Source standard / Review method
The source comes before the summary.
We build each factual item from an authoritative record, then keep the study frame and limitations visible.
Last reviewed July 21, 2026Review sequence
Define the exact claim or reading problem before searching.
Prefer regulatory records, trial registries, primary studies, and systematic reviews.
Record model or population, design, comparator, sample size, duration, and endpoint.
Describe what investigators measured and reported — not what a reader should expect.
Name uncertainty, translation gaps, conflicts, status, and what remains unestablished.
A human checks the exact source, wording, asset, caption, and link path before publication.
Evidence hierarchy
Regulatory sources and official study records establish status and registered design. Systematic reviews map an evidence landscape. Controlled human studies answer population- and endpoint-specific questions. Observational, animal, cell, and computational work can be informative without establishing a human outcome.
Authoritative starting points
Important distinction
A registry listing is not proof of benefit. PubMed indexing is not a quality guarantee. A mechanism is not a clinical outcome.