Computed from this object's own state — a structural condition is unmet. This is not a fixed disclaimer: the conditions below are each testable, and a build in which they are met renders READY.
Both trials register hierarchical endpoints analysed by win ratio, and the two hierarchies are not the same hierarchy.
In this section: Registration and administrative information (PROSPERO field set) · What this object claims to contain
| Review title | Tafamidis or acoramidis in transthyretin amyloid cardiomyopathy |
|---|---|
| Review question (PICO) | Both trials register hierarchical endpoints analysed by win ratio, and the two hierarchies are not the same hierarchy. |
| Population | Adults with wild-type or hereditary transthyretin amyloid cardiomyopathy |
| Intervention | TTR stabilizer (tafamidis, acoramidis) or silencer (patisiran, vutrisiran) |
| Comparator | Placebo |
| Background / rationale | Not recorded — this object holds no background field, and an introduction generated without one would be argument that no source in this review supports |
| Eligibility criteria | not recorded on the page this object was built from |
| Information sources | 0 source layers, listed on the Extraction tab |
| Search strategy | The executed strings, datetimes, filters and hit counts are on the Search tab |
| Study selection process | A second independent screen was NOT run. No second independent screen was run for this topic. This is recorded as owed rather than described: the remedy is to RUN one, and writing a duplicate-screening sentence without one would be the defect this whole exercise exists to prevent. |
| Number of screeners | One. No second independent screen was run. |
| Data extraction items | Extracted per trial on this object: registry id; primary publication; arms; per-arm counts and the published effect. Fields not listed here were not extracted. |
| Outcomes and prioritisation | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ |
| Risk of bias method | RoB 2, per result, five domains — see the risk-of-bias table below |
| Synthesis methods | Not recorded — no outcome on this object carries a pooled estimate, so no synthesis method was applied |
| Subgroup analyses | Not recorded — this object records no subgroup-analysis field at all, so whether any were prespecified is UNKNOWN -- which is not the same as none having been |
| Meta-bias assessment | Not recorded — no funnel, Egger or Peters value is held on this object, so no meta-bias assessment is claimed |
| Certainty assessment | GRADE, recorded for 1 outcome(s) (GRADE). The ratings are on the Certainty tab. |
| Funding | Not recorded — no funding statement is recorded for this review |
| Competing interests | Not recorded — no declaration is recorded |
| Built | 2026-08-18T00:00Z |
| Schema | 2-authored-from-source |
not recorded on the page this object was built from
Query as executed:
query.intr="tafamidis OR acoramidis OR vyndaqel OR vyndamax OR attruby OR AG10"; filter.advanced=AREA[StudyType]INTERVENTIONAL; pageSize=300; countTotal=true
In this section: Extracted values, and where each came from · Risk of bias (RoB-2) · Overall judgement per trial · Inter-assessor agreement · References · Included studies · Regulatory documents · Methods and guidance · Sources · Hierarchical composites, analysed by win ratio -- and the two hierarchies differ · Hierarchical composites, analysed by win ratio -- and the two hierarchies differ · What was measured · Pooled result · Contributing trials · Endpoint definitions, read from the registry
One row per extracted value. Every row carries the value, the verbatim sentence it was read from, a resolvable link to the source, and whether the number was read or derived. Where any of those is absent the row says so rather than omitting the value.
| Trial / outcome | Value as extracted | Verbatim source sentence | Source links | Read or derived |
|---|---|---|---|---|
| ATTR-ACT primary | no effect value held for this outcome | no source sentence recorded — this value cannot be checked against a quoted line here | NCT01994889 · source | DERIVED by us from not carried forward -- the estimate is withdrawn The counts above are what the page held for this trial. The endpoint definition beside them is verbatim from the registry, and is what the withdrawal turns on. |
| ATTRibute-CM primary | no effect value held for this outcome | no source sentence recorded — this value cannot be checked against a quoted line here | NCT03860935 · source | DERIVED by us from not carried forward -- the estimate is withdrawn The counts above are what the page held for this trial. The endpoint definition beside them is verbatim from the registry, and is what the withdrawal turns on. |
Variant:
Unit assessed: A result, not a study -- Handbook 8.2: 'risk of bias is assessed for a specific result'. Every judgement below names its outcome.
Assessor 1: (anthropic family). Assessor 2: (openai family).
| Trial | Result assessed | Domain | Assessor 1 (anthropic) | Assessor 2 (openai) | Agreed |
|---|---|---|---|---|---|
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D1 D1_randomisation | no information | no information | yes |
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D2 D2_deviations | no information | no information | yes |
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D3 D3_missing_outcome_data | no information | no information | yes |
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D4 D4_measurement | some concerns | LOW | NO |
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D5 D5_selection_of_result | some concerns | LOW | NO |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D1 D1_randomisation | no information | no information | yes |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D2 D2_deviations | no information | no information | yes |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D3 D3_missing_outcome_data | no information | no information | yes |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D4 D4_measurement | some concerns | LOW | NO |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | D5 D5_selection_of_result | some concerns | LOW | NO |
| Trial | Result assessed | Assessor 1 (anthropic) | Assessor 2 (openai) | Agreed |
|---|---|---|---|---|
| ATTR-ACT | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | no overall judgement reached — this assessor answered “no information”, which RoB 2 defines as a response to a signalling question and not as one of the three overall categories (low, some concerns, high) | no overall judgement reached — this assessor answered “no information”, which RoB 2 defines as a response to a signalling question and not as one of the three overall categories (low, some concerns, high) | yes |
| ATTRibute-CM | Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | no overall judgement reached — this assessor answered “no information”, which RoB 2 defines as a response to a signalling question and not as one of the three overall categories (low, some concerns, high) | no overall judgement reached — this assessor answered “no information”, which RoB 2 defines as a response to a signalling question and not as one of the three overall categories (low, some concerns, high) | yes |
This assessment is DUAL — two assessors from different model families, the second asked blind. The agreement rate is withheld pending adjudication. Both readers were given this review’s decision rule; an earlier version of this sentence said the blinding had withheld it, and that was wrong. What the 10 domain comparison(s) show is a reading difference that runs in OPPOSITE directions by domain — on domains 1 to 3 the second assessor more often declines to judge at all, and on domains 4 and 5 it more often judges low. A single agreement rate averages those two behaviours into one number that describes neither. Both assessors’ per-result judgements are shown in the table above, unadjudicated, and neither is this review’s finding. No adjudication has been performed, so this review holds no final risk-of-bias judgement for these results.
| Layer | Source | How it was obtained |
|---|
| Comparator | placebo |
|---|---|
| Estimand | hierarchical composite — random-effects |
| Effect scale | reported on the log scale; nothing is pooled on this outcome |
What this does not establish: not that tafamidis or acoramidis is ineffective in transthyretin amyloid cardiomyopathy. The trials here measure hierarchical win-ratio composites of death, hospitalisation and functional change, and nothing on this page estimates the question its title implies. What is confirmed: the original withdrawal was right and its one-line reason -- 'win-ratio composite reduced to 2x2 counts' -- is exactly the defect, now sourced from both registrations. Both identities are correct. Re-verification added the second, independent ground the original did not name: the two hierarchies differ in depth and in kind. THE replacement question is parked: a synthesis of win ratios is possible in principle but requires each trial's own win-ratio estimate and its variance, which the registry's protocol record does not carry.
Hierarchical composites, analysed by win ratio -- and the two hierarchies differ. k = 2.
Authored on re-verification.
What each trial COUNTED, in its own registry record's words, before anything here was pooled. A sentence saying what HAPPENED is not a sentence saying what was COUNTED, and only the second one licenses a pool. Read from the registry on 2026-08-18. Follow the registration link to disagree with any row.
| Trial | Rank in its own trial | Registered outcome measure | Description, verbatim | Analysis set / window |
|---|---|---|---|---|
| ATTR-ACT NCT01994889 | PRIMARY -- primaryOutcomes[0] on this registration | Hierarchical Combination of All-Cause Mortality and Frequency of Cardiovascular-Related Hospitalizations | All-cause mortality and frequency of cardiovascular hospitalization were analyzed using Finkelstein-Schoenfeld method. The method combines all-cause mortality and frequency of CV-related hospitalizations in a hierarchical fashion using all-cause mortality first. The method compares every participant with every other participant within strata, assigning a +1 to the "better" participant and a -1 to the "worse" participant and 0 if they are "tied". Participants who discontinued for transplantation (heart transplantation and combined heart and liver transplantation) or for implantation of a cardiac mechanical assist device, were handled in the same manner as death. 'Win' represents a participant doing better based on hierarchical comparison. The reported unit is the total "wins" for each treatment group from performing such a hierarchical comparison across all 4 strata in the study. | 30 months. Arm sizes here are 264 against 177, which is the pooled tafamidis doses against placebo. |
| ATTRibute-CM NCT03860935 | PRIMARY -- primaryOutcomes[0] on this registration | A Hierarchical Combination of All-Cause Mortality, Cumulative Frequency of CV-related Hospitalization, Change From Baseline in NT-proBNP and Change From Baseline in 6MWT at the Last Available Visit Where Both Subjects Had Non-missing Assessments. | The endpoint was analyzed using Finkelstein-Schoenfeld method. The method combines all-cause mortality, cumulative frequency of CV-related hospitalizations, change from baseline in NT-proBNP and change from baseline in 6MWT in a hierarchical fashion. The method compares every participant with every other participant within strata, assigning a +1 to the "better" participant and a -1 to the "worse" participant and 0 if they are "tied". Participants who had heart transplantation or implantation of a cardiac mechanical assist device were handled in the same manner as death. 'Win' represents a participant doing better based on hierarchical comparison. The reported unit is the total percent of "wins" for each treatment group from performing such a hierarchical comparison across stratification factors in the study. | At the last available visit where both compared subjects had non-missing assessments. |
In this section: Hierarchical composites, analysed by win ratio -- and the two hierarchies differ · Risk of bias, both assessors · GOSH · Trial-sequential analysis · Meta-regression
⬇ SVG (vector) 8 KB
Only the vector format is offered for this figure, by choice: it is line art — discrete marks, glyphs and labels — which SVG reproduces exactly at any size, so a raster would add weight without adding fidelity. The build was able to rasterise; it was asked not to.
Every cell carries BOTH independent cross-family assessments and they are not reconciled: showing one column would be a reconciliation presented as an observation. Glyph as well as colour, so the panel survives greyscale printing and colour-blind reading. A trial with no judgement is shown as a row of not-assessed markers rather than omitted, so a gap in the assessment cannot be mistaken for a clean assessment. Agreement as measured: {'per domain agreed': 6, 'per domain total': 10, 'per domain rate pct': 60.0, 'overall agreed': 2, 'overall total': 2, 'overall is derived': True, 'recomputed from the stored judgements': True, 'why recomputed': 'The stored disagreement tallies were frozen as text and 12 of 22 no longer match the judgements in the object -- a later guidance pass moved 51 the randomisation process judgements and the tallies were never recomputed. This is derived from what the object holds now.'}.
⬇ SVG (vector) 1 KB ⬇ PNG 3500x540 81 KB
All raster formats were generated at build time from the same SVG as the graphic above (SHA-256 9091822e99aa8781…), at 600 dpi for a 6-inch print width, on a white ground. They are not screenshots of the page and do not change with the theme you are reading in.
Submission formats written to the figures directory, not embedded here because an uncompressed 600 dpi TIFF is tens of megabytes: EPS (vector, journal's first choice for line art) — gosh.eps, 2 KB; TIFF (uncompressed, 600 dpi -- journal fallback) — gosh.tiff, 5,537 KB.
Deliberately not drawn. Every subset meta-analysis of 2 trials is 3 points; a cloud that small cannot show the multimodality GOSH exists to reveal, and a reader would take the picture as evidence of its absence.
⬇ SVG (vector) 1 KB ⬇ PNG 3500x650 87 KB
All raster formats were generated at build time from the same SVG as the graphic above (SHA-256 25ed9800fdf56614…), at 600 dpi for a 6-inch print width, on a white ground. They are not screenshots of the page and do not change with the theme you are reading in.
Submission formats written to the figures directory, not embedded here because an uncompressed 600 dpi TIFF is tens of megabytes: EPS (vector, journal's first choice for line art) — tsa.eps, 2 KB; TIFF (uncompressed, 600 dpi -- journal fallback) — tsa.tiff, 6,665 KB.
TSA boundaries depend entirely on a target information size that must be pre-specified. This review's protocol registers none, so any boundary drawn here would be a parameter chosen after seeing the data -- which is the practice TSA exists to protect against.
⬇ SVG (vector) 1 KB ⬇ PNG 3500x540 87 KB
All raster formats were generated at build time from the same SVG as the graphic above (SHA-256 e35ec93bbc9a1c10…), at 600 dpi for a 6-inch print width, on a white ground. They are not screenshots of the page and do not change with the theme you are reading in.
Submission formats written to the figures directory, not embedded here because an uncompressed 600 dpi TIFF is tens of megabytes: EPS (vector, journal's first choice for line art) — bubble.eps, 2 KB; TIFF (uncompressed, 600 dpi -- journal fallback) — bubble.tiff, 5,537 KB.
The protocol pre-specifies no moderator, and at k = 2 a meta-regression would be fitted on 2 points. Not drawn rather than drawn with a caveat: a bubble plot invites reading a slope, and there is no slope here that any reader should read.
In this section: Summary of findings · Reproducibility artifact · Other recorded qualifications
| Outcome | k | Pooled effect | Certainty |
|---|---|---|---|
| Hierarchical composites, analysed by win ratio -- and the two hierarchies differ | 2 | not pooled | See comment12 |
| Canonical object | — |
|---|---|
| Registered at | — — |
| Schema | 2-authored-from-source |
| Object last updated | 2026-08-18T00:00Z |
| Source object SHA-256 | 22f068a8979538b3 |
| Page generated | 2026-08-27 20:49 UTC |
| Generator build | 509dde275, built to STANDARD v1 |
Everything a third party needs to rebuild this page. Each figure on the Analysis tab downloads as an SVG carrying exactly the values shown.
Statements this review records about the limits of what it establishes, which are not shown elsewhere on this page. They are reproduced from the object verbatim and are not summarised.
| Evidence and why it does not settle | This is THE limb that decides THE topic AND no field settles it. Both included trials pass 'hierarchical win ratio' -- and they still cannot be combined, because ATTR-ACT's hierarchy is mortality plus CV hospitalisation (2 tiers) while ATTRibute-CM's adds NT-proBNP and 6-minute-walk change (4 tiers). A win ratio's estimand IS its hierarchy. Comparing two hierarchies for identity is a reading of two verbatim strings, not a field lookup, so it is recorded as derived. Two different hierarchies are two estimands, not heterogeneity. |
|---|---|
| Components as the registry states them | a hierarchical combination of all-cause mortality and frequency of cardiovascular-related hospitalizations. Two levels, analysed by win ratio -- every patient is compared with every other and the result is a ratio of wins to losses, not a count of events over a denominator. |
| Analysis set as the registry states it | 30 months. Arm sizes here are 264 against 177, which is the pooled tafamidis doses against placebo. |
| Components as the registry states them | a hierarchical combination of all-cause mortality, cumulative frequency of cv-related hospitalization, change from baseline in nt-proBNP and change from baseline in 6-minute walk test. Four levels, and the lower two are continuous change scores, not events at all. |
| Analysis set as the registry states it | At the last available visit where both compared subjects had non-missing assessments. |
| Why absent | k=2 and nothing was pooled, so there is NO model call to quote, NO pooled estimate, NO heterogeneity and NO package version. This is not a missing artefact; it is the correct state for a topic with no synthesis. |
| Previous why absent | k=1. No meta-analysis was performed, so there is NO model call to quote, NO pooled estimate, NO heterogeneity and NO package version. This is not a missing artefact; it is the correct state for a single-study topic. |
| Regated 2026 08 19 trailing placebo | THE only one of THE five gated topics whose cascade did not move. Re-run against the repaired classifier with the old code loaded from git at 7a08bcbe1 and the surfaced set re-executed (55 today, 55 stored): Zero of the 55 records changed role. Recorded as a measured null rather than left silent, because 'unchanged' and 'not checked' look identical on a finished page. k2 role located was separately corrected from 55 to 52 -- an arithmetic defect in the stage label, not a movement. Re-derived by scripts/regate_cascade_2026_08_19.py (old vs new) and by scripts/regate_across_revisions.py, which walks every commit that touched the classifier and reports at which of them the stored numbers reproduce: 48/3/1/3 appears first at f2bf16022 and is unchanged at every revision since, so this topic is current rather than merely unexamined. |
| Why this is stated rather than left implicit | A rating of some concerns with no explanation reads as a judgement against the trial. Omitting the reason would manufacture a criticism exactly as a low-by-default manufactures a reassurance -- the same defect pointed the other way. |
| This does not replace the stored judgement | This is the judgement the published algorithm proposes from the signalling responses named here. It does not replace the stored `judgement` beside it, and nothing reads it yet. The tool itself says its algorithms "provide proposed judgements, but users should verify these and change them if they feel this is appropriate" -- so this is an input to a person, not an output to a reader. |
| Registry facts backfilled | Read from ClinicalTrials.gov API v2 on 2026-08-21. THE judgements above were not revisited -- this adds the registry evidence beside a verdict that was made from it and did not record it, so a blind second assessor can be shown the same facts. A second assessment built on fewer facts is not a comparable assessment. |
| THE allow list THE rate is conditional on | A disagreement rate is meaningless without THE facts both assessors saw, so the list is stated here the way a denominator is stated. Fields emitted: trial, cohort, registered enrolment, registered masking, registered sites, registered comparator, number of registered arms, number of registered primary outcomes, registered primary outcome, registered analysis population, the rank the pooled result holds, the result being assessed, participants this review pools, registered enrolment minus participants pooled. Established by test, not argument: widening this list from 9 fields to 14 halved gepotidacin's disagreement -- its entire selection of the reported result divergence was a fact the second assessor had never been shown -- and left ceftaroline and tigecycline unmoved. |
| Which risk of bias domains drove the rating | Where a result is rated worse than LOW overall, the domain carrying that judgement is: the randomisation process on 2 result(s), deviations from the intended intervention on 2 result(s), missing outcome data on 2 result(s), selection of the reported result on 2 result(s). The overall rating is the worst domain, so these are the domains that determine it. |
In this section: Summary
Question. Both trials register HIERARCHICAL endpoints analysed by win ratio, and the two hierarchies are not the same hierarchy.
What was checked. ClinicalTrials.gov protocol records, read 2026-08-18
What was found. 2 registered trials:
| Registration | Trial |
|---|---|
| NCT01994889 | ATTR-ACT |
| NCT03860935 | ATTRibute-CM |
Why no result is pooled. Both trials register hierarchical endpoints analysed by win ratio, and A win ratio cannot be recovered from A 2x2 table. Attr-act's registered primary, verbatim, is a 'Hierarchical Combination of All-Cause Mortality and Frequency of Cardiovascular-Related Hospitalizations'. In such an analysis every patient in one arm is compared with every patient in the other and the result is a ratio of wins to losses; there is no numerator over a denominator to extract, and reducing it to counts of participants discards the hierarchy that defines it.
Whether these trials study this subject. Checked against each registration's arm structure on 2026-08-25. 1 of 2 trial was confirmed: the drug is part of the randomised comparison. 1 could NOT be decided this way, because the registered arm names do not name the intervention the way this review does. That is not a pass: it remains unverified, and no other check has looked.
What would change this. A pooled result becomes possible when trials reporting a shared, comparably measured outcome are available.
No pooled estimate, risk-of-bias assessment or certainty rating is published for this question, because there is no combined result for them to describe.
Built 2026-08-19 by ssot/build_to_standard.py against PAGE-STANDARD.md. 11 held, 3 refusing. A refusal is a complete outcome; nothing is generated to fill a slot.
| property | state | reason |
|---|---|---|
| P1_executed_search | HELD | 1 database queries recorded verbatim with dates and counts; PRISMA arithmetic reconciles and the 0-trial unscreened remainder is stated as a number rather than omitted. |
| P2_k_cascade | HELD | k at every stage: surfaced 55, located 52, experimental 48, comparator 3, included 2, unscreened remainder 0. |
| P3_inclusion_criteria | HELD | Criteria provenance block present with predefined=False on its face (state 'DERIVED_POST_HOC'). |
| P4_preconditions | HELD | All 8 recorded with verdict and cited authority: 2 FAIL, 1 NOT-ASSESSABLE. criteria_predefined is FAIL -- post hoc criteria, which R107 permits and C5/C7 does not satisfy. |
| P5_extraction_table | HELD | 5 cells: 2 READ with source path and verbatim text, 3 DERIVED with the method named. |
| P6_analysis_output | REFUSING | No quotable model output exists: k=2 and nothing was pooled. The absence is recorded as a finding with its cause and its trigger. |
| P7_published_comparison | REFUSING | No denominator yet. Blocked on external registration-id resolution; citation-string matching explicitly refused rather than substituted. |
| P8_registration_identity | HELD | 2 of 2 trial(s) verified live against the registry. |
| P18_restatement_is_reproducible | HELD | 1 restatement block(s), each naming a script that resolves on disk: regated_2026_08_19_trailing_placebo -> scripts/regate_across_revisions.py, scripts/regate_cascade_2026_08_19.py |
| P19_promotion_reaches_derived_blocks | HELD | inputs.trials, k_cascade.k_included_in_object and prisma_flow.included all say 2; no first-person field in published_comparison declares a k the pooled outcome(s) do not. |
| P20_cascade_reconciles | HELD | k2_role_located=52 == k3+k4+k5, and k0_surfaced=55 == k3+k4+k5+kNA. Both checked, and every failing limb would be reported rather than the first. |
| P21_ambiguous_question_split | REFUSING | This object records no split provenance. That is CORRECT for a topic whose question was never ambiguous and it is NOT a pass: this builder cannot tell the two apart, so the property is refused with the reason rather than held on an absence. What would settle it: a `split_provenance` block, or an explicit statement that the question admits one reading. |
| P22_sharing_recorded | HELD | No trial of this review appears in any other object's included set. COMPUTED over the corpus, not asserted: 2 trial(s) checked against every other topic object. |
| P23_recall_measured | HELD | All 1 recorded quer(ies) state their recall against this review's own included set: 2/2. |
| stage | k |
|---|---|
| k0_surfaced | 55 |
| k2_role_located | 52 |
| k3_experimental | 48 |
| k4_comparator | 3 |
| k5_background | 1 |
| kNA_not_assessable | 3 |
| k_included_in_object | 2 |
| k_unscreened_remainder | 0 |
| remainder_dispositions | EXCLUDED 31, ELIGIBLE NOT POOLABLE 2, ELIGIBLE NO RESULTS YET 13, what this says READ BY THE REMAINDER-DIAGNOSIS RULE: 31 of 46 fail a criterion, and the dominant ground is COMPARATOR -- only 6 of the 46 declare a placebo arm at all. This is a drug programme dominated by OPEN-LABEL EXTENSION and single-arm studies, so the query is well aimed and the randomised evidence is genuinely thin. Neither the too-broad-query shape nor the fragmented-estimand shape: a third reading, where the field itself has produced few controlled trials., no poolable candidates ZERO trials in the remainder are eligible AND poolable -- unlike alirocumab-lipid, where screening recovered two. Stated explicitly because a zero here is a real result about this evidence base rather than an absence of effort. |
| why_k_included_is_2_and_stays_2 | THE POOLING OBSTACLE IS THE ESTIMAND, NOT THE EVIDENCE BASE. Both included trials register HIERARCHICAL endpoints analysed by win ratio, and the two hierarchies are not the same hierarchy: ATTR-ACT combines all-cause mortality with CV-related hospitalisation frequency (2 tiers); ATTRibute-CM adds NT-proBNP change and 6-minute-walk change (4 tiers). Verified from the registry 2026-08-19. A win ratio's estimand IS its hierarchy and cannot be recovered from a 2x2 table. TWO DIFFERENT HIERARCHIES ARE TWO ESTIMANDS, NOT HETEROGENEITY -- so k=2 is the count of included trials and NOT a pooled k. |
| keyed_on | registration id |
| regated_2026_08_19_trailing_placebo | THE ONLY ONE OF THE FIVE GATED TOPICS WHOSE CASCADE DID NOT MOVE. Re-run against the repaired classifier with the old code loaded from git at 7a08bcbe1 and the surfaced set re-executed (55 today, 55 stored): ZERO of the 55 records changed role. Recorded as a measured null rather than left silent, because 'unchanged' and 'not checked' look identical on a finished page. k2_role_located was separately corrected from 55 to 52 -- an arithmetic defect in the stage label, not a movement. RE-DERIVED BY scripts/regate_cascade_2026_08_19.py (old vs new) and by scripts/regate_across_revisions.py, which walks EVERY commit that touched the classifier and reports at which of them the stored numbers reproduce: 48/3/1/3 appears first at f2bf16022 and is unchanged at every revision since, so this topic is current rather than merely unexamined. |
predefined: False | post_hoc: True — derived —
—
| limb | value | derived from |
|---|---|---|
| — | — | — |
| — | — | — |
| — | — | — |
| — | — | — |
—
k after screening: —. —
| failing limb | n | registrations |
|---|
—
—
| registration | rank | matching outcome |
|---|
| registration | verdict | limb | reason |
|---|
k=2 and nothing was pooled, so there is NO model call to quote, NO pooled estimate, NO heterogeneity and NO package version. This is not a missing artefact; it is the correct state for a topic with no synthesis.
What stands instead: —
Provenance: ——
Heterogeneity: Undefined: no between-study variance was estimated because no model was run.
What would change it: An estimand shared by two or more included trials. Where k is already above 1 and nothing pooled, the obstacle is the ESTIMAND rather than the evidence base, and this field must say which.
Denominator: —. Not yet established. Establishing it requires resolving published syntheses' included-sets to REGISTRATION IDs. A second lane has validated that pipeline (stated-k table identification plus two-hop citation->PMID->registration) and is running it now.
Explicitly not done: CITATION-STRING MATCHING. It would produce a comparison-shaped answer tonight and it is not identity: two syntheses citing the same trial under different labels would read as different, and the same label on different trials as the same. That is the substitution class this project audits for.
Blocked on: lane 2 registration-id resolution
Authority: Cochrane Handbook for Systematic Reviews of Interventions 6.5, 2024, verified 2026-08-19. Publishable: True.
| precondition | verdict | reason | authority |
|---|---|---|---|
| population_stated | PASS | present: question carries 118 characters | MECIR C62 'Ensuring meta-analyses are meaningful' (Mandatory, READ 2026-08-19): 'Undertake (or display) a meta-analysis only if participants, interventions, comparisons and outcomes are judged to be s |
| arm_role_resolved | PASS | inputs.trials: all 2 trial(s) carry arms with a readable role | MECIR C62 (Mandatory), 'interventions' limb -- this is a PRECONDITION FOR ASSESSING C62, not an enforcement of it: you cannot judge whether interventions are 'sufficiently similar' without knowing whi |
| comparators_identified | PASS | every outcome names a comparator type ['placebo']; none declares 'mixed'. Within-outcome consistency is visible only through the declared type, and cross-outcome differences are not judged. | MECIR C62 (Mandatory), 'comparisons' limb -- a synthesis pools one contrast. NOT cited to C7: C7 governs PREDEFINITION, and this check reads recorded comparator fields and tests them for consistency. |
| estimand_named | PASS | outcomes.estimand: all 1 outcome(s) name a quantity | MECIR C8 'Clarifying role of outcomes' (Mandatory, READ 2026-08-19); MECIR C62 (Mandatory), 'outcomes' limb |
| criteria_stated | FAIL | declared absent: screening.eligibility states its own absence: 'not recorded on the page this object was built from' | MECIR R29 'Eligibility criteria for types of participants', R30 '...types of interventions', R31 'Role of outcomes' -- ALL Mandatory REPORTING standards (READ 2026-08-19). Their verb is STATE. A block |
| criteria_predefined | FAIL | screening.eligibility_provenance.predefined is false: the criteria were DERIVED post hoc from this object's own question and recorded exclusions. That is auditable (R29/R30/R31) and is not pre-specification (C5/C7). This | MECIR C5 'Predefining unambiguous criteria for participants' AND C7 '...for interventions and comparators', BOTH Mandatory (READ 2026-08-19): 'Predefined, unambiguous eligibility criteria are a fundam |
| eligibility_met | NOT-ASSESSABLE | cannot assess: criteria are not stated (screening.eligibility states its own absence: 'not recorded on the page this object was built from'), so whether any trial met them cannot be decided | MECIR C5/C7 (Mandatory) with C6 'Predefining a strategy for studies with a subset of eligible participants' (Highly desirable, READ 2026-08-19) -- eligibility is assessed against the full report |
| contributes_a_randomised_contrast | PASS | inputs.trials.arms: all 2 readable trial(s) contribute one topic-vs-control randomised contrast | Handbook 6.5 s23.3.4 'How to include multiple groups from one study' (READ 2026-08-19) and s23.3.6 'Factorial trials'; MECIR C7 'Predefining unambiguous criteria for interventions and comparators' (Ma |
This page is stamped to the standard version above. If the standard rises, this page is below it and knowably so. No page is grandfathered; arni-hfref is presently unstamped and is therefore unknown-version, not compliant.
Every number on this page is projected from a single canonical object, and each measured cell names the source analysis it was read from. That is what is machine-checked. It does not establish that the underlying sources are right, only that this page faithfully reports them.