FDA panel asks for improvements in pulse oximeters

A Food and Drug Administration advisory panel suggested Tuesday that the agency improve how it regulates pulse oximeters, calling for clearer labeling and more rigorous testing of the devices. The widely used instruments monitor blood oxygen levels and have been shown to work less well on patients with darker skin, possibly exacerbating health disparities in many racial and ethnic groups. 

Panel members, even those who had initially been skeptical of the validity of those recent studies, said after reviewing the data they were persuaded that the devices were indeed less accurate on patients with darker skin and could lead to poorer health outcomes. “I am now convinced,” said Robert Loeb, a professor emeritus of anesthesiology at the University of Florida. “There is a problem and it has clinical significance.” 

No formal votes were taken; the panel met to hear from researchers and physicians about the issue and to offer guidance to FDA officials who will take the panel’s recommendations into account as they update regulations.

Many panelists and presenters were concerned that validation of the devices requires only small studies using a handful of healthy patients, few with dark skin, in carefully controlled lab studies that don’t correlate well to how the devices actually work in sick patients.

“Enlisting healthy volunteers is like a Phase 1 study, but there’s no Phase 2 or Phase 3 studies,” said Steven Nathan, medical director of the advanced lung disease and transplant program at Inova Fairfax Hospital in Falls Church, Va., and chair of the FDA’s anesthesiology devices panel. 

Current FDA guidelines require studies of just 10 patients, of which two, or at least 15%, must be darkly pigmented. Such studies, many panel members said, may be too small to determine how well the devices work in people with darker skin. 

“The sample size seems concerningly small for devices that affect the health of millions of patients,” said David Sterken, a hospitalist at the University of Wisconsin who spoke before the panel, which met virtually for eight hours.

The panel reviewed the many reasons pulse oximeters may not work as well in sick patients, including poor circulation caused by many medical conditions. Tests done to meet FDA requirements typically use healthy volunteers and techniques to get the best readings, including warming the fingers of test subjects and blocking ambient light. Many experts who addressed the panel said they were concerned FDA testing requirements radically differed from the real-world ICU and ER settings where the devices are relied on to guide clinical decisions.

“We wouldn’t tolerate that for a medicine, so I’m not sure we should tolerate that in a device,” said Eric Gartman, an associate professor of medicine at Brown University’s Warren Alpert Medical School, who spoke on behalf of the American College of Chest Physicians. Nathan, the panel’s chair, noted that drugs such as the Covid treatment remdesivir had much stricter standards for approval than the device that physicians used to determine whether or not to administer the drug. 

The FDA has requested new studies to assess pulse oximeter accuracy in hospital settings, for both adult and pediatric populations. Many panel members said future studies should contain far more patients with dark skin and that data for those patients should be analyzed separately so any issues were apparent and not lost when data from all patients in a study were pooled.

One reason for the relatively lax standards is that pulse oximeters are approved under the 510(k) review process, which means manufacturers must only prove their devices are as safe and effective as devices already legally marketed. This gives “manufacturers a reckless fast-track pathway to market,” said Michael Abrams, a senior health researcher for Public Citizen, which has been advocating for decades that the FDA include more racial and ethnic diversity in clinical studies. 

Many panelists agreed that current regulations that allow oxygen readings to be inaccurate by up to 3% on average should be tightened in the future to 2% or even 1.5%. Some members urged setting higher accuracy goals for the future — as the government does for auto manufacturers with gas mileage — so that device makers could work toward meeting those goals. William Wilson, an industry representative on the panel who serves as chief medical officer for pulse oximeter manufacturer Masimo, said he thought makers could easily meet a goal of 2% now and a goal of 1.5% in a few years. He said his company supported raising accuracy standards and testing devices on more patients with dark skin. 

Sam Ajizian, the chief medical officer for Medtronic, which also makes the devices, said any errors in readings in people with darker skin when using his company’s devices were small and not clinically relevant, but added, “We acknowledge more can be done to enhance accuracy and promote health equity.” 

Much of the day’s discussion centered on the differences between pulse oximeters used in medical settings and consumer devices. Medical-grade devices are regulated more strictly; the $10 to $25 pulse oximeters sold in drugstores or online face less scrutiny because they are considered to be for recreational use — something many on the panel found ridiculous. 

“I know a lot of people who have pulse oximeters at home. I don’t know anyone who uses it for recreational purposes. They are used for medical decision making,” said Arlene Hudson, who chairs the department of anesthesiology at the Uniformed Services University in Bethesda, Md. 

“These are not health and wellness items. This is not a toy,” added Murad Alam, vice chair of the department of dermatology at Northwestern University. “This is a misclassification problem.” 

Jessica Cocolin, a certified registered nurse anesthesiologist with a child with a congenital heart defect, added a poignant note to the discussion when she told the panel how often she had used a consumer device to make decisions about her child’s health. ”It literally meant life and death,” she said. 

The panel weighed several suggestions for consumer devices, including requiring that they read out a range, rather than a number that could give patients “a false sense of security;” putting them behind drugstore counters so pharmacists could counsel those who bought them about issues with use on darker skin; or issuing sterner warning labels that advise people to seek medical attention if they have trouble breathing despite what the devices say. Nathan said he would recommend that the FDA require manufacturers label their consumer devices “not for medical use.” 

But a number of panelists disliked the idea of relying on warning labels, saying few consumers do, or even can, read them. “I hope we don’t have another label that’s 19 pages long and you need a magnifying glass,” said Joseph O’Brien, the president of the National Scoliosis Foundation and the panel’s patient representative. He said he also disliked the idea of advising patients with darker skin to purchase more expensive and more accurate medical grade devices, which would cause financial disparities.

Others said a better solution would be to ensure the consumer devices are more accurate. “I think it sends a mixed message to allow companies to sell something but not have confidence in the results it puts out,” said Sean Hennessy, a pharmacist and epidemiologist at the University of Pennsylvania School of Medicine who served on the panel. 

A series of eight physicians who spoke to the panel representing their specialty medical societies expressed concern about the devices they rely on, and some frustration that the discrepancies in readings in darker skin patients had been known for decades and not corrected. Amal Jubran, who spoke on behalf of the American Academy of Sleep Medicine and published one of the first studies showing the discrepancy in 1990, said the devices had changed markedly since her publication, “but the inaccuracy in Black patients has not lessened.” 

The panel weighed several knotty issues, from how to most accurately measure skin tone, which is only indirectly related to race, to whether device makers should use algorithms to “correct” for race — a fix that is being discarded in other areas like assessing kidney disease because it has increased, rather than decreased, disparities. They also disagreed on whether physicians should rely more heavily on using arterial blood draws to better assess oxygen levels in patients with darker skin. 

Many physicians said they will remain conflicted about how much they can trust the devices until the FDA takes action to remedy the problem. “Earlier today, when I applied a pulse oximeter to a patient in the operating room, I paused,” said Jesse Ehrenfeld, an anesthesiologist who spoke on behalf of the American Medical Association, where he is president-elect. “I asked myself how can I know that this FDA-approved device will give me an accurate reading during the critical moments ahead in surgery?”

This is part of a series of articles exploring racism in health and medicine that is funded by a grant from the Commonwealth Fund.

Get your daily dose of health and medicine every weekday with STAT’s free newsletter Morning Rounds. Sign up here.