Pulse Oximeters Need to Work Better for Dark-Skinned Patients, FDA Panel Says

Pulse oximeters are less accurate in patients with darker skin, and the FDA should notify patients and providers about the issue and recommend that manufacturers correct the discrepancy, an FDA advisory panel said.

"I think the clinical evidence does demonstrate disparate performance in patients with darker skin pigmentation; we saw a lot of studies and everything pointed to the same," Steve Nathan, MD, chair of the Anesthesiology and Respiratory Therapy Devices Panel of the FDA's Medical Devices Advisory Committee, said Tuesday at a meeting to discuss ongoing concerns that pulse oximeters may be less accurate in individuals with darker skin pigmentations. "In my mind, I think the evidence we saw was quite clear."

Considerations for OTC Devices

In addition to the pulse oximeters that are used in the hospital, "a very large percentage of the population comes in contact with those inexpensive, non-medical pulse oximeters that they buy from department stores and grocery stores and pharmacies, and they rely on those to decide whether or not they should present to the physician," said Murad Alam, MD, professor of medical social science at Northwestern University in Chicago. "You have to really think about those as well." Committee members agreed that even though the studies they reviewed looked only at in-hospital devices, the non-prescription devices were likely to also have the same issues.

Panelists heard from several researchers who discovered a discrepancy between in-hospital pulse oximeter readings in white patients and those in Blacks, Hispanics, and other darker-skinned racial and ethnic minorities. Michael Sjoding, MD, of the University of Michigan Medical School in Ann Arbor, outlined his team's study of electronic health records for 1,333 white patients and 276 Black patients, in which the researchers compared pulse oximetry (SpO₂) readings with arterial blood gas measurements (SaO₂) taken within 10 minutes. They found that although occult hypoxemia was missed in 3.6% of white patients, it was missed more often among Black patients (11.7%).

"There was a significantly higher rate of occult or hidden hypoxia where the pulse oximeter was reading normally but the actual arterial oxygen saturation was less than 88%," he said. "The pulse oximeter was less able to identify low oxygen levels in Black patients compared to white patients."

Not everyone was convinced by the research, however. "As an editor of a journal, if any of these studies ... were submitted to my journal as an evaluation of pulse oximetry, I would have rejected all of them, because the time of the recording of the SpO₂ value to the recording of the blood gas value was anywhere from 5 to 30 minutes or closest one available in [the] electronic medical record," said Richard Branson, MSc, director of clinical research at the University of Cincinnati College of Medicine.

"I'm not saying that this is not real, but does it make sense that there are these large disparities," said Branson.

Nathan disagreed, saying, "How much can it change in 10 minutes? In some cases it might, but for the most part it's relatively stable."

Comments From Outside Groups

The committee also heard from people representing provider organizations and other healthcare groups. Eric Gartman, MD, speaking on behalf of the American College of Chest Physicians, noted one 2022 study of COVID patients looking for eligibility for COVID-specific treatment -- which often is based on pulse oximetry -- that found that Black and Hispanic patients were roughly 2% to 30% less likely to be offered COVID-specific treatment, "presumably because they were reading higher normal oxygen levels than they actually had."

"Particularly, what we see a lot is the qualification for supplemental oxygen, qualification for supportive equipment or disability parking placards -- do we know how severe a patient's disease is, and whether they're progressing or not; can we trust that pulse oximeter reading that we're seeing?" he added. And when it comes to sleep apnea testing, "the pulse oximetry criteria are very much tied into the definition of whether disease is present or not, and how bad it is. This is potentially a big issue for sleep medicine as well."

Michael Abrams, MPH, PhD, a senior health researcher at the Public Citizen Health Research Group, said the pulse oximeter issue "exemplifies two ongoing challenges: the issue of racial equity and the deficiencies of the 510(k) pathway for clearing medical devices." Currently, the FDA recommends that manufacturers submitting a pulse oximeter for approval under the 510(k) pathway include desaturation studies with 10 or more healthy subjects that vary in age and gender, including at least two darkly pigmented subjects or 15% of the study group, whichever is larger. "Lack of diversity in clinical trials has been a longstanding problem," Abrams said.

Manufacturer representatives also appeared before the committee. Sam Ajizian, MD, chief medical officer of patient monitoring at Medtronic, maker of pulse oximeters and other medical devices, said that Medtronic "stands behind its devices but acknowledges that more can be done to improve accuracy." He added that "currently, manufacturers rely on a variety of subjective scales to assess the impact of skin pigmentation on device performance. In order to improve device accuracy for everyone. We must replace these subjective scales with a uniform objective method of assessing skin pigmentation."

Pigmentation Versus Race

The panel also discussed whether the FDA should tailor its future guidance on pulse oximeters to address particular racial and ethnic groups, or limit its guidance to discussions of skin pigmentation; in the end, they decided the agency should address both. "Are we moving away from ethnic implications to more of a pigmentation [issue]?" asked Veverly Edwards, assistant professor of teaching at the University of Tennessee and the panel's consumer representative. "The African American community -- the disparities in care is how we got here. I just wanted to make sure it wasn't lost."

Branson said that he agreed that health disparities was an issue for Black patients, but added, "In the past, race-based differences were exploited by white supremacists to suggest there's actually a difference between us other than the color of our skin, so I think we have to be careful there. Objective measures are for science, but I do appreciate what she says that race definitely has to be included."

The panel also recommended that the FDA include a message to consumers buying over-the-counter pulse oximeters indicating that the device's accuracy has not been verified by the FDA, and also that various factors such as use of nail polish, skin pigmentation, and patient movement can affect the oximeter's accuracy. "When I get a prescription, they give me a whole big sheet of things the doctor was supposed to tell [me] about the side effects," said Robert Loeb, MD, professor of anesthesiology emeritus at the University of Florida. "I think the layperson should know that these are the things that affect [device] accuracy, including pigmentation."

The panel also suggested broadening the recommended number of patients -- overall as well as those of color -- in clinical trials, and both sick and healthy patients should be included. "There should be equal representation across the spectrum of skin pigmentation," Nathan said. And panel members agreed that the "Arms" -- a statistical calculation used to determine pulse oximeter accuracy -- needed to be tightened from the currently recommended ≤3% to ≤1.5% or 2%.

While the FDA is not required to act on the recommendations of its advisory committees, it usually does.

Comment