Where our information comes from
Evidence Health doesn’t generate medical opinions. It draws on authoritative sources, ranks them by how relevant and reliable they are to your question, and cites the exact one used.
How sources are ranked
Not all sources carry equal weight, and the right source depends on the question. We prioritise them in tiers:
Tier 1 — Regulators & official labels. Medicine product information, warnings, recalls and regulatory status. e.g. the U.S. FDA (openFDA), DailyMed.
Tier 2 — Clinical guidelines. Diagnosis and treatment pathways, standard of care. e.g. national health authorities and professional societies.
Tier 3 — Evidence syntheses. Systematic reviews and meta-analyses — the best synthesis of trial evidence and its uncertainty.
Tier 4 — Primary research. Individual studies, for recent or specialised evidence not yet in guidelines. e.g. PubMed-indexed trials.
Tier 5 — Trusted patient information. Plain-language explanations and patient education, used to explain — never to overrule higher-order clinical evidence. e.g. MedlinePlus.
Integrated now Live
- FDA drug labels (openFDA): indications, warnings, boxed warnings, interactions, side effects.
- RxNorm (NLM): canonical medicine identity, so brands, generics and codes resolve to one medicine.
- openFDA recalls & adverse-event reports: safety signals and recall notices.
Planned Soon
- MedlinePlus for patient-friendly explanations, PubMed for research, WHO ICD for conditions.
- A licensed drug-interaction database for reliable two-medicine checks.
One important caveat
Reported-reaction data (adverse-event reports) shows what people have reported to regulators. It is a signal, not a measure of how common a side effect is, and it does not prove a medicine caused the event. We always label it as such.
We store each source’s version and the date we retrieved it, so an answer can always be traced back to exactly what the source said at the time. See how that works →