Section 01Sourcing
Every protocol carries three to five sources and they are institution level. In practice that means NICE guidelines and Clinical Knowledge Summaries, the NHS, the British Association of Dermatologists, the Medicines and Healthcare products Regulatory Agency, the Cochrane Library, the indexed literature on PubMed, and UK government bodies where regulation is relevant.
We do not cite retailers, brand owned research pages, or publications whose business model is selling the thing they are describing. Where a claim rests only on manufacturer sponsored work, we say so.
Section 02What we never invent
No study, author, journal, sample size, effect size or statistic appears on this site unless we are confident it exists as described. Where the evidence base for something is thin, we describe its shape rather than manufacture a precise figure that would sound better. That produces sentences like moderate evidence from several small trials rather than sentences with numbers in them, and that is the correct trade.
The same applies to businesses. We do not name clinics, practitioners, prices, addresses or phone numbers we have not verified, and we do not publish rankings of businesses we have not assessed. Where a decision requires choosing a provider, we publish the criteria and the questions to ask instead of a list.
Section 03The protocol format
Every protocol on this site specifies four things, and a page that cannot specify all four does not get published as a protocol.
| Element | What it means | Why |
|---|---|---|
| Sequence | The order of steps, numbered | Order changes the outcome for layered products |
| Quantity | How much, in a measure you can apply | Under application is the commonest cause of failure |
| Frequency | How often, and how it changes over weeks | Tolerance is built by frequency, not strength |
| Window | How long before a judgement is meaningful | Judging early is the commonest cause of churn |
Section 04Branch blocks
Every protocol includes paired branch blocks in the form of a condition and a response. That format exists because the failure mode of instructional writing is describing the happy path and leaving the reader with no guidance when something deviates from it. If a protocol here does not tell you what to do when your face objects, it is incomplete and we would like to know.
Section 05How we describe evidence
| Term used | What it means here |
|---|---|
| Strong | Multiple controlled human trials, consistent, replicated outside the funding party |
| Reasonable | Controlled human evidence exists and points one way, typically at a modest effect size |
| Moderate | Several human trials, often small, short or narrow in who was studied |
| Mixed | Human evidence exists and does not point consistently in one direction |
| Emerging | Plausible mechanism, developing evidence base, not yet settled |
| Weak | Thin human evidence, often laboratory or animal work extrapolated |
| Not established | The claim as made has not been demonstrated at the level it is sold |
Those terms are used consistently across the site. Where a popular ingredient sits at the weaker end, we say so on the page rather than burying it, and several pages here carry assessments that the market for that ingredient will not enjoy.
Section 06Language rules
- British English throughout. Including the spellings, the units and the regulatory references.
- No em dashes. A stylistic constraint applied across the whole publication.
- Instructions in the imperative. Apply, wait, reduce, stop. A protocol written in the conditional is not a protocol.
- Timings in a consistent format. Minutes and seconds for steps, weeks for windows, and the same units on every page so they can be compared.
- No adjectives doing the work of evidence. Powerful, advanced and revolutionary are not findings.
Section 07The medical boundary
This publication is written for skin that is otherwise well. Where a protocol touches a diagnosable condition, it says what the UK pathway is and where the boundary sits, and it recommends assessment rather than attempting to substitute for it.
Several pages exist principally to tell readers to stop and see someone: the escalation section, the prescription routes page, and the warning blocks that appear throughout. Those are not disclaimers added for protection. They are the parts of a protocol that tell you the protocol has ended.
Section 08Commercial position
Set out in full on the about page and summarised here. The library is free. The pack is gated behind a newsletter subscription. Category assessment pages may in future carry retailer links, only where criteria were published in advance, and only with a statement at the top of the page. No such link is live today. Two archive pages each carry one editorial link, disclosed on the page. Everything else carries nothing.
Section 09Corrections and revisions
Every page carries a revision number and a last reviewed date. Corrections are made on the page rather than by deleting it. If a protocol changes because guidance changed, the page changes and the revision increments. If you think something here is wrong, tell us and include the page and the specific claim.