How we evaluate devices, and what we refuse to claim

This is the page the rest of the site is accountable to. If a recommendation here ever contradicts what follows, the recommendation is wrong.

1. Published specifications. Wavelengths in nanometers, diode counts, microamp or joule output, target temperatures, session lengths, protocols and warranty terms - exactly as the manufacturer publishes them.
2. Regulatory position. Whether a 510(k) record exists in the FDA's public database for that specific device, its number, decision date, applicant and product code.
3. What is missing. Which specification the manufacturer declines to publish. In the LED category that is almost always irradiance, which is the figure that governs dose.
4. Fit and practicality. Shell material, coverage, session length and whether the protocol is one a real person will complete. The best device is the one that gets used.
5. Who should not buy it. Every page names a reader for whom the device is the wrong purchase. If we cannot name one, we have not thought about it hard enough.

• Manufacturer specification pages, read directly and dated. Where a retailer and a manufacturer disagree, the manufacturer wins.
• The FDA 510(k) premarket notification database, the free public one at accessdata.fda.gov, searched by device name, by brand and by applicant. We also use the openFDA API for bulk queries.
• Peer-reviewed literature via PubMed, linked by DOI or PubMed ID where we reference a mechanism or a result.
• Publicly posted owner reports, linked, and always labeled as anecdote rather than evidence.

Every page carries a Sources section listing the URLs behind its factual claims. The underlying dataset covers 44 devices: we located 510(k) records for 20 and found nothing for 24.

This is the method, written out so you can repeat it and check us.

1. Search by device name first - the retail name as printed on the box.
2. If nothing returns, search by applicant: the brand, then its parent, then any contract manufacturer named on the packaging. Several consumer LED brands are cleared under a factory's name rather than their own.
3. Check the product code. OHS is light-based over-the-counter wrinkle reduction, OLP is acne, NFO is over-the-counter facial toning, OHT is hair removal, PAY is radiofrequency. A clearance in the wrong code does not support the claim being made.
4. Check the decision date. A 2008 clearance for a discontinued predecessor tells you nothing about the device on sale today.
5. Record the result either way. A blank is a finding, and we publish the blanks.

• We never claim to have tested a device. No lab, no review units, no test cohort. The phrase "in our testing" does not appear on this site.
• We never publish a star rating or a numeric product score. A score asserts a measurement we did not take.
• We never publish a price in the copy. Prices render only from a live data layer with a date stamp, and they expire after 48 hours. If the feed stops, numbers disappear and the buttons read "Check price on Amazon".
• We never present a brand-stated figure as measured. Every irradiance and energy-density number on this site is labeled brand-stated.
• We never write "FDA approved". No device in any category covered here is FDA approved. Some are cleared, which is a different and lower bar.
• We never fabricate a review, a testimonial or a reviewer. There are no testimonials on this site because we have none to publish.

Devices move up for publishing figures that can be checked and for holding a clearance record that names the exact model. They move down for asking the reader to take a claim on trust. Fit and session length are tiebreakers, because a device that is not used delivers nothing regardless of specification.

That produces a ranking of verifiability, not of outcome, and we want to be explicit about the limitation: we cannot rank these devices by effectiveness, because the data required to do so is not published by anyone. Any publisher claiming otherwise is ranking by feel.

Glow & Current earns affiliate commission through Amazon. Commission rates do not vary enough across these devices to make a cheaper recommendation a sacrifice, and several pages here recommend the cheaper option or recommend not buying at all. No brand pays for placement, no brand sees copy before publication, and no brand supplies us with product. Full disclosure here.

Specifications change when manufacturers revise a product page, and clearance records change when the FDA publishes a new decision. We re-check pages when something changes rather than on a fixed schedule, and we date every specification table. Our corrections policy is here, and the correction form is here.