How we review devices

1. 1. Technical specifications. Every figure carries its source — filing number and page when it comes from the FDA. If the manufacturer does not publish a figure, the entry reads not published and the field stays empty. We do not fill it with estimates, derivations, or values from similar products.
2. 2. Comparison with published research. We compare the parameters a device declares against the ones used in published studies. That is a comparison of numbers, not a judgement: we state what each one declares, not whether it is enough for a clinical goal.
3. 3. Usage notes. Build, fit, usability. Only when we have had the unit in hand.
4. 4. Who it suits and who it does not. Based on what the device declares and how it is built.

Every entry states which of the two it is, and the standard is the same for every brand.

Strongest to weakest. Every figure in an entry carries which one it is.

1. FDA 510(k) — taken from the manufacturer’s own public filing, with K number and page. It is the strongest because the manufacturer submitted it to a regulator.
2. Manufacturer documentation — a spec sheet or document supplied by the brand.
3. Claimed, undocumented — the brand asserts it and there is no document behind it. Published, and labelled as such.
4. Not published — we looked and it does not exist. That is not the same as not having looked, and the index shows the two differently.

Of the 60 face masks in the index, 3 do not publish their irradiance anywhere; across the full 129-device corpus it is 14. That a figure is not published is a fact about what is public, not a claim about the device or the manufacturer.

• Gifted units. If a brand sent us the product, the entry says so.
• Affiliate links. Several links on this site pay us a commission if you buy. They do not change what we recommend or the order things appear in.
• Commercial relationships. Any tie to a brand is disclosed in the entry and in the bio of whoever wrote it.

• We do not guarantee conclusions or ranking position to any brand.
• We do not sell placement in our lists.
• We do not claim to have measured something we did not measure. If an irradiance figure comes from the manufacturer’s filing, the entry says it comes from the filing — we have no way to measure it ourselves.

We do not give medical advice and we do not diagnose. No one on the team has clinical training. We do not interpret what any individual should use, and we do not state whether a dose is sufficient or adequate for a health outcome.

What we do instead: publish verified data with its source, and the arithmetic that relates it. If you want to know what suits you specifically, that is a conversation with a dermatologist, not with a table.

This site used to show editorial scores out of 5. We removed them: there was no written methodology behind the number, and a score without a method is worse than no score at all, because it implies a precision that does not exist and cannot be audited.

They come back when a method is published on this page stating what is measured, with what weighting, and against what source — so that anyone could recalculate the score and arrive at the same result. Until then the entries show the figures and where they come from, which is what can be checked.

A device can appear more than once in the index under different K numbers: a manufacturer files again for a new model, a new indication, or a revised design. easy Claire, reVive Essentials and the Ulike Reglow UM10 each hold two. Those are separate, valid filings, not duplicates — and their declared figures can legitimately differ, because they describe different submissions.

We keep the rows separate and show every K number, rather than merging them into one product row. Merging would force us to pick which filing’s figures to show, and that choice is not ours to make silently.

We have published 2 figures incorrectly. Both were our extraction errors, not manufacturer errors, and both are listed here permanently. Correcting in public only counts if the corrections can be found.

Across all 129 filings we read we found no case where a manufacturer’s own published figures contradict each other. When they publish, the numbers add up.

If you find a figure that is wrong, write to us and we will correct it with the date of the change visible. The technical data comes from public documents: if a filing says something different from our entry, the filing wins.