Direct comparison with blood tests reveals patterns of variability correlated with pigmentation

The differences, or bias, between estimates of blood oxygen saturation levels as measured with pulse oximeters compared to the gold-standard method of measuring oxygen saturation in arterial blood varied significantly between patients with darkly pigmented skin compared to lighter skin pigment, according to research presented at the American College of Cardiology’s Annual Scientific Session (ACC.25).

The new findings are from the largest prospective real-world study to evaluate the relationship between skin pigmentation and pulse oximeter bias in critically ill patients, a topic that garnered public attention during the COVID-19 pandemic when some studies suggested that pulse oximeters may provide false reassurance about a patient’s health status by overestimating oxygen saturation in certain racial groups. In contrast with some previous studies, the current study’s results indicate that pulse oximeter readings, on average, underestimated blood oxygen levels. However, the results also showed that pulse oximeters overestimated oxygen saturation, showing a positive bias in 20% of observations, and the proportion of observations with positive bias differed across skin pigment groups.

“Although pulse oximeter bias on average was negative for all people, it was less negative in the darkly pigmented people than in the people with lighter pigment, meaning that pulse oximeters do not perform the same across different skin pigment categories,” said Carolyn Hendrickson, MD, associate professor of medicine at the University of California San Francisco and the study’s first author. “We also found that the proportion of positive bias—the one that goes in the worrisome direction meaning that someone might have dangerously low oxygen saturation that is not detected with a non-invasive monitor—was higher in patients with dark skin pigment compared to those with medium and light skin pigment.”

Blood oxygen saturation is an important tool that clinicians use to triage patients and make treatment decisions. Hospital admission, administration of supplemental oxygen and the use of advanced treatments for COVID-19 all rely on understanding a patient’s level of blood oxygen saturation, Hendrickson said. With pulse oximeters, clips or disposable adhesive probes are applied to the fingertip or earlobe where they shine certain wavelengths of light through the tissue and measure how much light gets absorbed or reflected. For the study, researchers compared pulse oximeter readings with functional oxygen saturation from blood gas analysis, a gold-standard blood test for oxygen saturation.

The study enrolled 631 patients averaging 62 years of age who received treatment in the intensive care unit at Zuckerberg San Francisco General Hospital between 2022 and 2024. About one-quarter of participants were identified in medical records as White, one-quarter identified as Hispanic, one-fifth identified as Black and one-fifth identified as Asian.

Researchers measured each participant’s skin pigmentation using both the subjective Monk Skin Tone Scale and objective measurements of melanin content with a spectrophotometer, a non-invasive device that measures reflectance and absorbance patterns of light. According to the objective measurements, 53% of patients were classified as having medium pigmentation, 33% were classified as having light pigment and 14% were classified as darkly pigmented.

Pulse oximeter and blood gas analysis readings were taken simultaneously, and most study participants had at least two such readings, for a total of 1,760 paired measurements among the 631 study participants.

In their analysis, the researchers accounted for over 30 factors that could potentially influence blood oxygen saturation readings including demographic factors, comorbidities, medications and physical attributes. After adjusting for these factors, the results showed the same overall trends, with patients overall seeing a bias toward underestimates with the pulse oximeter but people with darker pigmentation showing less negative bias and a greater likelihood of an overestimate compared to those with lighter pigment.

Based on the results, researchers emphasized the importance of recognizing the uncertainty in pulse oximeter readings and suggested that future devices could be designed with alarms or lights to indicate when the reading is less certain.

“Our study shows that the oximeters have a lot more uncertainty in the critically ill patients than they do in the healthy volunteers who participate in validation studies,” she said. “More discussion is needed between manufacturers, regulators and clinicians to draw attention to times when the oximeter is uncertain.”

Hendrickson said the study findings can also help to inform how pulse oximeters are tested and regulated.

“Making sure there’s representation across skin pigmentation categories is really important for future studies evaluating pulse oximeter performance,” Hendrickson said. “We think that the social construct of race is important and impacts health outcomes, and it is not the same thing as skin pigmentation. We’re advocating for the use of skin pigment data to be collected in addition to race when trying to understand equitable performance in a variety of patient populations.”

Researchers said that several aspects of the study limit its generalizability. First, the study involved only one type of pulse oximeter made by a single manufacturer; in addition, the medical-grade devices used in the study are held to regulatory standards while those that are available for home use are not typically tested in validation studies. Also, many of the oxygen saturation readings were in a relatively high range—with most being above 90%—so it is not clear whether the same patterns would be seen in patients with lower oxygen levels. In addition, only a small portion of study participants had very dark pigmentation.

Looking forward, Hendrickson said that additional large, prospective studies with a greater representation of more darkly pigmented patients and more observations in lower oxygen saturation ranges would help to further clarify patterns in pulse oximetry bias.

The study was funded by the U.S. Food and Drug Administration.

ACC.25 will take place March 29-31, 2025, in Chicago, bringing together cardiologists and cardiovascular specialists from around the world to share the newest discoveries in treatment and prevention. Follow @ACCinTouch, @ACCMediaCenter and #ACC25 for the latest news from the meeting.

The American College of Cardiology (ACC) is the global leader in transforming cardiovascular care and improving heart health for all. As the preeminent source of professional medical education for the entire cardiovascular care team since 1949, ACC credentials cardiovascular professionals in over 140 countries who meet stringent qualifications and leads in the formation of health policy, standards and guidelines. Through its world-renowned family of JACC Journals, NCDR registries, ACC Accreditation Services, global network of Member Sections, CardioSmart patient resources and more, the College is committed to ensuring a world where science, knowledge and innovation optimize patient care and outcomes. Learn more at ACC.org.

###

Hendrickson will be available to the media in a press conference on Saturday, March 29 at 4:15 p.m. CT / 21:15 UTC Room N226.

Hendrickson will present the study, “The EquiOx Study: Evaluating Pulse Oximeter Bias across a Range of Skin Pigment in Critically Ill Adults,” on Sunday, March 30, 2025, at 8:00 a.m. CT / 13:00 UTC in the Main Tent, North Hall B.