Editorial Policy

Our Sources

Vellito exclusively covers research from established, peer-reviewed sources. Our content is drawn from:

• PubMed / MEDLINE — peer-reviewed biomedical literature from the National Library of Medicine
• ClinicalTrials.gov — registered clinical trial results from the U.S. National Institutes of Health
• FDA — drug approvals, safety communications, and regulatory decisions
• European Medicines Agency (EMA) — European regulatory actions and assessments
• CDC MMWR — Morbidity and Mortality Weekly Reports from the Centers for Disease Control
• medRxiv — preprints in health sciences (clearly labeled as non-peer-reviewed)
• Cochrane Library — systematic reviews and meta-analyses
• Frontiers — open-access peer-reviewed research
• NIH News — research findings from the National Institutes of Health

How Our Summaries Are Made

Summaries on Vellito are generated by an automated AI pipeline working directly from the primary source — the peer-reviewed publication, trial registration, or regulatory announcement. We label this clearly on every article because readers deserve to know what they are reading.

Before anything is published, every draft passes a multi-stage automated accuracy review:

• Deterministic fact-check — every number, percentage, p-value, sample size, and drug name in the summary must appear verbatim in the source text, or the draft is rejected.
• Independent AI review — a second, separate model scores each draft 1–10 for factual fidelity against the source; drafts scoring below 8, or with any critical error, are held back from publication.
• Daily independent audit — a random sample of published articles is re-reviewed every day by a third, unaffiliated model; flagged articles are pulled back for correction or re-generation.

Editorial Oversight

Vellito's editorial standards — what we cover, how findings may be characterized, what language is banned, and when content must be corrected or removed — are set and supervised by our editorial team of medical writers with physician and PhD backgrounds. Editors audit published content, review articles flagged by our automated systems, and are responsible for the corrections process below.

Individual articles are not pre-reviewed by a human before publication. The named editor on each article identifies who is responsible for editorial oversight of that specialty area, not a per-article reviewer.

Quality Scoring

Articles are scored on evidence strength, methodology, source prestige, and clinical relevance. Only articles meeting our quality threshold are published. Borderline drafts are held out of publication entirely — they are re-generated, or in some cases manually reviewed in our editorial queue, before they can go live.

Dual-Audience Summaries

We produce two versions of every summary:

• Healthcare Professional (HCP) summaries — detailed clinical analysis including methodology, statistical significance, confidence intervals, number needed to treat (NNT), and implications for clinical practice
• Patient summaries — accessible, plain-language explanations of what the research means, written for a general audience without medical training

Corrections Policy

We take accuracy seriously. If an error is identified in any published summary:

1. The article is immediately flagged for review
2. A corrected version is published with a note indicating the nature of the correction
3. The "Last Updated" date is revised to reflect the correction

To report an error, please email [email protected] with the article URL and a description of the issue.

Editorial Independence

Vellito does not accept sponsored content and has no financial relationships with pharmaceutical companies, device manufacturers, or healthcare organizations that could influence our editorial decisions. Our coverage is determined solely by the scientific merit and clinical relevance of the research.

Disclaimer

Vellito provides summaries of published medical research for informational and educational purposes only. Our content is not medical advice. Healthcare decisions should be made in consultation with a qualified healthcare provider based on individual circumstances. Always refer to the original peer-reviewed publication for complete study details.