This device is crucial in the fight against COVID-19. It may not work on dark skin

In the fight against CVOID-19, the pulse oximeter has been an essential tool for doctors and other medical professionals. But the small device that monitors oxygen levels may not work well for people with dark skin, according to multiple studies and government health agencies.

Pulse oximeters, which have increased in use during the pandemic, may yield inaccurate results, the US Food and Drug Administration warned Friday.

Earlier in week, the US Centers for Disease Control and Prevention also updated its coronavirus clinical guidance to warn doctors and nurses that data from several studies suggest skin pigmentation can affect the accuracy of the devices.

“While pulse oximeters may be useful for estimating blood oxygen levels, these devices have limitations that can result in inaccurate readings,” Dr. William Maisel, director of the Office of Product Evaluation and Quality in the FDA’s Center for Devices and Radiological Health, said in a statement.

Pulse oximeters are small clamp-like devices that attach painlessly to a patient’s finger and constantly monitor the amount of oxygen in their blood. COVID-19 is a respiratory illness, meaning it attacks the lungs first, so low oxygen levels are a sign a patient may be getting worse. CDC data shows Black, Latino and Native Americans are four times more likely to be hospitalized with Covid-19 than others.

The recent warnings come after a study published in the New England Journal of Medicine in December. Dr. Michael Sjoding and several colleagues from the University of Michigan analyzed data from over 10,000 patients. For each patient, they compared the oxygen levels recorded by a pulse oximeter to those measured by arterial blood gas — a much more accurate, but painful and invasive, procedure.

The researchers wanted to know how often the pulse oximeter was showing a relatively normal oxygen level when it should have been registering something more concerning.

In White patients, the pulse oximeter gave a misleading number 3.6% of the time. In Black patients, it was 11.7% of the time.

The takeaway, Dr. Sjoding says, is that pulse oximeters were three times as likely to miss significantly low oxygen levels — or hypoxemia — in Black patients. The study suggests one in every 10 Black patients may be getting deceptive results.

Why the readings aren’t precise

There’s a relatively simple explanation for why. Pulse oximeters work by sending two types of red light through your finger. A sensor on the other side of the device picks up 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. If the device isn’t calibrated for darker skin, the pigmentation could affect how the light is absorbed. Dark nail polish can cause a similar effect.

Experienced physicians don’t rely on the pulse oximeter alone to make a diagnosis or decide a patient’s treatment, says Dr. Michelle Ng Gong, chief of Critical Care Medicine and the chief of Pulmonary Medicine at Montefiore Medical Center in New York.

“They would never tell a patient that I don’t care how badly you feel, as long as that number is OK, don’t worry about it,” says Dr. Gong. “It is a tool. And as a tool, we need to be able to use it properly in the context of other information.”

But, Dr. Gong says, during the pandemic, when hospitals are overwhelmed, and doctors who may not normally be in the emergency room are being brought in to see and triage patients, numbers from a pulse oximeter may hold more weight. An inaccurate reading could be particularly problematic if the patient’s oxygen level is borderline.

Doctors also need to be aware of their own unconscious bias, Dr. Gong says. Studies show, for example, when African Americans complain of pain, they are less likely to be given as much (or any) pain medication as Whites. If a patient says they are having trouble breathing, physicians shouldn’t dismiss them just because a pulse oximeter reading is normal.

“The only way health disparities can be reduced,” Dr. Gong says, “is if we attack it from multiple ends, from both the larger contributions as well as our personal interactions.”

Doctors are also not the only ones using pulse oximeters. As hospitals reached capacity during the pandemic, some Covid-19 patients with less severe symptoms were sent home to monitor their progress. According to market research, sales of at-home pulse oximeters have increased 500% since the novel coronavirus arrived in the US.

Amy Moran-Thomas, a professor of anthropology at MIT, started researching pulse oximeters last year when her husband was sent home with one.

The findings aren’t new

Moran-Thomas uncovered studies going back to the 1990s that suggested there was a problem with pulse oximeters in darker-skinned patients.

In 2005, a study done at the University of California, San Francisco’s Hypoxia Lab, found that three different models of pulse oximeters overestimated oxygen levels in dark-skinned patients. They did a follow-up study in 2007 with similar results.

“A reading of 77 like my husband’s could hide a true saturation as low as 69 — even greater immediate danger. But EMTs or intake nurses might not be able to detect those discrepancies. The number appears objective and race-neutral,” Moran-Thomas wrote in Boston Review.

Moran-Thomas wondered why the problem, identified decades ago, hadn’t already been fixed.

“I am a pulmonary and critical care physician,” says Dr. Sjoding, who began his study after reading Moran-Thomas’ article. “One of our coauthors is a prominent Black physician from the University of Michigan. None of us knew this. None of us knew about these studies from the mid 2000s. It was not a part of our training.”

Public attention seems to be picking up. On January 25, several US senators sent the FDA a letter urging them to look into the issue. The FDA is responsible for approving any medical pulse oximeters before they go to market. The agency currently requires pulse oximeters to be tested on a variety of skin tones, meaning “at least 2 darkly pigmented (test subjects) or 15 % of the subject pool, whichever is larger.”

The FDA says it is carefully reviewing the available data to determine whether additional guidance or studies from the agency are needed. It recently approved a smartphone pulse oximeter that claims to work on all skin tones, using a 10-20 second automatic calibration period to accurately measure how light is moving through the user’s skin.

“These devices are not intended to be the sole or primary use of information to make a clinical diagnosis or treatment decision,” Dr. Maisel told CNN in an interview Wednesday. “Someone shouldn’t overly rely on a pulse oximeter reading, even if it’s the most accurate product.”

“Just make sure you’re aware of that when you’re making decisions,” says Dr. Sjoding. “The pulse oximeter reading could be a couple of points off. And if that is the case, would you care for this patient in a different way?”