A new study by Johns Hopkins University School of Medicine and Baylor College of Medicine indicates that inaccurate pulse oximetry readings for black and Latino patients with the emerging coronavirus in the United States are associated with significant treatment delays or ineligibility for certain treatments.
Although it has been a concern for several decades, awareness about the overestimation of oxygen levels in dark-skinned patients with inaccurate pulse oximeter readings has grown during the COVID-19 pandemic. The new study, published Tuesday in the JAMA Internal Medicine, adds to a growing body of research investigating racial and ethnic biases and contrasts and their impact on patients receiving care.
Among the most surprising findings was the fact that about 25 percent of people should have received more aggressive treatments but did not, and the majority of these individuals were black or Hispanic, according to co-author Dr. Tianshi David Wu. Additionally, the pulse oximeter never set the oximetry value at or below 94 percent for a significant proportion of people, more than half of whom are black.
“Social determinants of health, racial disparities, and outcomes of COVID-19 are well described,” said Wu, assistant professor of medicine in pulmonary and critical care at Baylor, in a phone interview with CTVNews.ca on Monday.
“People who are considered to be a racial or ethnic minority have a higher likelihood of dying from COVID-19 and we speculate that this type of technology bias may fundamentally contribute to some aspects of this difference.”
Wu and his colleagues wanted to study the problem in the context of COVID-19 as it relates to treating the disease and how pulse oximeters are used everywhere in the world.
“Unlike other diseases, the classification of severity and COVID is based on the value of the pulse oximetry or the oxygen saturation number,” Wu said.
“For this reason, our treatment decisions are overly dependent on this oxygen saturation number…and you can’t pass into the hospital without measuring oxygen saturation with a pulse oximeter.”
Wu noted that the National Institute of Health and other international guidelines, along with prescribing information such as remdesivir and dexamethasone, all base their eligibility on either a diagnosis of severe COVID-19 or a low value of pulse oximetry.
A retrospective cohort study, which collected data from five hospitals within the Johns Hopkins Health System, found persistent overestimations of arterial oxygen saturation among Asian, black and Hispanic individuals.
“When comparing pulse oximetry measurements with arterial oxygen saturation, approximately one-third of patients from each racial or ethnic minority group had at least one unspecified episode of hypoxia compared to less than one-fifth of the white patients,” the authors wrote in the study. Hypoxia occurs when there is not enough oxygen in the tissues.
“In addition, we found a systematic failure to identify black and Hispanic patients who were most likely to be eligible to receive COVID-19 treatment and a statistically significant delay in recognizing the recommended guideline threshold for initiating treatment among black patients compared to white patients. “
How do pulse oximeters work and why can they be inaccurate?
A pulse oximeter is a non-invasive method for monitoring the amount of oxygen in a person’s blood that can quickly detect changes in those oxygen levels. It has become a particularly useful tool in patient care decisions during a pandemic, with recommendations for hospitalization and licensing of certain drugs based on specific thresholds measured by oximeters.
The device is usually attached to a finger, although other parts of the body such as the earlobe can also be used. First developed in the 1970s in Japan, the device typically works by sending LED light through the skin at two different wavelengths and analyzing how much light is absorbed and passed to the other side at those wavelengths. Melanin, the pigment found in darker skin, tends to absorb more of the light that travels through the skin.
The use of the devices became more common in some medical settings by the late 1980s, but as early as 1989, researchers reported significantly more problems with getting an adequate reading in dark-skinned patients. One 1990 study found that pulse oximetry readings seemed about two and a half times less accurate in black patients, and researchers speculated that they were less reliable because the calibration data were mostly from white patients.
A letter from a team of physicians from the University of Michigan Medical School published in the New England Journal of Medicine in December 2020 sheds light on this known problem and how differences in pulse oximetry measurements posed an increased risk to black patients. Specifically, their research found that black patients had three times more undetected occult hypoxia than white patients in two large groups.
Occult hypoxemia occurs when the arterial oxygen saturation is less than 88 percent despite a pulse oximeter reading of 92 to 96 percent.
In an article published in the Annals of Intensive Care last January, Dr. Martin Tobin and Amal Gibran, the researchers behind the 1990 study, wrote, “In the 31 years since we made this recommendation, we are not aware of any manufacturer attempting to incorporate modified algorithms into pulse oximeters.”
“The inaccuracy of pulse oximetry in black patients is another example of how medical information being generated in (and for) white people contributes to poor clinical outcomes in patients of color,” they added.
gamma study results
The authors of the Johns Hopkins/Baylor study note that overestimating oxygen levels can be associated with cutting back on early treatments or early discharge of a hospital patient, or it can be associated with delaying or withholding treatment that would help shorten the duration of the disease, slow its progression, or reduce the chances of Death.
Of the 7,126 patients with COVID-19 analyzed in the JAMA study, occult hypoxemia occurred in 30.2% of Asian, 28.5% of black patients, and 29.8% of non-Hispanic patients. Among white patients, it was 17.2 percent.
The researchers found that black patients had a 29 percent lower risk, while non-Hispanic black patients had a 23 percent lower risk of identifying treatment eligibility. A total of 451 patients, or 23.7 percent, were not recognized as eligible for treatment; 54.8 percent of these patients were black.
Of the remaining 1,452 patients, or 76.3 percent, who were eventually identified as eligible, black patients had an average delay of 1 hour longer than white patients.
The use of pulse oximeters has also bypassed hospital settings, particularly during the pandemic, and their affordable cost has led to an increase in the number of people using them at home as well. Wu and his colleagues expressed concern about how this also affects how quickly patients seek professional care and how their disease is assessed.
“A black person who tests positive for COVID-19 in an outpatient setting may be advised not to seek care or decide to postpone it based on false reassurance from normal pulse oximetry readings,” the authors wrote in the paper.
“Such a scenario was reflected in the overrepresentation of black patients with unrecognized eligibility for treatment with remdesivir and dexamethasone.” The issue could also be linked to treatment and triage decisions for other respiratory illnesses such as pneumonia and acute respiratory distress syndrome, they added.
Some of the study’s limitations included the fact that the race and ethnicity used in the study were self-reported, so the researchers were unable to measure or explain differences in skin color within each racial or ethnic group. The researchers also said that the results may not be generalizable to healthy individuals or those with less severe disease. They also noted that the use of the oxygen saturation threshold and drug eligibility was an underestimate of the actual delay in treatment, since there are other contributing factors as well, such as drug availability.
“While pulse oximetry has become an essential tool in diagnosis, triage, and management decisions in acute care settings, device deficiency in certain populations has not been adequately investigated or addressed, although recognized for several decades and highlighted in the 2020 Communication on Safety of ahead of the US Food and Drug Administration,” the researchers wrote.
In the short term, Wu says spreading the word to raise awareness of the issue is important and that thresholds for measuring oxygen saturation should become more liberal.
However, long-term he and his colleagues believe the solution requires changing the technology so that it can respond to differences in skin tone. “Technology reform… is certainly possible. It has not been done,” he said.
“It’s a difficult situation that clinicians and clinicians on the front lines face when we have this bias in such a vital tool.”