The new research suggests that skin color–related differences in pulse oximeter readings are in fact impacting clinical decision-making, lead author Eric R. Gottlieb, MD, of Brigham and Women’s Hospital and Massachusetts Institute of Technology, both in Boston, and colleagues wrote. This suggests that technology needs to updated to improve health equity, they continued, in their paper published in JAMA Internal Medicine.
Dr. Eric R. Gottlieb
“It has been known for decades that these readings are affected by various surface pigmentations, including nail polish and skin melanin, which may affect light absorption and scattering,” the investigators wrote. “This increases the risk of hidden hypoxemia [among patients with darker skin], in which patients have falsely elevated SpO2 readings, usually defined as 92% or greater, with a blood hemoglobin oxygen saturation less than 88%.”
Although published reports on this phenomenon date back to the 1980s, clinical significance has been largely discounted, they said, citing a 2008 paper on the topic, which stated that “oximetry need not have exact accuracy” to determine if a patient needs oxygen supplementation.
‘We’re not providing equal care’
Questioning the validity of this statement, Dr. Gottlieb and colleagues conducted a retrospective cohort study involving 3,069 patients admitted to intensive care at the Beth Israel Deaconess Medical Center in Boston between 2008 and 2019, thereby excluding patients treated during the COVID-19 pandemic. The population consisted of four races/ethnicities: White (87%), Black (7%), Hispanic (4%), and Asian (3%).
Aligning with previous studies, multivariable linear regression analyses showed that Asian, Black, and Hispanic patients had significantly higher SpO2 readings than White patients in relation to hemoglobin oxygen saturation values, suggesting falsely elevated readings.
Further modeling showed that these same patient groups also received lower oxygen delivery rates, which were not explained directly by race/ethnicity, but instead were mediated by the discrepancy between SpO2 and hemoglobin oxygen saturation values. In other words, physicians were responding consistently to pulse oximetry readings, rather than exhibiting a direct racial/ethnic bias in their clinical decision-making.
“We’re not providing equal care,” Dr. Gottlieb said in an interview. “It’s not that the patients are sicker, or have other socioeconomic explanations for why this happens to them. It’s us. It’s our technology. And that’s something that really has to be fixed.”
The investigators offered a cautionary view of corrective algorithms, as these “have exacerbated disparities and are subject to ethical concerns;” for example, with glomerular filtration rate estimations in Black patients.
Dr. Gottlieb also cautioned against action by individual physicians, who may now be inclined to change how they interpret pulse oximeter readings based on a patient’s race or ethnicity.
“I don’t think that we can expect physicians, every time they see a patient, to be second guessing whether the number basically reflects the truth,” he said.
Instead, Dr. Gottlieb suggested that the burden of change rests upon the shoulders of institutions, including hospitals and device manufacturers, both of which “really need to take the responsibility” for making sure that pulse oximeters are “equitable and have similar performance across races.”
While Dr. Gottlieb said that skin color likely plays the greatest role in measurement discrepancies, he encouraged stakeholders “to think broadly about this, and not just assume that it’s entirely skin color,” noting a small amount of evidence indicating that blood chemistry may also play a role. Still, he predicted that colorimetry – the direct measurement of skin color – will probably be incorporated into pulse oximeters of the future.