Evidence policy

A citation counts only when it supports the exact claim.

Evidence levels must stay separate

An ingredient study is evidence about the tested ingredient, form, dose, population and outcome. It is not automatically evidence that a multi-ingredient UriSmooth formula produces the same result. A certificate of analysis can support specified quality facts; it does not prove efficacy. Manufacturing documentation can support a manufacturing statement; it does not prove health outcomes.

Preferred source order

  1. Binding rules, official drug labels and primary product records.
  2. Current government or professional clinical guidance.
  3. Relevant systematic reviews and meta-analyses.
  4. Randomized controlled human trials.
  5. Other human evidence.
  6. Mechanistic, animal and laboratory evidence for background only.

Manufacturer pages can verify current product facts but are not treated as independent clinical proof. Search snippets, competitor pages, testimonials and AI answers are leads—not final substantiation.

Directness questions

  • Does the exact ingredient form match?
  • Is the studied dose comparable?
  • Is the population relevant?
  • Is the measured outcome the one described?
  • Was the finished product tested or only one component?
  • Is there stronger or contradictory evidence?

Contrary evidence

Material contrary evidence must appear in the discussion. We do not select one favorable study and hide a stronger synthesis. When the evidence is uncertain or mixed, the wording must say so plainly.

Prohibited substitutions

Testimonials cannot replace trials. Ingredient presence cannot replace effective dose. A standard supplement disclaimer cannot turn a disease-treatment claim into a permitted claim. The number of citations cannot compensate for poor relevance.

Policy published: September 21, 2026.