Home Chronic conditions · Segment guide

Wearables for health monitoring

Continuous glucose, cardiac monitoring, epilepsy alerts, oxygen. What's clinically validated, what's wellness, and what a clinician will actually accept data from.

One distinction runs through everything here

A device is either cleared by a regulator for a stated medical purpose, or it is a wellness product. The hardware can be near-identical. What differs is what the manufacturer is permitted to claim, what evidence sat behind it, and whether a clinician can act on the reading. "Medical grade" is a marketing phrase; the clearance is the fact.

Continuous glucose monitoring: three products, three different permissions

The short version: you can now buy a CGM without a prescription — but the over-the-counter versions are cleared for different purposes, and one of them is explicitly not for managing diabetes.

What each OTC monitor is actually cleared for

The FDA cleared the first over-the-counter CGM, Dexcom's Stelo, in March 2024. Abbott's Lingo and Libre Rio followed in June 2024. As of mid-2026 Stelo and Lingo are the two available in the US; Libre Rio has clearance but no announced launch.

  • Lingo is a wellness CGM for adults and is not cleared to help manage diabetes. It's aimed at people without diabetes interested in general health.
  • Libre Rio is cleared for adults with type 2 diabetes who manage it through diet and activity — not those using insulin.
  • Stelo was the first cleared for over-the-counter purchase; check its current indication against your situation before relying on it.

The practical consequence: two sensors that look the same and sit on the same arm can carry completely different permissions. If insulin dosing is involved, an OTC wellness sensor is the wrong product — that's a prescription CGM decision made with a clinician.

SpO2: the accuracy problem regulators are still fixing

The short version: pulse oximeters are less accurate on darker skin, the effect is well documented, and the FDA moved in 2025 to tighten testing. If you're relying on oxygen readings, know this before you trust a number.

The evidence, and what the new guidance requires

Melanin absorbs the red and infrared wavelengths pulse oximetry depends on, which biases readings. The clinical consequence is occult hypoxemia — oxygen genuinely low while the device reads acceptably. Among patients showing SpO2 of 92–96%, Black patients were found to have roughly three times the frequency of occult hypoxemia compared with White patients.

In January 2025 the FDA issued draft guidance to improve accuracy across skin pigmentations for oximeters used medically. It proposes substantially stricter validation:

  • 3,000 data points instead of 200
  • Study samples of 150 or more people rather than 10
  • At least 25% of participants with dark skin, and 30% of data points from darker-skinned individuals
  • Prominent labelling that skin pigmentation may affect sensor performance

This guidance addresses oximeters used for medical purposes. The SpO2 feature on a general-purpose smartwatch is typically a wellness feature and sits outside it entirely — so the accuracy concern applies at least as much, with less oversight.

What a clinician will actually accept

Consumer wearable data is rarely admitted directly into a clinical record. What it does well is prompt a conversation — a watch flagging an irregular rhythm leads to a proper ECG, which is the thing that gets acted on. Treat the wearable as the reason to make an appointment, not as the result. If you intend the data to inform treatment, ask the clinician beforehand what form they can use.

Alerting devices: ask what happens when nobody responds

The short version: seizure alerts, cardiac alerts and similar are only as good as the response behind them. The sensor is the easy part.

The questions that separate a system from a gadget
  • What is it validated to detect? Seizure detection, for example, performs very differently across seizure types — a device validated for one may be near-blind to another.
  • What's the false-alarm rate in real life? Nightly false alerts get devices switched off, which is worse than not having one.
  • Who is alerted, and what if they don't answer? A notification to a phone on silent is not a response plan.
  • Does it work without a phone nearby? Many alerting wearables are Bluetooth-tethered.
  • What happens if the company stops trading? Cloud-dependent medical alerting stops when the servers do.

Seven questions before buying

  1. Is this cleared by a regulator for a medical purpose, or sold as wellness?
  2. If cleared — for which purpose, and which population? (They're narrower than the marketing.)
  3. Has my clinician said they can use data from it?
  4. What are the documented accuracy limits, and do any apply to me?
  5. Who receives alerts, and what's the plan if nobody answers?
  6. Does it need a phone nearby to work?
  7. Where does this health data go, who else receives it, and for how long is it kept?

What we're working on

Assessments for this segment will state each device's regulatory status and indication in plain language, alongside documented accuracy limits, alerting behaviour, data handling, certification status and published SAR figures. The weekly brief is where they land first.

Not a test laboratory. ViaMondo does not measure SAR, RF emissions or any other physical property of a device, and makes no independent measurement claims. Figures we publish are taken from manufacturers' published data and public regulatory filings, cited to source.

Not medical advice. This page is general and educational. Nothing here diagnoses, treats, prevents or cures any condition, and no device described replaces clinical care. Do not change medication, insulin dosing or treatment based on a wearable reading. Always consult a qualified healthcare professional about your own situation or that of someone you care for.