The Pulse Oximeter Is Racially Biased
Inaccurate readings mean some Black people have dangerously low oxygen levels that doctors don’t catch
Credit: Grace Cary / Getty Images
Pulse oximeters, the most commonly used tool to measure a person’s blood oxygen levels, over-estimate the oxygen saturation of many Black patients. As a result, these patients may be receiving inadequate supplemental oxygen in hospitals and other settings. It also means that home pulse oximeters that people use to monitor their respiratory function if they have Covid-19 or emphysema could be giving inaccurate readings to Black users.
The research, published yesterday in the New England Journal of Medicine, was conducted by doctors at the University of Michigan who were taking care of hospitalized people with Covid-19. The physicians noticed that they were getting different results from pulse oximeters and arterial blood gas, a more invasive but more precise way to measure oxygen levels directly from the blood, and this discrepancy seemed to be happening more often in Black patients.
The doctors decided to look into the observation further, focusing on a specific question that would have clinical significance: How often were normal pulse oximeter readings between 92% and 96% missing arterial blood gas oxygen levels below 88%, the point at which a person would be considered hypoxic, meaning they had dangerously low oxygen levels. They discovered that this scenario occurred in 11.6% of Black patients and 3.6% of white patients, more than a three-fold difference.
“In this range, a saturation difference of a couple of points would make a difference in how you treat the patient,” says Michael Sjoding, MD, a pulmonary specialist at University of Michigan Medical School who led the research. “You definitely would increase a patient’s supplemental oxygen that they’re receiving if you actually learned that the saturation was really below 88%. It’s even conceivable that if a patient came to the emergency department and the saturation was actually 88% percent but you’re reading it at 94%, it may influence your decision to admit the patient to the hospital for monitoring.”
Pulse oximeters indirectly estimate oxygen levels based on how much light travels through a person’s finger. Skin pigment, specifically melanin, can interfere with light transmittance through the skin, making the reading less accurate. Like so many drugs and medical devices, the pulse oximeter was developed and calibrated on white people, and, as a result, it does a poor job of reading information from people with darker skin tone.
Utibe Essien, MD, MPH, an assistant professor of medicine at the University of Pittsburgh, says that this finding is just the latest example of racially biased medical tools and calculators used to make clinical care decisions that end up harming Black patients. Other examples include the spirometer used to measure lung capacity, the estimated glomerular filtration rate (EGFR) that’s critical for kidney disease diagnosis and treatment, and a cardiovascular calculator that estimates a person’s risk for heart attack or heart disease.
“White is the baseline, white is the quote-unquote normal. And whether that’s in our education, in our research, and ultimately in our clinical care, we end up with conversations like we’re having now as to ‘Why does this tool fail Black individuals?’” Essien says. “I think we have to look back and wonder, well, who are the individuals who are actually studied, who are actually part of the initial trials, whether we’re looking at a pulse oximeter or we’re talking about a lab value that helps patients get the care that they need.”
Both Sjoding and Essien say that while they can’t know for sure, they worry that this error could be contributing to the racial disparities seen with Covid-19, a disease that has killed Black Americans at nearly three times the rate of white Americans.