Racial bias in pulse oximeter calibration

[lady_*nix]

Well. This is fucking grim.

https://bostonreview.net/science-nature ... acial-bias

TL;DR pulse oximeters are mostly calibrated for and tested on people with light skin, so very dark skin can cause them to have falsely high readings of up to 8% - enough to impact triage a whole lot, even before considering medical workers' own racial biases.

[Ikyoto]

sad, but hopefully this will be good intel for recalibration going forward

[lady_*nix]

Yeah, no kidding.

One of my medically inclined friends says the article isn't quite accurate, because a lot of nurses know to manually compensate for the inaccuracy (and will treat sustained readings below 90 the same regardless, i.e. as a dire emergency). But it sounds like there's still a huge possibility for bias. And also honestly (from experience) I wouldn't trust most docs to correct the measurements manually as nurses tend to do.

[SciFiFisher]

Yeah, no kidding.

One of my medically inclined friends says the article isn't quite accurate, because a lot of nurses know to manually compensate for the inaccuracy (and will treat sustained readings below 90 the same regardless, i.e. as a dire emergency). But it sounds like there's still a huge possibility for bias. And also honestly (from experience) I wouldn't trust most docs to correct the measurements manually as nurses tend to do.

the docs rarely look at the devices. The nurses do. So, when the doc asks for a pulse ox reading he will get it from the nurse who will do the correction.