As an emergency medicine physician, Dr. Owais Durrani sees this issue regularly first-hand: When he clamps a pulse oximeter onto a patient’s fingertip to measure their blood oxygen levels, the small device can yield less accurate readings in people with darker skin tones, reporting levels higher than what they actually are.
The possibility of an inaccurate oxygen reading for Black and brown patients worries Durrani in the emergency room on a daily basis.
“Was there a point where I may have discharged a patient that had a reading of, say, 95%, but the actual reading was lower and they maybe could have benefited from observation or being admitted to the hospital? Possibly, and that scares me, and that frustrates me,” he said. “So that’s why it’s so important that the FDA is looking into this.”
A panel of the US Food and Drug Administration’s Medical Devices Advisory Committee met Tuesday to review clinical data about the accuracy of pulse oximetry in patients with darker skin and to discuss recommendations on using these devices on people with dark skin tones and whether they should have labels – such as a black box warning – noting that inaccurate readings may be associated with skin color.
Because the public continues to use these devices to check oxygen levels, “I think we should give more detailed instructions on what they’re supposed to do with them,” Dr. Murad Alam, a member of the panel and professor at Northwestern University, said in the meeting.
“I understand we don’t know exactly what their accuracy is, but I would suggest that FDA might want to consider a labeling change,” he said. “In the meantime, it would allow the FDA the ability to investigate these products and give consumers some guidance as to how to use them – and have some power to actually, at the very least, compel the manufacturers to convey that simplified and expanded labeling to patients who will in fact be using these to try to save their own lives.”
The meeting was intended to help inform the FDA’s ongoing evaluation of pulse oximeters, their accuracy and performance, and what can be done to fix racial biases. The Anesthesiology and Respiratory Therapy Devices Panel heard from patients, regulators, researchers and developers of medical devices.
“The performance of medical devices needs to be well understood to mitigate any negative, unintended consequences for patients in different populations,” Adm. Richardae Araojo, FDA associate commissioner for minority health, said during the meeting. “So today’s discussion on the real-world performance of pulse oximeters and factors that impact their accuracy will inform FDA as we consider the regulation of these medical devices.”
Pulse oximeters work by sending light through your finger; a sensor on the other side of the device receives this light and uses it to detect the color of your blood. Bright red blood is highly oxygenated, while blue or purplish blood is less so. If the device isn’t calibrated for darker skin tones, the pigmentation of the skin could affect how the light is absorbed by the sensor, leading to flawed oxygen readings.
Ultimately, the pulse oximeter can estimate the amount of oxygen in the blood without the need for a blood sample. But on a person with dark skin, the oximeter could indicate that oxygen levels are normal, suggesting that the person may be discharged – when, in reality, a blood sample might show low oxygen levels, suggesting that they need not only additional care but oxygen support.
To compensate for the possible discrepancy, Durrani said, he sometimes conducts a “walking” pulse oximeter reading on patients with darker skin.
“So having a patient walk for like five minutes while having a pulse ox on, and usually if they are at higher risk, what will happen is, their oxygen numbers might drop further,” said Durrani, who practices at a hospital in Houston.
A walking test can help him gauge whether the standard pulse oximeter reading overestimated a person’s oxygen levels, but he is still waiting for the day when he can trust the device to provide an accurate oxygen level measurement for all of his patients – no matter the color of their skin.
“I’m definitely very happy that this critical issue is being addressed on a national level at the FDA,” said Durrani, who is not involved in the meeting or with the agency.
“But as an ER doc who’s in the trenches and making a lot of fast-paced decisions based on this data, I’d like to see more discussion hopefully in this meeting or future meetings around what the clinical implications of this are going to be,” he added. “We would need a more kind of multipronged guidance when it comes to how physicians should then interpret these results.”
The FDA announced in February 2021 that although pulse oximetry can be useful for estimating blood oxygen levels, the devices “have limitations” and “a risk of inaccuracy under certain circumstances,” referring to a 2020 report on racial bias in pulse oximetry measurements that was published in the New England Journal of Medicine.
However, evidence of flawed pulse oximeter readings in people with dark skin dates to the 1980s. In 2013, the FDA issued premarket guidance for developers of pulse oximeters, recommending that they have “a range of skin pigmentation” represented in their clinical studies of the devices, including at least two “darkly pigmented subjects or 15% of the study group, whichever is larger.”
The issue of inaccurate readings for dark-skinned individuals has gotten more widespread attention as the use of pulse oximeters increased dramatically at the start of the Covid-19 pandemic.
“I think that the Covid-19 pandemic has exacerbated this issue,” Rutendo Jakachira, a Ph.D. student in Brown University’s Department of Physics who is studying racial disparities in pulse oximetry, told CNN in July.
“People have known that there’s an issue with the overestimation of these oxygen saturation levels, particularly in dark-skinned patients, but they did not appreciate the significance of this issue until the Covid-19 pandemic,” she said. “The pandemic has seen an increased use of pulse oximeters in the hospital and in the home setting and thus the consequences of the neglect of the issues with errors in these pulse oximeter has been revealed.”
One study on flawed pulse oximeter readings, published in the journal JAMA Internal Medicine in July, found that among more than 3,000 hospitalized patients receiving intensive care, Asian, Black and Hispanic patients got less supplemental oxygen than White patients, and that was associated with differences in their pulse oximeter readings.
Other research published that month in the medical journal BMJ found that Black patients had higher odds than White patients of having low blood oxygen noted in their drawn-blood readings but not detected by pulse oximetry. That finding was based on data from the Veterans Health Administration in which pulse oximeter readings were paired with oxygen level measurements from drawing blood.
Another study of about 7,000 Covid-19 patients, published in May in the journal JAMA Internal Medicine, found that compared with White patients, pulse oximetry overestimated oxygen levels in the blood by an average of 1.7% among Asian patients, 1.2% among Black patients and 1.1% among Hispanic patients. That overestimation may have contributed to an unrecognized or delayed recognition of someone’s eligibility for Covid-19 therapies.
Some experts in Tuesday’s meeting questioned whether differences in how pulse oximeters perform among people with dark or light skin are significant.
“The data that we see from some of these studies, it really begs for, as some other people have said, prospective data collection with a measurement of skin pigmentation,” said Richard Branson, a voting member of the panel and professor emeritus at the University of Cincinnati.
Studies of Medtronic’s Nellcor pulse oximeters found that they reported blood oxygen levels that were within 2% of participants’ drawn-blood oxygen levels – regardless of skin color, Dr. Sam Ajizian, chief medical officer of patient monitoring at Medtronic, said in an emailed statement Tuesday.
“Still, the data shows a small statistical discrepancy between results for those with light pigmentation and patients with darker skin pigmentation,” he said.
“Medtronic is seeking to make improvements in our devices based on a greater understanding of the impact skin pigmentation has on pulse oximetry readings,” he said. “Through better information-sharing and an industry-wide commitment to continued innovation, we are advocating for improvements in the methods we use to validate pulse oximeters, including standardization of how we assess skin pigmentation and an increase in representation of patients with darker skin pigmentations in clinical trials.”
Pulse oximeters, invented in 1974 by Japanese bioengineer Dr. Takuo Aoyagi, are not the only medical devices that have been found to perform more accurately in people with fair skin than in those with darker skin – largely because of a lack of testing in racially diverse groups of people.
“The first one that jumps into my mind is forehead thermometers, infrared thermometers: It’s a similar issue, in folks with darker skin, they tend to underreport fever in those patients,” Durrani said. “A lot of times, I’ve seen those thermometers, anecdotally speaking, just not even giving us a reading when it comes to darker folks.”
Temporal thermometers – used to measure body temperature on the forehead – use light-sensing technology similar to that in pulse oximeters. Research published in September in the medical journal JAMA found that the devices may be less accurate than oral thermometers at detecting fevers among hospitalized Black patients, and undetected fevers could lead to delays in medical care.
A study published in July in the journal ACM Transactions on Graphics describes how wearable devices that measure heart rate may not perform as well in people with darker skin tones because they also use similar technology to forehead thermometers and pulse oximeters.
Some research groups across the United States are seeking solutions to address these biases. Scientists at the University of California, Los Angeles are developing a new approach for wearable heart-rate sensors that involves radar technology.
Another group, involving Jakachira and her colleagues at Brown University, is developing methods to make pulse oximeters more accurate in people with dark skin tones by canceling out the effect that melanin – the substance that produces pigmentation in our skin, hair and eyes – can have on how these devices perform.
In theory, any medical equipment that leverages infrared sensing in how it works “will likely show discordance in its performance across people with different skin tones,” said Dr. Leo Anthony Celi, clinical research director and principal research scientist at the MIT Laboratory for Computational Physiology, told CNN in September.
That’s because the technology in such devices, like pulse oximeters and temporal thermometers, involves emitting and capturing light to make a measurement. If a device isn’t calibrated for darker skin tones, the pigmentation could affect how the light is absorbed and how infrared technology works.
“Skin tone will interfere with the reflection of light,” said Celi, who has also studied disparities in pulse oximeters. “We think that this is just the tip of the iceberg. I suspect that over the next year, we will have an abundance of papers showing how there are products that are in use for care that are themselves not objectively accurate across different demographics.”