Discover
Scientific literature, trial registries, official statistics, regulatory resources and explicitly labelled discovery leads are collected according to source role.
Methodology
Collection is not synthesis.
The observatory separates discovery, data capture, classification, human review and conclusion-sensitive synthesis. That separation is a safety and auditability feature, not a presentation detail.
Validate
Records are normalised, deduplicated, quarantined where relevance or quality rules fail, and retained with provenance rather than silently disappearing.
Classify
Automated metadata classification supports triage and navigation. It does not itself establish effect, certainty, causality or safety.
Review
Comparator, population, outcome, effect estimate, confounding, funding, conflicts and risk-of-bias fields require source-grounded human review before synthesis.
Current readiness: the latest pre-migration public release contained 986 candidate records, 0 human-reviewed records and 0 conclusion-sensitive synthesis-ready records. Those are operational readiness facts, not an evidence-certainty rating.
Publication boundary
The public Observatory is intentionally restricted to publication-safe aggregate and audit metadata. Raw records, news-lead content, retailer/product-level investigative findings and private matching procedures are excluded from the public release.
Migration provenance
SCC Vaping succeeds the Vaping26 public observatory. Legacy technical identifiers can remain in migrated datasets where changing them would weaken provenance, schemas, tests or reproducibility. Public-facing branding changes do not rewrite historical evidence artefacts.