Scientific Claim Registry (SCR)

A registry for scientific claims — it accumulates evidence, it doesn't arbitrate it.

Papers have DOIs. Researchers have ORCIDs. Clinical trials have registration numbers. Scientific claims have no adopted public registry. SCR gives each claim a persistent ID and accumulates — automatically, over time — the evidence that supports, contradicts, or refines it.

What a registered claim carries

Deliberately minimal. Human-curated first. Reuses existing standards (GRADE) instead of inventing new scores.

Persistent identifier

A permanent, citable ID — e.g. SCR-LIP-000001 — that endures as evidence accumulates.

Canonical statement + context

Population, condition, exposure, comparator, outcome and methodological scope. A claim is meaningless stripped of context.

Supporting & contradicting evidence

Linked at the level of individual studies, with design and risk-of-bias — not just citation counts.

Evidence certainty (GRADE)

Reuses the international standard rather than a novel confidence metric.

Separated dimensions

Evidence confidence, expert consensus and knowledge state are tracked separately. Consensus is not truth.

Version history

Every revision is recorded — author, timestamp, rationale, evidence added or removed. Knowledge that can answer "what changed?"

A claim, made into an object

A real entry from the lipedema pilot registry.

SCR-LIP-000015 clinical association

Women meeting lipedema screening criteria have a higher prevalence of positive ADHD self-report (ASRS-18) than women without lipedema (76.9% vs 54%; RR 1.424).

Knowledge stateEmerging

Evidence certaintyLow (GRADE)

Consensusnot yet assessed

Evidence: 1 cross-sectional study, n=354 — Amato et al., 2023 · DOI:10.7759/cureus.35570

Gaps: self-reported screening; convenience sampling; no confounder adjustment; single study, not replicated.

Honest by design: the registry shows what is uncertain, not only what is known.

Three layers — and only the first must exist

The durable asset is the registry. It registers; it does not arbitrate truth. Consensus and clinical recommendation are optional layers above — a future AI could even generate consensus on demand by reading the registry.

Layer 1 · the product

Registry

Automated. Each claim gets a persistent ID; the system accumulates the evidence that supports, contradicts, or refines it, with full history. No judgment.

Layer 2 · optional

Consensus

Experts may endorse, dissent, or qualify a claim. Useful — but not required for the registry to exist.

Layer 3 · optional

Recommendation

Medical societies turn claims into clinical guidance. A separate, human, accountable process.

The founding question

Publications, researchers, and clinical trials all have adopted, persistent registries. Why don't the scientific claims themselves?

SCR is not a replacement for journals, systematic reviews or peer review. It is the adopted, domain-organized index of claims — and, increasingly, the structured, machine-readable substrate that scientific AI can consult and cite.

Built on existing work — honestly

SCR does not invent claim identity. It is the adopted, domain-organized index that accumulates evidence per claim over time.

Pilot domain: lipedema

Chosen not because it is easy, but because it is hard — a rapidly expanding, fragmented literature with active controversies, multidisciplinary stakes, a diagnostic delay measured in years, and an engaged international community. If claim identifiers prove their worth here, the framework is disease-agnostic by design.

50 founding claims · SCR-LIP-000001 … 000050