Methodology

How we rate evidence

We separate primary evidence from AI interpretation and show the confidence of each source. Ratings are designed to help users prioritize evidence, not to replace clinical judgment.

Alternative therapies

We rate alternative therapies using four components: human evidence, safety, clinician support, and source reliability.

Evidence tiers are patient‑friendly versions of evidence levels used by authoritative sources.

  • • Tier A: strong human evidence
  • • Tier B: some human evidence
  • • Tier C: limited/low‑quality human evidence
  • • Tier D: no human evidence

Safety tiers: Low risk, Use with caution, High risk/avoid. If there is no human evidence and material safety concerns, we cap the rating and display a prominent warning.

Trials

Trial ratings combine relevance (match to user search terms) and a baseline internal score based on trial status, recency, and data completeness. Active and recruiting trials are prioritized, and stale or incomplete records score lower.

Papers

Paper ratings prioritize recency, study type, source reliability, and whether the publication is peer‑reviewed. Preprints are labeled and typically score lower than peer‑reviewed literature.

Transparency

We label evidence clearly and always link to primary sources. AI summaries can be incomplete or wrong, so please verify with the original record.