Do pulse oximeters have racial bias?

On Sunday, the UK’s Department of Health and Social Care announced an independent review to find any potential bias in medical items such as pulse oximeters and their impact on patients from various ethnic groups. The development comes amid concerns that many medical devices are designed and calibrated for White patients.

The UK government said in a release that the coronavirus pandemic had exposed health disparities across the country with death rates higher among ethnic minorities.

How a pulse oximeter works
A small clip-like device that calculates the amount of oxygen in the bloodstream, pulse oximeters can be attached to the fingertip, toes, or earlobe.

The device has a small electronic processor and two LEDs — one emitting infrared light and another emitting red. The lights pass through the finger and the device senses what comes through the other side.

Blood contains haemoglobin which, when saturated with oxygen, is called oxygenated haemoglobin and is bright red in colour. Haemoglobin without oxygen is called deoxygenated haemoglobin.

Oxygenated haemoglobin and deoxygenated haemoglobin absorb infrared and red light differently. Oxygenated haemoglobin allows more red light to pass through and absorbs more infrared light than deoxygenated haemoglobin. The processor measures the amount of light transmitted and the device displays the oxygen saturation level or the percentage of oxygenated haemoglobin in the blood. The normal blood oxygen saturation rate is between 95% and 100%.

Why the device is racist
Light passing through the skin is the initial step in an oximeter. Most of these devices are calibrated for light skin and can have erroneous results in non-White people.

The Hypoxia Research Laboratory at the University of California (San Francisco), which tests pulse oximeter performance, conducted several experiments on the devices. Hypoxia Research Laboratory Director Dr Philip Bickler told The New York Times that the simplest way to explain the inaccuracies in dark-skinned patients was that the pigment scattered the light around, reducing the signal. He added that it was like adding static to a radio signal — there is more noise, less signal.

Last year, a letter to the editor published in The New England Journal of Medicine, titled ‘Racial Bias in Pulse Oximetry Measurement’, noted that the devices’ results could be misleading for Black patients.

The team studied 10,789 pairs of oxygen saturation measures obtained using arterial blood gas test and pulse oximetry. The researchers studied 276 Black and 1,333 White patients. They found that the pulse oximetry overestimated oxygen saturation 3.6% of the time in White patients. For Black patients, it overestimated 12% of the time.

The researchers conducted a second study on 37,308 pairs of oxygen saturation measures from 1,050 Black patients and 7,342 White patients with similar results.

Lead author Dr Michael W. Sjoding told The New York Times that most of the medical community had been working on the assumption that pulse oximetry was accurate.

Dr Sjoding, an assistant professor of internal medicine at the University of Michigan Medical School, said as a trained pulmonologist and critical care physician, he had no understanding that the pulse oximeter could be potentially inaccurate, adding that he was missing hypoxemia in a certain minority of patients.

A study published in Anesthesiology in 2005 had raised the issue of racial bias in oximeters. The study, ‘Effects of Skin Pigmentation on Pulse Oximeter Accuracy at Low Saturation’, found that the device overestimated the arterial oxygen saturation by 3% in patients with dark pigments, compared with 0.37% in lightly pigmented patients.

The team wrote that in its 18 years of testing accuracy of pulse oximeters, the majority of subjects had been light-skinned. It is likely that most pulse oximeters have been calibrated using light-skinned individuals based on the assumption that skin pigment did not matter, the team added.