How we rate evidence
Not all health claims are equal. Here is how we assess the claims made in the videos we summarise.
Why evidence ratings matter
A YouTube creator can state almost anything with confidence. Our job is to give you context: is this claim backed by solid science, mixed findings, or mostly anecdote? Without that context, it is hard to know what to act on.
Every claim we extract from a video is assigned one of five evidence ratings. These ratings reflect our best reading of the available literature at time of review, not a comprehensive systematic review. They are a starting point for your own thinking — not a substitute for professional medical advice.
What we look at
When rating a claim, we consider:
- Whether multiple high-quality randomised controlled trials (RCTs) or meta-analyses support the claim
- Whether findings replicate across different populations and study designs
- Whether clinical guidelines from major health bodies (NHS, AHA, ADA, etc.) endorse the claim
- The recency and relevance of the studies cited (if any) by the creator
- Whether the claim has plausible biological mechanisms or is implausible based on current understanding
Our limitations
Our evidence ratings are AI-assisted, not produced by clinicians. For high-risk topics — testosterone, mental health, supplements, fertility, prostate health, erectile dysfunction — ratings go through an additional human review step before publishing.
Science evolves. A claim rated “mixed” today may be re-rated as “strong” (or “not supported”) in future updates. We aim to refresh ratings periodically, but we cannot guarantee every rating reflects the very latest research.
If you believe a rating is wrong, please contact us.
How our evidence labels work
Every claim we review gets one of five evidence ratings.
Supported by multiple high-quality human studies or clinical guidelines.
Supported, but still context-dependent or limited to certain populations.
Promising but uncertain. Results vary across studies or populations.
Mostly anecdotal, small studies, animal studies, or influencer claims.
Contradicted by existing evidence or scientifically implausible.
We have not yet reviewed the claims in this video.