The Food and Drug Administration convened an advisory panel on Friday to discuss concerns that pulse oximeters may be less accurate in people with dark skin pigmentation.
The agency asked panelists for their thoughts on new clinical trial requirements to ensure the devices are accurate across all skin tones and whether prescription and over-the-counter pulse oximeters should be held to the same standards.
The scope of the problem
The FDA issued a warning in 2021 that pulse oximeters pose a “risk of inaccuracy under certain circumstances,” and a group of attorneys general last year urged the agency to “act with urgency” to address the devices’ race and color bias.
Edward McClure, a patient with chronic obstructive pulmonary disease, told the panel that sometimes it would take up to three pulse oximeter readings to get a number that seemed correct. McClure’s daughter offered a potential explanation for the faulty results: she had read about a study that found pulse oximeters were more likely to provide inaccurate readings for people with melanated skin.
“As I mentioned that to my many medical professionals … none of them except for my African American primary care physician had even heard of the problem. So that is a big, big, big, big problem in terms of the risk that people are being placed in,” McClure said.
FDA officials cited studies that demonstrate the scope of the problem. In six of seven systematic reviews published between 2020 and 2023, the agency found records of over-estimates in people with dark skin pigmentation.
Those discrepancies translated to real-life consequences.
Ashraf Fawzy, a pulmonary and critical care physician at Johns Hopkins Medicine, said in two retrospective studies of COVID-19 patients, pulse oximeters overestimated oxygen saturation by .93% to 1.2% in Black individuals compared to white individuals. Researchers found that Black patients accounted for over half of the patients whose need for COVID-19 treatment was never recognized.
“We urge the panel and FDA to require pulse oximeters be tested on a diverse population using objective measurements of skin tone and actual clinical data in real world settings to ensure adequate and equitable performance when making important clinical decisions,” Fawzy said.
Clinical trial recommendations
The FDA panel discussed recommendations that clinical trials of pulse oximeters must include at least 24 participants, an increase from the previous requirement of 10.
The panelists agreed on a recommendation that participants in clinical trials should span the Monk Skin Tone (MST) scale, a measurement developed by Ellis Monk, an associate professor at Harvard University, that is designed to represent a broader range of skin tones.
Currently, many studies use the Fitzpatrick scale to measure skin pigmentation. The Fitzpatrick scale was developed by a dermatologist in the ‘70s to measure the UV sensitivity of different skin types and skews toward lighter skin tones.
As a second form of measurement, the FDA also recommends using the Individual Typology Angle (ITA) at the sensor site. The panelists agreed with using these measurements but recommended requiring larger clinical trials.
Rachel Brummert, a consumer representative on the panel, said 24 patients is just the “bare minimum,” and suggested doubling the requirement.
The FDA also is considering a non-disparate performance requirement to ensure the devices don’t perform differently based on skin pigmentation.
Scott Lucas, vice president of device safety at ECRI, supported using MST and ITA but recommended doing a pilot test of the MST protocol to see if users can get consistent readings on the same patient. Lucas also recommended using colorimetry devices that have been validated across a diverse group of users for ITA.
“It wasn’t long ago in the height of the pandemic, before COVID tests were readily accessible, when these pulse ox devices were critical in determining when a patient was in grave danger, and device performance issues led directly to patient harm,” Lucas said. “So we must keep in the forefront of our minds that this is a significant health equity issue. The color of a patient’s skin should never degrade the quality or the effectiveness of the tools that healthcare providers use to give life-saving care.”
Requirements for OTC devices
Another of the FDA’s questions was whether the panelists thought over-the-counter pulse oximeters should meet the same requirements as prescription-use devices. The panelists agreed they should have to meet the same criteria and debated how to reflect that on the devices’ labeling.
Anne Whitney Brown, a pulmonologist and voting member on the panel, recommended holding them to the same standards because for people at home, a pulse oximetry reading might be the main deciding factor in whether to go to the emergency room, while in a clinical setting, other variables are taken into consideration.
“If it doesn’t pass this criteria, there’s a big label that says, ‘This has never been tested. It is not considered appropriate for medical purposes,’” said Jeffrey Feldman, a non-voting member and a professor of clinical anesthesiology and critical care at the University of Pennsylvania School of Medicine. “My recommendation would be for the purpose of medical use, anything that doesn’t go through that [criteria] gets labeled very clearly.”
Brummert, the consumer advocate, said that patients are unlikely to read a label or box warning, and recommended keeping them behind the pharmacy counter, where a pharmacist could explain the difference.
“A consumer is just going to walk into a CVS or … buy it on Amazon. They’re not going to read the labeling. They just know that this is something they can put on their finger and they’re going to get some sort of result,” Brummert said.
The FDA doesn’t usually decide if products should be put behind the pharmacy counter, but that decision could be determined at the state level, said Malvina Eydelman, office director for CDRH’s office of health technology for ophthalmic, anesthesia, respiratory, ENT and dental devices.
Industry comments
Medtronic and Masimo, both of which sell pulse oximeters, agreed with many of the suggested changes. They supported using a validated scale, such as the MST, and the minimum sample sizes recommended for studies.
They also supported using non-disparate bias to evaluate the performance of pulse oximeters with respect to skin pigmentation but raised some questions about the specifics.
Steven Barker, Masimo’s chief science officer, said the accuracy requirements for when a patient’s oxygen saturation is below 85% are too broad, as these patients are at a higher risk of hypoxia, and recommended tightening the range.
Meanwhile, Sam Ajizian, Medtronic’s chief medical officer for its patient monitoring operating unit, recommended further inquiry and data collection before setting performance thresholds.
When asked if Medtronic would consider recalling its pulse oximeters, Ajizian said the company’s devices conform to current FDA standards.
“Currently, we think that if there was a recall by any manufacturer around this, it would undermine public safety because this is a foundational device in operating rooms and ICUs, ERs and ambulances, and everywhere,” Ajizian said.
At the same time, he discouraged pulse oximeter results alone being used as a threshold for making care decisions.
Ajizian added that Medtronic has been working to educate its users about the issues around skin pigmentation, but believes the benefits of the products “far outweigh” the risks.