Mastodawn

Pulse oximetry has previously been noted to have a bias against people with higher skin melanin content. It over-measures blood oxygen for Black people and others with more melanin, because most oximeters were calibrated for people with light skin. This can lead to someone with a real blood O2 reading of 91% reading as 94%. A new study shows that has led to different levels of follow-up care, with Black patients being less likely to receive supplemental oxygen as a result of this systematic mismeasurement. I was recently caring for my mom when she had a severe infection. During the recovery phase, she had supplemental oxygen at home and I literally had a pulse oximeter hanging on my neck like a pendant, taking spot checks frequently of her oxygen. It was already scary enough and I can't imagine not being able to trust the readings of what should be a life-saving device! This is why systemic racism is so insidious and pops up in all sorts of subtle but fundamental ways, especially when science is contaminated by the racist thinking! It makes me wonder what other effects this has had? Has this contributed to the known under-diagnosis of Black patients with sleep apnea? It's scary to think about.

https://news.stanford.edu/stories/2026/05/pulse-oximeter-bias-black-patients-follow-up-care

Pulse oximeter bias linked to gaps in care for Black patients

New research reveals that racial bias in pulse oximeters leads to significant disparities in follow-up care between Black and white patients.

Sam Levine 1d ago

@dantheclamman It is known to have caused deaths early in COVID among Balck patients because they weren't put on oxygen soon enough. https://pmc.ncbi.nlm.nih.gov/articles/PMC9257583/ (the note about the correction is that a couple of figure labels got reversed, which is corrected now, nothing to do with the data or paper conclusions)

The situation was maddening to anyone who has done any spectroscopy work. It's incredibly obvious the background absorbance level has to take into account skin tone the same way when I take an absorbance reading in the lab I take a blank using a cuvette made of the same material for reliable subtraction of the background. The fact the companies who made these devices screwed that up this badly was just so, so incompetent.

Checking your browser - reCAPTCHA

Poloniousmonk 1d ago

@SRLevine @dantheclamman

Isn't that how they determine how much pain you are in? I know Black pain is chronically undertreated. This could be the mechanism for it?

Sam Levine 1d ago

@Uair @dantheclamman Pain is as far as I am aware still judged by asking people to rate it on a scale of 1 to 10. A pulse oximeter is strictly for measuring the proportion of hemoglobin molecules in red blood cells that are bound to oxygen, which I don't believe has any direct correlation to pain measurements.

(was going to caveat this with "I am not a doctor", but I technically am? just not the medical type)

Poloniousmonk 1d ago

@SRLevine @dantheclamman

I know that pain correlates with blood O2 because you breathe deeper and heavier. I also think I know that MDs have a way to empirically measure pain, or at least think they do, and am under the impression that the ten scale is just a placebo to trick the patient into thinking someone cares about their pain.

I do know that patient self-reporting on a ten scale is basically meaningless. My knees used to dislocate as a kid, and I'd just pop them back in. I'd rank a dislocated knee at an 8 and setting it at a 9, but my scale has more range than people who've never been in that kind of pain. The worst pain you ever felt is always the worst pain you ever felt. It's why little kids freak out over an injury an adult would hardly notice.

I'd like to hear an MD chime in on this.

Twotired 1d ago

@Uair @SRLevine @dantheclamman
Perhaps @mcnado would care to chime in.

McNadoMD 17h ago

@Twotired @Uair @SRLevine @dantheclamman pain is determined by subjective reporting from the patient. The pain scale is used and charted by nursing staff, but I can’t remember the last time I asked someone that. I ask them if they want something for pain, and then I order it. If, for some reason, the pain is requiring too much medication to control, I have more discussion about it.

McNadoMD 17h ago

@Twotired @Uair @SRLevine @dantheclamman it is not terribly common for me to see folks “drug seeking” these days, and usually it is pretty obvious. I always let folks with bad pain know that the medication won’t make it go away, it just makes you forget it a bit. I also always use multimodal pain control (NSAIDs, acetaminophen, opioids, lidocaine, ketamine, ice, etc.).

McNadoMD 17h ago

@Twotired @Uair @SRLevine @dantheclamman for folks who can’t communicate well, I use a combination of physical exam and vital signs to decide if we need more pain control. Blood pressure and heart rate going up are good indicators of distress, and of course people in pain often look like they are in pain. Beyond that though, we don’t have a good objective pain-o-meter. Main thing is listening to the patient/caregiver.

McNadoMD 17h ago

@Twotired @Uair @SRLevine @dantheclamman and to the point of the OP here — there are myriad studies showing that we grossly undertreat pain in people of color in the US. It is a classic example of systemic racism that I use when I teach about bias.

Poloniousmonk 16h ago

@mcnado @Twotired @SRLevine @dantheclamman

Thank you for letting us know what's what. The rest of us were all guessing :)

@dantheclamman I wonder if anyone has studied the various FitBit watches for similar inconsistencies.

Yamelenko🪖💪1d ago

@kimlockhartga @dantheclamman everyone thinks same thing honestly

Dr. Dan Killam 1d ago

@kimlockhartga me too. From what I've read, wrist-based smartwatches are just inherently less accurate, but unlike the little standard finger ones a fitbit could possibly be set to different calibration curves by the app. So someone could possibly set their skin tone!

seachanged 1d ago

Talking specifically about possible under-diagnosis of sleep apnea, there are other components of diagnosis that might reduce that error, namely the nasal cannula used in a home sleep study, as well as the chest band.

Two of the three measures are not affected by skin tone, so there is a chance that they can help offset the bias in SpO2 measurement.

seachanged 1d ago

Quantitatively, the error rate for melanin-rich skin has been reported as:

In a December 2020 .. Valley .. found that Black COVID patients were three times as likely to have a significant discrepancy between pulse oximeter and ABG readings. Twelve percent of the time, when Black patients had a pulse oximeter reading in the “safe range”—92% to 96%—their actual saturation on an ABG test was below 88%. In white patients, this discrepancy occurred 4% of the time.

Dr. Dan Killam 1d ago

@seachanged horrible. In the 80s is an emergency. After decades using pulse ox. I guess there's a few models that have multispectrum approaches, but still very uncommon. They should be standard

So oximeters need to measure reflectance as well as transmission, and should have from the start.

Jeff Horton 🇨🇦1d ago

@dantheclamman is that the same problem as this (story today in CBC no less) https://www.cbc.ca/news/canada/kitchener-waterloo/gurnoor-kaur-canada-wide-science-fair-project-innovation-award-wow-9.7217194

Kitchener teen wins innovation award at Canada-Wide Science Fair, fixes 35-year-old problem | CBC News

Gurnoor Kaur, a Grade 11 student at Cameron Height Collegiate Institute in Kitchener, has won the best project award for innovation at the Canada-Wide Science Fair in Edmonton.

CBC

Dr. Dan Killam 1d ago

@jeffhorton sounds related!