Since the start of the pandemic, the popular medical device has been indispensable for measuring oxygen in the blood, but dangerously inaccurate in patients with darker skin. Now the FDA is in the hot seat and Black researchers are working on a cure.
At the University of Maryland Medical System, which has 11 hospitals in and around Baltimore, about half the patients are Black. So as discussion about how medical devices called pulse oximeters—the plastic hinges that clamp onto your finger to measure the oxygen content of your blood—don’t work as well for Black people and other people of color came to the forefront last year, the hospital system decided to assess the scope of the problem. The results were mindblowing: In the 12 months that ended in July, the hospitals estimated that they’d sent home 1,012 African Americans from emergency rooms who had potentially life-threateningly low oxygen levels in their blood despite normal readings from pulse oximeters.
Dr. Roderick King, the system’s chief diversity, equity and inclusion officer, was shaken but not surprised. “Once we realized the inaccuracy of the pulse oximeter, we knew the number was going to be big,” he told Forbes.
Pulse oximeters are used in every hospital in America and, since the start of the pandemic, in a lot of homes, too. But darker skin, calloused hands, cold fingers or even dark nail polish can cause inaccurate readings. A patient could have low blood oxygen, a symptom of worsening Covid or serious lung disease, and the device might still display a number within the acceptable range. The threshold for whether a patient is considered to have too little oxygen in their blood can be a deciding factor in whether someone suffering from lung disease receives necessary oxygen at home, as well as whether insurance will cover the costs of any treatment.
For any patient, especially one with symptoms bad enough to drag themselves to the ER, a missed diagnosis is terrifying. If the University of Maryland hospital system, which sees about 500,000 patients in its emergency rooms each year, sent home an average of nearly three patients every day due to faulty readings, how many more people across the country are being denied treatment they need? For how many of those people did the undiagnosed symptoms turn into something worse? And how do we make sure more people aren’t harmed by the flaws of a diagnostic tool that’s been called the most common medical device on Earth aside from the thermometer?
Particularly galling is the fact that researchers have known for years that pulse oximeters are less accurate on people with darker skin because of how melanin interferes with light absorption. But it took the Covid-19 pandemic and a series of damning studies for the medical community and the Food and Drug Administration, whose job it is to ensure that Americans don’t die due to product flaws, to even start figuring out a solution.
Studies already show that Black patients’ health complaints are more likely to be ignored or minimized by health providers, said Dr. Uche Blackstock, founder of Advancing Health Equity. This just makes things worse. “I think we haven’t fully appreciated the magnitude of what these inaccuracies have done in terms of the lives harmed,” she said.
Now, engineers at Brown University and Tufts University are working separately to redesign the pulse oximeter so that it’s accurate for everybody. Major manufacturers of pulse oximeters used in hospitals, including Medtronic (maker of the Nellcor brand) and Masimo, are tweaking their products while touting their own studies they say show any racial differences are too small to matter clinically. And the Food and Drug Administration, which held a public hearing in November, is in the hot seat to improve its approval process and to explain how so many cheap versions of the device, many made in China and available on Amazon, have flooded the market without oversight.
A remedy is most likely a long way off. In the meantime, pulse oximeters continue to be popular and relied upon by hospitals, nursing homes and families across America, and, increasingly, the world. Global 2022 sales of $2.7 billion are expected to balloon to $4.8 billion by 2030.
When Kimani Toussaint, a Brown University optics researcher, learned from his wife, an epidemiologist, about racial-bias issues with pulse oximetry, he was shocked. “I was actually very surprised that in 2020 this was an issue that had not been solved,” said Toussaint, who is Black and serves as the school of engineering’s senior associate dean. “This is the challenge of the lack of diversity in STEM areas. … Irrespective of intent, not having the right people in the room in terms of designing and developing a technical solution can have unintended serious consequences.”
A December 2020 study by doctors at the University of Michigan is generally credited with opening a lot of eyes to the pulse oximeter’s shortcomings. It showed that Black patients were nearly three times more likely than white patients to have “hidden hypoxia,” or low blood-oxygen levels that the devices didn’t detect. Subsequent studies have shown that patients with darker skin receive less supplemental oxygen during stays in the intensive care unit, that hidden hypoxia is also a problem for Black patients in non-ICU hospital settings, and that bad oximeter readings may have led to delayed Covid-19 treatment for Black and Hispanic patients.
Preliminary research from the University of California at San Francisco shows that in hospitalized patients with dark skin and reduced blood flow to the fingers, which can be caused by low pressure, anemia, shock or even just cold hands, the combination was enough to miss a diagnosis of low blood oxygen in about 20% of pulse-oximeter readings.
BY THE NUMBERS
Pulse Oximeters And Racial Bias
Sources: Research and Markets, FDA, University of Michigan study, University of California San Francisco preprint, University of Maryland Medical System
Notes: Testing requirement is for at least two people, or 15%, whichever is larger. Pulse oximeter market and UMMS discharge numbers are estimates.
To find a remedy, Toussaint is working with graduate student Rutendro Jakachira on a new type of pulse oximeter that plays with the electrical properties of light so that sensors will be able to work on patients of all skin tones. He has partnered with physicians at the Miriam Hospital in Providence, R.I., to do a pilot clinical study. While it could take years to get through the regulatory process, Toussaint said that he hopes the FDA will consider racial bias an urgent issue and move quickly. “It really is an emergency,” he said.
Valencia Joyner Koomson, an associate professor of electrical and computer engineering at Tufts University, is also working on a new oximeter design. Koomson, who is Black, noted that the issue goes beyond melanin and to all the variables that can affect a pulse oximeter’s performance, such as calloused hands. If the device’s light signal never reaches the blood it’s impossible to get an accurate reading, she said. “It’s similar to the loss of signal for a cellphone. We’re trying to design a device that can adapt.”
While her research is also in its early stages, Koomson said she intends to file patents on her technology with hopes of commercializing the research. While the average American might know about pulse oximeters because of the pandemic, their uses go beyond any one disease. Readings can impact whether someone suffering from lung disease receives necessary oxygen at home, for example, or whether insurance will cover the costs of treatment. Koomson is working with a data-science student to look at how pulse oximeter errors affect maternal health based on research data from Beth Israel Deaconess Medical Center in Boston.
The efforts of Toussaint and Koomson are part of a broader push to begin to fix long-standing inequities in healthcare that go beyond medical devices. Facilities like the University of Maryland Medical System have begun ditching a race-based algorithm for scoring kidney disease that made Black patients appear healthier than they were and resulted in denying them transplants. While that’s a different issue than the one with pulse oximeters, it’s why some researchers are wary of advocating corrections to existing pulse oximeter readings for dark-skinned patients.
“Race-based corrections lead to a lot of confusion later on that exacerbates bias,” said Achuta Kadambi, an assistant professor of electrical and computer engineering at the University of California at Los Angeles, who’s working on more equitable remote heart monitoring. “It’s a slippery slope. You want the physics of the device to be sensing things in a manner that’s robust.”
Since the alternative is a painful blood draw from an artery, measuring blood oxygen with a clip-on device is a great idea.
Early non-invasive blood-oxygen measurement was designed with equity in mind. As MIT anthropology professor Amy Moran-Thomas wrote in Wired in 2021, Hewlett-Packard worked with NASA in the 1970s to create devices for astronauts “where the precise measurement of oxygen played a vital role.” As the company expanded into the hospital market, it designed sensors that sat on top of a patient’s ear and took into account different skin tones using eight different wavelengths of light.
Eventually, H-P pulled back from medical devices to focus on the emerging and more lucrative manufacture of personal computers, and other companies relied on then-new technology of pulse oximetry, developed by Japanese researcher Dr. Takuo Aoyagi in 1974. Companies that include Nellcor, which Medtronic owns and plans to spin off as part of a new company, Nonin Medical and Masimo, founded by Iranian immigrant Joe Kiani developed fingertip devices for the hospital market. The companies say that any differences in how their FDA-approved technology works with different skin tones are too small to be clinically significant. Masimo chief medical officer William Wilson, citing results from a recent study by the company, said, “It’s not perfect, but it’s close to perfect.”
Even so, the big makers of hospital-grade pulse oximeters said they’re continuing to work on improving accuracy. Sam Ajizian, Medtronic’s chief medical officer of patient monitoring, said in an emailed statement to Forbes that “we believe that more can and should be done to continuously improve the accuracy for those patients with darker skin tones.”
Only hospital-grade medical devices need FDA approval. Over-the-counter pulse oximeters, available online for as little as $14.99, use similar technology but are largely unregulated. So, too, are Apple Watches and other fitness gizmos that often include pulse oximetry and may be used as an early warning on whether to see a doctor or what issues to discuss during a telehealth appointment.
A month after the University of Michigan study hit the internet with a splash, three U.S. senators — Elizabeth Warren (D-Mass.), Ron Wyden (D-Ore.) and Cory Booker (D-N.J.) — wrote to the FDA asking for a review.
In February 2021, the agency alerted physicians and consumers that the devices may be inaccurate in cases of “poor circulation, skin pigmentation, skin thickness, skin temperature, current tobacco use, and use of fingernail polish.” In November, it convened a virtual public meeting of the agency’s medical-devices group to investigate the pulse oximeter’s racial disparities. The FDA is now reviewing the committee’s recommendations, and has called for health providers to be aware of the devices’ limitations in making diagnoses and treatment decisions.
Industry watchers told Forbes to expect the FDA to move slowly and in coordination with international efforts. But there’s one way the agency can make medical devices in general work for all patients — change the requirements for testing.
The FDA requires just 15% of study participants, or two people, to have dark skin, well below a number that would be representative of American demographics. There are no rules on how to quantify pigmentation, an issue that’s crucial for pulse oximeters because of how the light-based technology works, or doesn’t, along a spectrum.
Dr. Philip Bickler, director of UCSF’s Hypoxia Research Lab, argues that the requirement should be at least 30% and needs to include people with a variety of skin tones, including those who are Asian, Hispanic and Black, and that there needs to be a way to measure and categorize that skin tone. “This is fundamentally a skin-pigment issue,” he told Forbes. “There’s a problem with some of the studies in recent years that skin pigment is inferred from self-reported race.”
Masimo’s Wilson and Nonin CEO John Hastings both said they would like to see increased requirements for inclusion of dark-skinned patients in testing. They’d also favor some form of oversight of over-the-counter pulse oximeters. “We think one of the next steps is going to be guidance from the FDA,” Hastings said.
When it comes to over-the-counter pulse oximeters for home use, the FDA considers them general wellness products, and as the agency warns, “should not be used for medical purposes.”
“What does it mean to buy a pulse oximeter for recreational use?” Tufts’ Koomson said. “Is it a toy? I think our regulators and government need to give a bit more scrutiny here.”
While this plays out, doctors across the country are left to figure things out on their own. “What should my residents do when they’re working on a patient who identifies as Black?” said Dr. Utibe Essien, an assistant professor in medicine at UCLA who studies racial and ethnic disparities. “Should they get another blood test or another image to be certain?”
At the University of Maryland Medical System, Dr. King and his colleagues are grappling with the issue while doing further research. They had hoped to study the differences between the variety of models of pulse oximeters that their hospitals use, but didn’t get the funding. Instead, they’re working with electronic medical records firm Epic Systems to get a better sense of inequities across millions of patients.
“I expect a lot of finger pointing as we lead into this,” King said. “I don’t want to get caught up in ‘this pulse oximeter works better than that pulse oximeter.’ We probably need to replace all of them.”
MORE FROM FORBES
Credit: Source link