Every research evidence label on WikiRemedy is assigned by a human reviewer — not generated by AI or scraped automatically. We read the studies and make a judgement.
Our editorial team reviews published research from the following databases and institutions before assigning an evidence level to any remedy.
The US National Institutes of Health database — the largest freely accessible archive of biomedical literature. Our primary starting point for any evidence review.
Systematic reviews and meta-analyses that synthesise evidence from multiple studies. Cochrane reviews are considered the gold standard for clinical evidence.
Clinical reference material from Mayo Clinic is used to cross-check evidence levels, particularly for pharmaceutical and lifestyle remedies.
Studies from journals including The Lancet, BMJ, JAMA, and specialist publications for natural health, nutrition, and integrative medicine.
We do not use a formula. Our research advisors read the available literature and make a judgement based on study quality, sample size, consistency of findings, and replication. A single small trial does not qualify as "Moderate" — it takes multiple consistent findings.
If we cannot find published studies for a remedy, we assign "No evidence" — which means no studies found, not that the remedy doesn't work. Many traditional remedies pre-date modern clinical trial culture by centuries. Community experience fills that gap on WikiRemedy.
Research evolves. A remedy labelled "Limited" today may qualify for "Moderate" next year if new studies are published. Our research advisors review evidence labels on a rolling basis — prioritising remedies with active community engagement and new research.