From a database of millions to one honest page

1. Finding

Each topic has a fixed search, anchored on the medical subject headings that the National Library of Medicine assigns. Only publications that carry evidence get through: randomised trials, clinical trials, cohort studies, systematic reviews and meta-analyses. Review articles are left out — they retell rather than show.

So is research that is not about you. Studies whose title says they are about children, pregnant women, students or athletes are excluded, as are patient groups unrelated to the question.

2. Grading

Every study gets a number from 1 to 5. One rule comes before all others: who it was studied in weighs more than how it was studied. A meta-analysis of mouse experiments is still about mice, so it lands on level 2 and not level 5. That distinction is exactly where this market makes its money.

3. Selecting

Not the most recent two hundred studies — those are mostly cell cultures and mice. What the archive carries is the state of the science: the landmark studies, the meta-analyses, and the recent work that genuinely matters. Weight of evidence, how much a study was built on, and how recent it is, all count.

Near-identical studies are held back. Six meta-analyses on daily steps and mortality are one answer, not six pages.

4. Writing

Each study gets a plain-language write-up in two layers. The verdict — what it examined, how strong the evidence is, what it means — is free for everyone. The full write-up is for subscribers.

The free layer is genuinely free: it is in the page for anyone and for search engines. The subscriber layer is left out of the page entirely when you have no access, rather than hidden with styling. That is a deliberate difference.