Ingredient Framework

Evidence tiers we use

We categorise source strength into four pragmatic tiers to support consistent interpretation.

1. Tier A, Regulatory & systematic reviews: National or regulatory agency assessments (e.g., SCCS, FDA summaries), systematic reviews, and meta-analyses.
2. Tier B, Peer-reviewed clinical and toxicology studies: Human clinical studies and animal studies using established dermal exposure models.
3. Tier C, In vitro, mechanistic, or case reports: Laboratory assays, mechanistic studies, and well-documented clinical case reports.
4. Tier D, Industry data, preprints, or non-peer-reviewed sources: Used cautiously and explicitly labelled; considered when higher-tier evidence is limited or unavailable.

Note: We prioritise higher-tier evidence (Tier A–B) where available; in many cases, Tier B forms the primary working evidence, with Tier C–D used to provide supporting context.

Scoring & flagging rules (short)

For consumer clarity we apply three visible flags on ingredient pages: No Common Risk Signals, Heightened Context Sensitivity, and Evidence Evolving. The internal scoring is a structured, interpretive model combining four weighted components to support consistent evaluation. It is not a clinical or regulatory scoring system.

score >70 = No Common Risk Signals; 40–70 = Heightened Context Sensitivity; <40 = Evidence Evolving.

Step-by-step evaluation (applied method)

1. Collect INCI and synonyms. We compile the standard INCI name plus common trade names and CAS numbers to avoid ambiguity.
2. Identify concentration data. Search regulatory dossiers, publicly available safety data sheets and regulatory dossiers, and product labels for typical concentration ranges.
3. Gather evidence. Retrieve Tier A–C literature, regulatory opinions, and clinical reports relevant to dermal exposure.
4. Assess exposure context. Evaluate typical user patterns: leave-on vs rinse-off, frequency, higher-sensitivity populations.
5. Score components and compute flag. Apply the weighted matrix above and generate a preliminary flag.
6. Peer review & publish rationale. A second reviewer within the project validates the evidence selection and the final text includes linked sources and a changelog.

Every ingredient page contains a short "how we decided" box summarising these steps and linking to the primary sources we used.

Limits & common caveats

No framework can remove all uncertainty. Common limitations include:

• Incomplete concentration data: Many consumer products do not publish exact ingredient concentrations, we therefore use ranges and conservative assumptions where necessary.
• Population diversity: Individual sensitivity varies, an ingredient flagged "No Common Risk Signals" may still cause allergy in a small proportion of people.
• Rapidly evolving evidence: New studies or regulatory decisions can change conclusions; hence the "last reviewed" date and changelog are central to our transparency.

This framework provides consumer-facing educational context and is not a substitute for clinical diagnosis or professional toxicology consultation where required.

Reproducibility & transparency

To ensure reproducibility:

• We publish a list of primary references for each ingredient entry (with direct links where publicly available).
• We provide a short changelog for substantive updates explaining the evidence that prompted the change.
• Internal scoring worksheets are archived and available on request to qualified researchers where feasible (contact us for access).

This openness is designed to allow external experts to verify our process and suggest improvements.

Corrections, disputes, and appeals

If you believe an ingredient entry is incorrect, please submit a correction via our contact page with supporting evidence. We investigate disputes using the same tiered evidence approach and publish our findings and any resulting changes.