Every item carries an evidence label. Labels come from the publication types and study-subject tags that PubMed and Europe PMC assign, and from the ClinicalTrials.gov record for registered trials. Where a record is too new to be tagged yet, we label it from its title.
Human RCT / meta-analysisRandomized controlled trials (RCTs) in people, or meta-analyses that pool earlier studies. The strongest designs, still not proof on their own.
Human study, non-randomizedStudies in people without randomization: open-label, observational, post-hoc analyses, case reports.
ReviewReviews and systematic reviews that summarize earlier work. No new data.
Registered trialA ClinicalTrials.gov record of a planned or running trial. A registration is not a result.
Animal / in-vitroStudies in animals, cells or other lab models. Findings often do not carry over to humans.
PreprintShared before peer review. Provisional.
Editorials, letters, and third-party videos. Expert opinion, not new data.
Forum posts, personal reports and social media. Defined here for completeness; the Radar never lists them.
Titles and bibliographic details are shown only to identify each source; all rights remain with the publishers, authors and creators. We link to the original and never host articles, abstracts, figures, PDFs or video thumbnails. Rights-holders and authors can request a correction or removal at support@nstresearch.com.