Pulse Oximeters and Skin Tone: Why Accuracy Hasn't Been Equal for Everyone

A pulse oximeter works by shining light through your fingertip and measuring how much is absorbed. Oxygen-rich and oxygen-poor blood absorb light differently, and the device reads oxygen saturation from that difference. It is elegant, cheap, and everywhere.

It also has a known problem. Skin pigment absorbs light too. And because many oximeters were calibrated and validated mostly on lighter-skinned people, they can read less accurately on darker skin — often overestimating oxygen levels, which is the dangerous direction to be wrong in. An overestimate can make someone look fine when they are not, delaying care that hypoxia needs urgently.

This is not a fringe concern. It became a prominent issue during the COVID-19 pandemic, and a regulatory communication in 2022 flagged the accuracy gap on darker skin tones explicitly. For a population like ours in West Bengal, where most patients have medium to dark skin, a device that quietly performs worse on darker skin is not a device we can use as-is and trust.

Why this matters for the work we do

If you are going to screen oxygen saturation across a community where darker skin tones are the norm, you cannot treat skin-tone accuracy as an afterthought. A screening tool that is accurate only for some of the people it screens is not really accurate. It just looks accurate on average while failing the people it was supposed to help.

So when we designed SamaClip's sensing and set out to validate it, accuracy across skin tones was not one item on a checklist. It was a core requirement. We deliberately built our validation to recruit across the full range of skin tones we actually serve, using the Monk Skin Tone scale — a more inclusive reference than the older scales — so that performance could be reported for darker tones specifically, not buried in an average.

We are running that validation study now. When it is complete, we will publish what it shows for each skin-tone group, rather than a single headline number that hides the variation. If a device claims to work for everyone, the burden is to show it works for the people most often left out of the testing. That is the standard we are holding ourselves to.

The broader point

Medical technology inherits the assumptions of whoever built and tested it. When the testing population is narrow, the blind spots get shipped to everyone else. Closing the skin-tone gap in something as basic as oxygen measurement is not a nice-to-have. It is part of what "equitable screening" has to mean if the phrase is going to mean anything.