Addressing racial bias in pulse oximetry

Article Type
Changed
Fri, 08/05/2022 - 15:17

Pulse oximetry is a vital monitoring tool in the ICU and in pulmonary medicine. Regrettably, re-emerging data show that pulse oximeters do not accurately measure blood oxygen levels in Black patients, presumably due to their skin tone. Patients with darker skin are, therefore, more likely to experience occult hypoxemia (i.e., low arterial oxygen saturation despite a seemingly normal pulse oximetry reading). While inaccuracy of pulse oximeter measurements in patients with darker skin has been recognized for decades, recent studies have highlighted this as an ongoing problem with potentially severe consequences for Black patients and other patients of color.

One recent study found that Black patients had almost three times the likelihood of occult hypoxemia compared with White patients (Sjoding, MW, et al. N Engl J Med. 2020;383[25]:2477-8).

Subsequent studies have confirmed this to be a widespread problem across various clinical settings in hundreds of hospitals (Wong AI, et al. JAMA Netw Open. 2021;4[11]:e2131674; Valbuena VS, et al. Chest. 2022;161[4]:971-8). A recent retrospective cohort study of patients with COVID-19 found that occult hypoxemia in Black and Hispanic patients was associated with delayed eligibility for potentially lifesaving COVID-19 therapies (Fawzy AF, et al. JAMA Intern Med. 2022; published online May 31, 2022).

Dr. Emily A. Harlan

Now that numerous studies have demonstrated the inaccuracy of pulse oximetry with the potential to cause harm to historically marginalized racial and ethnic groups, must we abandon the use of pulse oximetry? We would argue that pulse oximeters remain valuable tools, but for now, we must adapt our practice until better devices are widely adopted.

First, it is crucial that health professionals are aware that pulse oximeters may underestimate the true extent of hypoxemia for all patients, but particularly for patients with darker skin. Acknowledging this device flaw is essential to avoid harm to our patients.

Second, clinicians must have heightened skepticism for seemingly normal pulse oximetry values when caring for symptomatic patients at risk of occult hypoxemia.

Until better pulse oximeters are widely available, clinicians must consider workarounds aimed at ensuring timely identification of hypoxemia in Black patients and other patients of color.

These patients may need invasive monitoring of arterial oxygenation, including arterial blood gas checks or an arterial catheter. However, invasive monitoring comes at the cost of discomfort to patients and potential complications, such as vessel or nerve damage.

Dr. Daniel Colon Hidalgo

Invasive monitoring of patients at risk for occult hypoxemia is not an equitable or acceptable long-term solution for this problem. As advocates for patients, clinicians and professional organizations should lobby regulatory bodies to ensure pulse oximeters are accurate for all patients.

We must also call on government leaders to move this process forward. For example, in response to efforts by the United Kingdom’s Intensive Care Society, the Health Secretary of the UK, Sajid Javid, has called for a review of pulse oximeters as part of a larger review assessing structural issues in health care that lead to worse outcomes in racial and ethnic minorities (BBC News. https://www.bbc.com/news/uk-59363544. Published online Nov. 21, 2021).

Device companies are largely for-profit corporations with obligations to their shareholders. It seems that existing incentives are insufficient to motivate investment in less biased technology and real-world evaluations of their devices.

We previously called for buyers of pulse oximeters to change the incentives of device companies – that is, for “hospitals to commit to only purchasing pulse oximeters that have been shown to work equally well in patients of colour.” (Hidalgo DC, et al. Lancet Respir Med. 2021;9[4]:E37). And, indeed, we worry that hospitals are putting themselves at medicolegal risk by not raising their purchasing standards. Since it is now widely known that pulse oximeters are inaccurate in certain patients, could there be liability for hospitals that continue to use devices we know to be disproportionately inaccurate by race?

Dr. Thomas S. Valley


Device companies must commit to fixing racial bias in pulse oximeters. Change is feasible, and pulse oximeters can be redesigned to be accurate and reliable among all patients using existing technology that is decades-old.

In the 1960s and 1970s, Hewlett Packard worked with NASA to noninvasively measure oxygen saturation in astronauts (Moran-Thomas, M. Wired. Published online June 4, 2021. https://www.wired.com/story/pulse-oximeters-equity). The device was designed to work for all skin tones and could be calibrated based on an individual’s skin tone. However, Hewlett Packard moved away from medical devices in the 1980s, shelving their design while other companies took over the oximeter market.

Lastly, as new devices are designed, they must be proven to work for all patients. Testing should be conducted in real-world clinical settings using metrics aligned with clinical care, since we know testing in artificial environments may not generalize to critically ill patients. Testing standards historically used by the FDA, such as only requiring device testing in a small number of non-White individuals, may miss clinically relevant hypoxemia. Non-inferiority studies are particularly susceptible to poor design or under-powering, and rigorous standards are needed from unbiased sources.

While potential solutions are currently being evaluated, the fact remains that the inaccuracy of pulse oximeters has been known for decades without any meaningful action taken to correct the problem.

As Valeria Valbuena, author of a study demonstrating inaccuracy of pulse oximetry in patients about to undergo ECMO, points out, “Using White patients as the standard in biomedical design has led to both differential care and innovation inertia for optimizing the way devices and algorithms work for patients of racial and ethnic minoritized groups” (Valbuena VS. JAMA Intern Med. 2022. doi: 10.1001/jamainternmed.2022.1903).

We know that hypoxemia is dangerous for our patients and that this is only one example of the long-standing systemic racism leading to harm in historically marginalized racial and ethnic groups. It is unacceptable that the devices we rely on to care for our patients are disproportionately inaccurate in non-White patients.

We hope that with increased awareness of this problem, meaningful action will be taken by device companies to ensure pulse oximeters work accurately for all patients.




From the Division of Pulmonary and Critical Care, Department of Medicine and the Center for Bioethics and Social Sciences in Medicine, University of Michigan Medical School (Drs. Harlan and Valley), and the Institute for Healthcare Policy and Innovation (Dr. Valley), University of Michigan, Ann Arbor, MI; and the Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado School of Medicine, Aurora, CO (Dr. Colon Hidalgo).

Publications
Topics
Sections

Pulse oximetry is a vital monitoring tool in the ICU and in pulmonary medicine. Regrettably, re-emerging data show that pulse oximeters do not accurately measure blood oxygen levels in Black patients, presumably due to their skin tone. Patients with darker skin are, therefore, more likely to experience occult hypoxemia (i.e., low arterial oxygen saturation despite a seemingly normal pulse oximetry reading). While inaccuracy of pulse oximeter measurements in patients with darker skin has been recognized for decades, recent studies have highlighted this as an ongoing problem with potentially severe consequences for Black patients and other patients of color.

One recent study found that Black patients had almost three times the likelihood of occult hypoxemia compared with White patients (Sjoding, MW, et al. N Engl J Med. 2020;383[25]:2477-8).

Subsequent studies have confirmed this to be a widespread problem across various clinical settings in hundreds of hospitals (Wong AI, et al. JAMA Netw Open. 2021;4[11]:e2131674; Valbuena VS, et al. Chest. 2022;161[4]:971-8). A recent retrospective cohort study of patients with COVID-19 found that occult hypoxemia in Black and Hispanic patients was associated with delayed eligibility for potentially lifesaving COVID-19 therapies (Fawzy AF, et al. JAMA Intern Med. 2022; published online May 31, 2022).

Dr. Emily A. Harlan

Now that numerous studies have demonstrated the inaccuracy of pulse oximetry with the potential to cause harm to historically marginalized racial and ethnic groups, must we abandon the use of pulse oximetry? We would argue that pulse oximeters remain valuable tools, but for now, we must adapt our practice until better devices are widely adopted.

First, it is crucial that health professionals are aware that pulse oximeters may underestimate the true extent of hypoxemia for all patients, but particularly for patients with darker skin. Acknowledging this device flaw is essential to avoid harm to our patients.

Second, clinicians must have heightened skepticism for seemingly normal pulse oximetry values when caring for symptomatic patients at risk of occult hypoxemia.

Until better pulse oximeters are widely available, clinicians must consider workarounds aimed at ensuring timely identification of hypoxemia in Black patients and other patients of color.

These patients may need invasive monitoring of arterial oxygenation, including arterial blood gas checks or an arterial catheter. However, invasive monitoring comes at the cost of discomfort to patients and potential complications, such as vessel or nerve damage.

Dr. Daniel Colon Hidalgo

Invasive monitoring of patients at risk for occult hypoxemia is not an equitable or acceptable long-term solution for this problem. As advocates for patients, clinicians and professional organizations should lobby regulatory bodies to ensure pulse oximeters are accurate for all patients.

We must also call on government leaders to move this process forward. For example, in response to efforts by the United Kingdom’s Intensive Care Society, the Health Secretary of the UK, Sajid Javid, has called for a review of pulse oximeters as part of a larger review assessing structural issues in health care that lead to worse outcomes in racial and ethnic minorities (BBC News. https://www.bbc.com/news/uk-59363544. Published online Nov. 21, 2021).

Device companies are largely for-profit corporations with obligations to their shareholders. It seems that existing incentives are insufficient to motivate investment in less biased technology and real-world evaluations of their devices.

We previously called for buyers of pulse oximeters to change the incentives of device companies – that is, for “hospitals to commit to only purchasing pulse oximeters that have been shown to work equally well in patients of colour.” (Hidalgo DC, et al. Lancet Respir Med. 2021;9[4]:E37). And, indeed, we worry that hospitals are putting themselves at medicolegal risk by not raising their purchasing standards. Since it is now widely known that pulse oximeters are inaccurate in certain patients, could there be liability for hospitals that continue to use devices we know to be disproportionately inaccurate by race?

Dr. Thomas S. Valley


Device companies must commit to fixing racial bias in pulse oximeters. Change is feasible, and pulse oximeters can be redesigned to be accurate and reliable among all patients using existing technology that is decades-old.

In the 1960s and 1970s, Hewlett Packard worked with NASA to noninvasively measure oxygen saturation in astronauts (Moran-Thomas, M. Wired. Published online June 4, 2021. https://www.wired.com/story/pulse-oximeters-equity). The device was designed to work for all skin tones and could be calibrated based on an individual’s skin tone. However, Hewlett Packard moved away from medical devices in the 1980s, shelving their design while other companies took over the oximeter market.

Lastly, as new devices are designed, they must be proven to work for all patients. Testing should be conducted in real-world clinical settings using metrics aligned with clinical care, since we know testing in artificial environments may not generalize to critically ill patients. Testing standards historically used by the FDA, such as only requiring device testing in a small number of non-White individuals, may miss clinically relevant hypoxemia. Non-inferiority studies are particularly susceptible to poor design or under-powering, and rigorous standards are needed from unbiased sources.

While potential solutions are currently being evaluated, the fact remains that the inaccuracy of pulse oximeters has been known for decades without any meaningful action taken to correct the problem.

As Valeria Valbuena, author of a study demonstrating inaccuracy of pulse oximetry in patients about to undergo ECMO, points out, “Using White patients as the standard in biomedical design has led to both differential care and innovation inertia for optimizing the way devices and algorithms work for patients of racial and ethnic minoritized groups” (Valbuena VS. JAMA Intern Med. 2022. doi: 10.1001/jamainternmed.2022.1903).

We know that hypoxemia is dangerous for our patients and that this is only one example of the long-standing systemic racism leading to harm in historically marginalized racial and ethnic groups. It is unacceptable that the devices we rely on to care for our patients are disproportionately inaccurate in non-White patients.

We hope that with increased awareness of this problem, meaningful action will be taken by device companies to ensure pulse oximeters work accurately for all patients.




From the Division of Pulmonary and Critical Care, Department of Medicine and the Center for Bioethics and Social Sciences in Medicine, University of Michigan Medical School (Drs. Harlan and Valley), and the Institute for Healthcare Policy and Innovation (Dr. Valley), University of Michigan, Ann Arbor, MI; and the Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado School of Medicine, Aurora, CO (Dr. Colon Hidalgo).

Pulse oximetry is a vital monitoring tool in the ICU and in pulmonary medicine. Regrettably, re-emerging data show that pulse oximeters do not accurately measure blood oxygen levels in Black patients, presumably due to their skin tone. Patients with darker skin are, therefore, more likely to experience occult hypoxemia (i.e., low arterial oxygen saturation despite a seemingly normal pulse oximetry reading). While inaccuracy of pulse oximeter measurements in patients with darker skin has been recognized for decades, recent studies have highlighted this as an ongoing problem with potentially severe consequences for Black patients and other patients of color.

One recent study found that Black patients had almost three times the likelihood of occult hypoxemia compared with White patients (Sjoding, MW, et al. N Engl J Med. 2020;383[25]:2477-8).

Subsequent studies have confirmed this to be a widespread problem across various clinical settings in hundreds of hospitals (Wong AI, et al. JAMA Netw Open. 2021;4[11]:e2131674; Valbuena VS, et al. Chest. 2022;161[4]:971-8). A recent retrospective cohort study of patients with COVID-19 found that occult hypoxemia in Black and Hispanic patients was associated with delayed eligibility for potentially lifesaving COVID-19 therapies (Fawzy AF, et al. JAMA Intern Med. 2022; published online May 31, 2022).

Dr. Emily A. Harlan

Now that numerous studies have demonstrated the inaccuracy of pulse oximetry with the potential to cause harm to historically marginalized racial and ethnic groups, must we abandon the use of pulse oximetry? We would argue that pulse oximeters remain valuable tools, but for now, we must adapt our practice until better devices are widely adopted.

First, it is crucial that health professionals are aware that pulse oximeters may underestimate the true extent of hypoxemia for all patients, but particularly for patients with darker skin. Acknowledging this device flaw is essential to avoid harm to our patients.

Second, clinicians must have heightened skepticism for seemingly normal pulse oximetry values when caring for symptomatic patients at risk of occult hypoxemia.

Until better pulse oximeters are widely available, clinicians must consider workarounds aimed at ensuring timely identification of hypoxemia in Black patients and other patients of color.

These patients may need invasive monitoring of arterial oxygenation, including arterial blood gas checks or an arterial catheter. However, invasive monitoring comes at the cost of discomfort to patients and potential complications, such as vessel or nerve damage.

Dr. Daniel Colon Hidalgo

Invasive monitoring of patients at risk for occult hypoxemia is not an equitable or acceptable long-term solution for this problem. As advocates for patients, clinicians and professional organizations should lobby regulatory bodies to ensure pulse oximeters are accurate for all patients.

We must also call on government leaders to move this process forward. For example, in response to efforts by the United Kingdom’s Intensive Care Society, the Health Secretary of the UK, Sajid Javid, has called for a review of pulse oximeters as part of a larger review assessing structural issues in health care that lead to worse outcomes in racial and ethnic minorities (BBC News. https://www.bbc.com/news/uk-59363544. Published online Nov. 21, 2021).

Device companies are largely for-profit corporations with obligations to their shareholders. It seems that existing incentives are insufficient to motivate investment in less biased technology and real-world evaluations of their devices.

We previously called for buyers of pulse oximeters to change the incentives of device companies – that is, for “hospitals to commit to only purchasing pulse oximeters that have been shown to work equally well in patients of colour.” (Hidalgo DC, et al. Lancet Respir Med. 2021;9[4]:E37). And, indeed, we worry that hospitals are putting themselves at medicolegal risk by not raising their purchasing standards. Since it is now widely known that pulse oximeters are inaccurate in certain patients, could there be liability for hospitals that continue to use devices we know to be disproportionately inaccurate by race?

Dr. Thomas S. Valley


Device companies must commit to fixing racial bias in pulse oximeters. Change is feasible, and pulse oximeters can be redesigned to be accurate and reliable among all patients using existing technology that is decades-old.

In the 1960s and 1970s, Hewlett Packard worked with NASA to noninvasively measure oxygen saturation in astronauts (Moran-Thomas, M. Wired. Published online June 4, 2021. https://www.wired.com/story/pulse-oximeters-equity). The device was designed to work for all skin tones and could be calibrated based on an individual’s skin tone. However, Hewlett Packard moved away from medical devices in the 1980s, shelving their design while other companies took over the oximeter market.

Lastly, as new devices are designed, they must be proven to work for all patients. Testing should be conducted in real-world clinical settings using metrics aligned with clinical care, since we know testing in artificial environments may not generalize to critically ill patients. Testing standards historically used by the FDA, such as only requiring device testing in a small number of non-White individuals, may miss clinically relevant hypoxemia. Non-inferiority studies are particularly susceptible to poor design or under-powering, and rigorous standards are needed from unbiased sources.

While potential solutions are currently being evaluated, the fact remains that the inaccuracy of pulse oximeters has been known for decades without any meaningful action taken to correct the problem.

As Valeria Valbuena, author of a study demonstrating inaccuracy of pulse oximetry in patients about to undergo ECMO, points out, “Using White patients as the standard in biomedical design has led to both differential care and innovation inertia for optimizing the way devices and algorithms work for patients of racial and ethnic minoritized groups” (Valbuena VS. JAMA Intern Med. 2022. doi: 10.1001/jamainternmed.2022.1903).

We know that hypoxemia is dangerous for our patients and that this is only one example of the long-standing systemic racism leading to harm in historically marginalized racial and ethnic groups. It is unacceptable that the devices we rely on to care for our patients are disproportionately inaccurate in non-White patients.

We hope that with increased awareness of this problem, meaningful action will be taken by device companies to ensure pulse oximeters work accurately for all patients.




From the Division of Pulmonary and Critical Care, Department of Medicine and the Center for Bioethics and Social Sciences in Medicine, University of Michigan Medical School (Drs. Harlan and Valley), and the Institute for Healthcare Policy and Innovation (Dr. Valley), University of Michigan, Ann Arbor, MI; and the Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado School of Medicine, Aurora, CO (Dr. Colon Hidalgo).

Publications
Publications
Topics
Article Type
Sections
Disallow All Ads
Content Gating
No Gating (article Unlocked/Free)
Alternative CME
Disqus Comments
Default
Use ProPublica
Hide sidebar & use full width
render the right sidebar.
Conference Recap Checkbox
Not Conference Recap
Clinical Edge
Display the Slideshow in this Article
Medscape Article
Display survey writer
Reuters content
Disable Inline Native ads
WebMD Article