Integration map — propagation of the surrogate-endpoint validity review

This internal file maps each section of surrogate-validity-review.md to its landing site in the audit-visible technical file. It exists so the internal team can track which parts of the review have been propagated and where. The map is internal only; no audit-visible document references this file.

Propagation targets — what lands where

Review section	Target audit-visible document	Target location
§2 + §3 (Regulatory framework + Methodology)	R-TF-015-011 State of the Art	New section: "Surrogate endpoint validity — review methodology"
§4 body (Diagnostic accuracy / benefit 7GH)	R-TF-015-011 State of the Art	New subsection: "Surrogate endpoint validity — diagnostic accuracy"
§5 body (Severity scoring / benefit 5RB)	R-TF-015-011 State of the Art	New subsection: "Surrogate endpoint validity — severity scoring"
§6 body (Referral optimisation / benefit 3KX)	R-TF-015-011 State of the Art	New subsection: "Surrogate endpoint validity — referral optimisation"
§7 (Cross-domain synthesis / causal pathway)	R-TF-015-003 Clinical Evaluation Report	Pillar 1 VCA subsection — expanded narrative with causal-pathway diagram / table
§8 (Limitations and residual uncertainty)	R-TF-015-003 Clinical Evaluation Report	Existing limitations subsection — cross-referenced; PMCF commitments flagged
§9 (Conclusion)	R-TF-015-001 Clinical Evaluation Plan	Evidence-hierarchy table — Pillar 1 row strengthened with reference to the new SotA section
§10 (Reference list)	R-TF-015-011 State of the Art	Appended to existing SotA reference list (deduplicated)

BSI round-1 response propagation

• Item 3b response (docs/bsi-non-conformities/clinical-review/round-1/item-3-clinical-data/rb-data-sufficiency-justification/response.mdx): a paragraph cites the new SotA section by name as the closure of the Horiana §2.1.3 bullet 1 mitigation. Reviewer-friendly language — no engineering leaks, no reference to the internal task folder.
• Round-1 kanban (docs/bsi-non-conformities/clinical-review/round-1/index.mdx): new card i3b_surrogate_literature_review advances received → analysed → answered → sent as the propagation progresses.

Status tracking

Actual propagation was delivered per the Option B minimum-viable rescope documented in do-we-need-this.md on 2026-04-21. The scope decision rationale: the task as originally written (≥ 20-reference full structured review) would re-curate ~80 % of evidence already in R-TF-015-011; only the ~20 % genuinely-new anchoring bodies (stage-at-detection → survival, regulator-accepted severity endpoints, care-pathway outcome equivalence) were added to the audit-visible technical file. The internal reference folder (32 appraised references) is retained in full as a Round-2 defensive asset; it is not propagated into audit-visible docs.

Work-stream	Status
References — Domain 1 (diagnostic accuracy)	13 internal reference files — complete (11 of which are also in the primary SotA corpus; 2 genuinely new: Gershenwald 2017, Conic 2018)
References — Domain 2 (severity scoring)	10 internal reference files — complete (7 genuinely new: EMA 2004, Schmitt 2014 HOME IV, Simpson 2016, King 2022, Olsen 2004, Mattei 2014, Mrowietz 2011)
References — Domain 3 (referral optimisation)	9 internal reference files — complete (4 genuinely new: Whited 2013, Armstrong 2018, Moreno-Ramirez 2007, Snoswell 2016)
appraisal-log.md rolling CRIT table	complete
surrogate-validity-review.md §1–§10 draft	complete (retained as Round-2 asset; not propagated under Option B)
Propagation into R-TF-015-011 SotA §"Surrogate endpoint validity"	complete — 13 anchoring references; three subsections (per-benefit); cross-domain synthesis; CRIT1-7 appraisal table; AMA-style reference list
Propagation into R-TF-015-003 CER Pillar 1 row + causal-pathway subsection	complete — VCA row expanded to reference SotA anchoring section; new subsection §"Causal pathway and clinical meaningfulness of the selected endpoints" per Class IIb indirect-benefit scrutiny
Propagation into R-TF-015-001 CEP evidence-hierarchy planned source	complete — VCA planned source updated; benefit 3KX surrogate-to-outcome chain citations aligned with SotA anchoring references
Item 3b response.mdx §6 Pillar 1 VCA surrogate-endpoint anchoring	complete — new section 6 inserted; sections 7 and 8 renumbered
Round-1 kanban i3b_surrogate_literature_review card	complete — added to review column with rescope rationale
Build verification	pending (run npm run build:qms)
audit-deliverable-reviewer pass	pending
bsi-clinical-auditor pass	pending
celine-clinical-consultant pass	pending

Wording boundaries (per QMS audit boundary and docs/bsi-non-conformities/CLAUDE.md)

• Internal task folder path (task-3b6-...) must not appear in any audit-visible document or in response.mdx.
• No references to Horiana by name in audit-visible documents; use regulator-level phrasing ("external methodological expert advice received on 2026-04-17").
• BSI ticket codes (T0088560, M1.1, round-1) do not appear in audit-visible documents; they are internal organisational artefacts.
• Company and device brand names follow QMS CLAUDE.md rules — "we" / "the device" — except in the CEP / SotA Stream-B manufacturer-specific literature-search register (MEDDEV 2.7/1 Rev 4 Annex A5 reproducibility exception already documented in apps/qms/CLAUDE.md).
• Reviewer-friendly language in response.mdx: no engineering-leak references (no .mdx paths, no component names, no commits). Instead: "the SotA document has been expanded with a new section on surrogate endpoint validity (red-lined version provided)".

Red-lined-version narrative for BSI

BSI expects red-lined versions for any documentation change. Since the technical file is MDX-rendered, the practice already agreed with Maggie in previous rounds is to describe what changed and where in response.mdx, referencing specific document names and section headings. This same approach applies to the SotA additions and the CER / CEP expansions.