Θ" }, "msd_before": { "momentum": "Little new work", "stability": "Stable", "divergence": "Aligned (gap 0)" }, "msd_after": { "momentum": "Little new work", "stability": "Stable", "divergence": "Aligned (gap 0)" }, "sources_total_before": 23, "sources_total_after": 23, "distinct_studies": 22, "claims_total": 34, "pmids_checked": 23, "pmids_failed": 0, "retractions_or_eocs_found": 0, "claims_spot_checked": 12, "claims_spot_check_failures": 2, "validator": { "tool": "framework/5.4/validate_review.py", "before": "PASS - 0 fail, 2 warn, 34 pass", "after": "PASS - 0 fail, 2 warn, 34 pass", "warnings": "Both are the disclosed calibration warnings: the calibration entry was not graded blind, and the expected band was written by an author who had seen the review. Both are already stated in the review and are not fixable by editing the card.", "output_file": "validator.txt" }, "findings_by_severity": { "critical": 0, "major": 1, "minor": 1, "editorial": 1 }, "findings": [ { "id": "C-1", "severity": "Editorial", "category": "imprecise characterization of a source's method", "location": "Limits and genuine alternative causes (second paragraph), and the matching claim-register row", "original": "confirmed infected under a two-test definition", "corrected": "confirmed infected under a confirmatory definition - positive culture, or positive histology and CLOtest together, with at least two negative tests required to call a patient uninfected", "evidence": "Ciociola et al. (PMID 10406244) define infection as culture growth OR both histology and CLOtest positive, and non-infection as at least two negative tests. 'A two-test definition' implies two positives were always required. The 73%-of-2394 figure the sentence carries is correct and unchanged.", "artifacts_changed": [ "h-pylori-peptic-ulcer.review.md" ] }, { "id": "C-2", "severity": "Minor", "category": "asymmetric caveat / incomplete limitation", "location": "Clinical evidence and what it establishes - 'The finding generalizes beyond duodenal recurrence'", "original": "Graham et al. observed lower recurrence of both gastric and duodenal ulcers after an eradication regimen, with no maintenance antiulcer treatment after healing.", "corrected": "... after healing; its triple therapy also contained bismuth subsalicylate, so the same mucosal-action caveat noted below applies, and the bismuth-free attribution rests on the duodenal trials.", "evidence": "The review flags the mucosal action of a bismuth-containing regimen for Sung (PMID 7800005) in the next clause but not for Graham (PMID 1558340), whose triple therapy was tetracycline, metronidazole and bismuth subsalicylate. Those two are the review's only gastric-site sources, while the 'narrow causal attribution' limb of its Theta case rests on the bismuth-free duodenal/NSAID-context trials (PMID 8419816, PMID 11809180), so the gastric generalization and the clean-attribution argument do not overlap. The review already concedes the parallel gap on the pathway limb ('satisfied for duodenal ulcer and only partly for gastric ulcer').", "level_impact": "None. SC1 is a disease-class claim the duodenal evidence alone supports, and SC3 (generalization) already holds at Eta.", "artifacts_changed": [ "h-pylori-peptic-ulcer.review.md" ] }, { "id": "C-3", "severity": "Major", "category": "false provenance claim about adversarial review", "location": "Consensus provenance reconciliation block, and consensus-addendum.md", "original": "The earlier adversarial approval refers to its original edition; this provenance supplement is not a new independent or blinded adjudication.", "corrected": "No earlier adversarial approval is attached to this case; this provenance supplement is not a new independent or blinded adjudication.", "evidence": "integrate_consensus.py line 71 appends this sentence for every case with no approval check. Exactly 26 reviews carry the 2026-10-06 approval line; they were enumerated and h-pylori-peptic-ulcer is not among them.", "lead_flagged": true, "artifacts_changed": [ "h-pylori-peptic-ulcer.review.md", "consensus-addendum.md" ] } ], "rejected_findings": [ { "id": "R-1", "checked": "Three references print a year one off the PubMed publication-date year (Levenstein 2015/2014; Lee 2013/2012; Medel-Jara 2026/2025)", "disposition": "Rejected - accepted print-versus-e-pub convention." }, { "id": "R-2", "checked": "The Eta-to-Theta upgrade itself", "disposition": "Rejected, i.e. Theta stands. Theta requires minimal, not zero, unresolved exceptions, and each candidate second exception fails: the ~quarter of strictly tested US non-NSAID duodenal ulcers that were uninfected bears on near-universality, which the claim disclaims; the Vergara null in prior NSAID users is a baseline-determinable context boundary the v5.3 scope rule permits; Cochrane's very-low quality rating concerns precision, not direction. HEAT's later window remains the single unresolved within-scope exception, is carried by name rather than absorbed by redefinition, is penalised once (SC3, not SC1 twice), and edition B's case for Eta is retained verbatim as dissent." }, { "id": "R-3", "checked": "A Theta headline from a subclaim set containing an Eta", "disposition": "Rejected. The rule is primary-subclaim, pre-registered in edition A before any subclaim was scored, and SC1 is coextensive with the graded claim. Scope-matched returns Theta for the same reason, and weakest-link's Eta is disclosed in three separate places. This is the opposite of the scope-matched misread the brief warns about." }, { "id": "R-4", "checked": "Role printed in AT A GLANCE as 'Primary driver, within the infection-mediated causal route' rather than the bare canonical label", "disposition": "Rejected. The scorecard classification is exactly 'primary driver', one of the five permitted values; the table adds a scope qualifier the derivation justifies, and the review states explicitly that the rate-limiting limb is an inference from outcome changes." }, { "id": "R-5", "checked": "Momentum 'Little new work'", "disposition": "Rejected. Artifact named and checkable: HEAT (2022 and 2025 reports) is the last completed causal trial with an ulcer endpoint, and the one current trial in the ledger is identified as a regimen trial with an eradication endpoint." }, { "id": "R-6", "checked": "Divergence 'Aligned: field framing implies Theta; assessed Theta; gap 0'", "disposition": "Rejected. Arithmetic stated, and the implied level is sourced to a verbatim sentence verified in PMID 15106235: 'Helicobacter pylori infection is the main etiologic factor for peptic ulcer disease.'" }, { "id": "R-7", "checked": "Lay summary says deliberate ingestion inflamed the stomach lining of healthy 'volunteers' (plural) while each report is n=1", "disposition": "Rejected. Two separate self-ingestion reports are cited (PMID 3982345, PMID 3826027), and the technical sections label both as n=1 with no ulcer endpoint. The plural is accurate across the two records." }, { "id": "R-8", "checked": "Known upstream builder defects: collective author with leading comma, fused numerals, empty DOI fields, 'PMIDnot supplied'", "disposition": "Rejected, all absent. All 23 references carry a named individual first author, every inline PMID is spaced, no empty DOI field exists, and the one reference without an indexed DOI says so explicitly." }, { "id": "R-9", "checked": "Lead-flagged Oct-9 revision artifact: a spliced Consensus section coexisting with preserved REVIEWER NOTES still asserting that no Consensus work was performed", "disposition": "Rejected - checked and not present in this case. The spliced block is headed \"Consensus provenance reconciliation - 9 October 2026\" and is self-limiting: it states that the located records are historical checks, \"not newly credited complete-paper examinations\", that no source-comparison count is inferred, and that the supplement \"is not a new independent or blinded adjudication\". The preserved reviewer notes and the scorecard literature_tools_note make no negative claim about Consensus; they positively enumerate the retrieval tools used by the authoring passes, which predate the 2026-10-05 enrichment the Consensus records come from. No sentence in either review contradicts the spliced block, so there was nothing to scope and nothing was edited." }, { "id": "R-10", "checked": "Renderer bug: 'Best-supported rival' printing models_evaluated[0], often the nominal", "disposition": "Rejected - this is the one case in this group where models_evaluated[0] IS the nominal (M1, rank 1, is_nominal true), so the bug could have fired; but the prose names M5, the rank-2 heterogeneous multiple-cause model, and says explicitly that M5 contains the nominal. The prose is correct and no change was made." }, { "id": "R-11", "checked": "Confidence band set from access depth", "disposition": "Rejected - not present. This case's rationale is already purely evidence-based (independent investigators, countries, ulcer sites, endpoints and designs replicating the intervention-sensitive pattern; HSEF confidence separated from GRADE certainty). Access depth is not offered as a ground for High." } ], "residual_limitations": [ "Access depth is abstract and indexed metadata for 22 of 23 records plus targeted passages of the HEAT full report. Three weight-bearing figures could not be reached at abstract depth and rest on that targeted reading: HEAT's second-period event counts (12 vs 9), the repeat-breath-test result at a median 3.95 years (146/161 negative in the eradication arm vs 41/161 in control), and Sung's one-year recurrence subset. All are consistent with the published abstracts; none is flagged as an error.", "The two validator warnings are real and disclosed: the calibration entry was not graded blind and the sealed band was written by an author who had seen the review, so Theta falling inside an Eta-to-Theta band is not evidence of calibration - which the review also states.", "Rubric 04 is an acknowledged imperfect fit and the Beta/Gamma/Delta boundary rests on the review's explicit reading of an internally uneven rubric (eradication provision versus transfer/consortium wording). The reading is stated, quoted and consistent with cap precedence, but it is an interpretation." ], "scorecard_changed": false, "scorecard_change_reason": "No level, confidence, count or role moved; both corrections are prose-precision fixes. C-1's wording also appears in the scorecard's claim register; left for the lead to mirror or leave, since the figure it carries is unaffected.", "package_level_propagation": { "strings_to_replace": [ { "what": "h-pylori-peptic-ulcer.review.md sha256 in this case folder's SHA256.txt (and any package MANIFEST/BATCH_COMPLETION entry carrying it)", "old": "9d3349c208ed2682270bc61b79f93f822b4fbd578f93d57400cb6a0d10539c46", "new": "bab452a02ff4fd95148280a8d7f3ba36aa10873c6c7d7250c9d702c6e36db071" }, { "what": "h-pylori-peptic-ulcer.review.md sha256 in this case folder's SHA256.txt (and any package MANIFEST/BATCH_COMPLETION entry carrying it)", "old": "9d3349c208ed2682270bc61b79f93f822b4fbd578f93d57400cb6a0d10539c46", "new": "2a75e06a3b6ba9d1c65f2c469ff34b7db7f94c3f13726ad5e67976bd066fa9a6", "supersedes_earlier_entry_in_this_file": true } ], "note": "No headline string, letter, confidence, count or title changed, so OVERVIEW.md / READER.html / EDITORIAL_HANDOFF.md values are not stale. NOTE: an R-9 row was added to the review's rejected-findings table after the first hash was recorded; the final digest is the one in changed_file_hashes." }, "changed_file_hashes": { "h-pylori-peptic-ulcer.review.md": { "sha256": "c7a43ccfa41a4e3d563bd76eb1930ced15239e354d79020844d688ec3924f379", "bytes": 108147, "previous_sha256": "9d3349c208ed2682270bc61b79f93f822b4fbd578f93d57400cb6a0d10539c46" }, "validator.txt": { "sha256": "8fee60c8a2e684b27db2e8156b7a3c002001e0164a66600ecb4cc830759ed3f5", "bytes": 3161, "new_file": true }, "ADVERSARIAL_REVIEW.md": { "sha256": "2cc2aad1b5df15ac1677f25901e64e22628ac4432a630119f59b5228b87b7f53", "bytes": 10535, "new_file": true }, "consensus-addendum.md": { "sha256": "e0c5787455712a1c553909e6f0c523ffd3f830abb3c433bd3399266774226f56", "bytes": 1880, "previous_sha256": "5266903373b4f4b2c95a4c1c67ee81bf8bd02a29023d63c4be89cac41118e5a5", "change_detail": "Rescoped the false earlier-adversarial-approval sentence (same correction as in the review)." } }, "propagate_to": [], "divergent_copies": [], "consensus_questions": [ "In randomized trials of H. pylori eradication versus control among patients STARTING long-term aspirin or NSAID therapy, is the reduction in ulcer bleeding sustained beyond about three years in participants with documented durable clearance?", "In contemporary low-prevalence populations, what share of non-NSAID peptic ulcers is H. pylori-positive under confirmatory testing?" ], "lead_flagged_artifact_checks": { "false_approval_boilerplate": "present; corrected (C-3)", "confidence_from_access_depth": "absent (R-11)", "models_evaluated_0_renderer_bug": "absent, near miss (R-10)", "reviewer_notes_contradiction": "absent (R-9)", "false_hiv_draft_clause": "absent", "review_metadata_reviewer_left_unedited": true }, "line_endings_note": "CRLF preserved on every edited file, so each diff against sol61-original is confined to the corrected content.", "lead_policy_2026_10_14": { "confidence_two_step_test": { "step1_strike_access_depth": "no rewrite needed; this rationale never leaned on access depth (rejected finding R-11)", "step2_rederive_band": "High stands", "surviving_valid_determinants": [ "High rests on replication across independent investigators, countries, ulcer sites, endpoints and designs, with pooled syntheses confirming direction, which is the High text", "Cochrane's very-low quality rating concerns effect precision, not replication breadth, and the review correctly separates HSEF confidence from GRADE certainty" ], "band_before": "High", "band_after": "High", "band_changed": false }, "momentum_stricter_rule": { "value": "Little new work", "result": "holds", "basis": "The conservative value, argued FROM completed work rather than crediting it as momentum: HEAT (2022 and 2025 reports) is named as the last completed causal trial with an ulcer endpoint, and the one current trial in the ledger is identified as a regimen trial with an eradication endpoint. Nothing being run is claimed to be threshold-crossing.", "completed_work_counted_as_momentum": false }, "policy_as_stated_by_lead": "Confidence: strike access depth as a determinant everywhere, then re-derive; raise the band only if access depth was the only thing holding it down and the cited evidence meets the High text. Documented evidence-quality or reproducibility concerns are valid determinants and the band stands. Momentum: 'Decisive work underway' requires work actually being run that could cross the next threshold; a completed publication or trial never satisfies it." } }