Editorial policy

Which sources we use, in order

We cite in this order of preference, and stop as high up the list as the claim allows.

• Government and regulatory sources. National health authorities, official gazettes, ministry publications.
• Recognised professional medical bodies. Specialty societies and royal colleges.
• Peer-reviewed medical literature. Preferring systematic reviews and randomised trials over single-centre series, and always reporting the certainty the reviewers themselves assigned.
• University and major hospital patient information.
• Other reputable medical publications.

Every source in our library was opened by hand, confirmed to be published by an authority, a journal or a professional society, and confirmed to return a working page directly rather than through a redirect. A commercial page selling the thing being described is not a source, however confidently it is written.

What medical review means here

It means a clinician has read the page and confirmed that what it says about anatomy, technique, recovery and risk is accurate and is not overstated. Advice about your own nose comes from a consultation, where somebody can examine it.

A page is labelled medically reviewed only where a named clinician actually read it. Where none has, the byline says "Editorially reviewed, not yet read by a clinician" and names the editor instead. We would rather carry that label on a page than dilute the other one.

Reviewers are named, and the specialty they trained in is stated on their profile. See medical review and conflicts of interest, which is the more important of the two.

Two dates on every page, never one

Published, and last reviewed. Evergreen medical content is close to meaningless with only a publication date: a page written two years ago and re-read by a clinician last month is more current than one published last month and never checked since. Our guides index sorts on the review date for the same reason.

Clinical pages are re-read on a twelve month cycle. Every dated stamp on this site is a commitment to future work, and each one is entered in a review calendar on the day it is published.

The standards every page has to meet

• Every figure carries its source, with the population it was measured in, the timepoint and the comparator. A number without those is a number about nothing.
• A success rate is quoted only where a named study measured one, and it is quoted with the study, the population and the certainty the reviewers themselves assigned.
• Costs are explained rather than quoted. The cost pages teach you to read a written quote line by line, because that itemised number for your own nose is the one that actually decides anything.
• Photography here is illustrative and consented. Results are shown to you by the surgeon, in a consultation, where you can ask how many cases the set was chosen from.
• Every claim is traceable. Where a claim needs a citation we do not have, the claim is cut rather than softened.
• Coverage stays on rhinoplasty and the people who perform it. Named private individuals are written about only where they have published about themselves.

Where AI is used, and where it is not

Drafting and condensing on this site are assisted by AI tools. Every clinical page produced that way is then checked line by line against its sources by a person, and read by a clinician before publication. No page is published on the strength of a model output alone.

AI is never used to generate a citation, a statistic, a study, a quotation or a surgeon record. Those are the specific things language models invent most convincingly, and every one of them is checked against the primary source by hand.

User-generated content

Everything published here is written and edited by the named people on it. Comments and reviews on a medical page create moderation duties that have to be resourced properly to be discharged properly, and a small team that moderates badly does more harm than good.