Pulse Oximeters are Important for Keeping all Patients Safe
On December 17, 2020, Sjoding et al. published a retrospective analysis of pulse oximetry (SpO2) data from two patient cohorts indicating that in some patients, occult hypoxemia was not detected when compared to paired oxyhemoglobin saturation measured by laboratory co-oximetry (SaO2).1 Occult hypoxemia was defined as an SaO2 of < 90% when the paired SpO2 measurements were 92% or greater. The authors compared sub-groups from the cohorts self-identifying as Black and White, and found that the incidence of occult hypoxemia was three times greater in Black patients (11.7%) compared with White patients (3.6%). As the authors noted, these findings, if correct, have important patient safety implications since patient triage based upon pulse oximeter measurements could fail to lead to appropriate escalation of care. As a retrospective, uncontrolled study without objective measurements of skin tone, the analysis performed by Sjoding et al. has important limitations. Nevertheless, it is important to verify these findings to understand if there is the potential for pulse oximeter measurements to mislead clinicians, especially in patients with dark skin tones.
What is the Evidence?
The impact of skin tone on pulse oximeter measurements has been documented in the scientific literature since at least 2005. The putative source of bias in measurement is overlapping absorption of light in the red region (660 nm) for both oxyhemoglobin and the skin pigment melanin. Laboratory studies into the impact of skin tone on pulse oximeter measurements have documented a bias, although not of the magnitude identified in the Sjoding data. Bickler et al. found that SpO2 measurements overestimated SaO2 measurements to a greater degree in patients with dark skin tones. The bias increased as saturation decreased and varied with the type of oximeter. They found a maximum bias of 3.56 ± 2.45% for test subjects with dark skin in the 60–70% saturation range but no more than 0.93 ± 1.64% for saturations above 80%.2 The same group studied additional pulse oximeters in test subjects with dark and light skin tones and concluded that several factors were predictive of errors in pulse oximeter measurements including skin tone, probe type, saturation level, and sex. They also stated that bias would be important for patients with a saturation less than 80%.3 Of note, Jubran and Tobin prospectively studied a cohort of ICU patients to determine if SpO2 measurements could be used to titrate oxygen to maintain a PaO2 > 60. Those authors identified a greater bias in SpO2 measurements in patients with dark skin tones and recommended that a threshold of 95% be used for oxygen titration versus 92% for White patients.4 No known studies to date investigate the impact of sex and skin tone together, which could potentially lead to greater measurement bias in female patients with dark skin tones.
The measurement bias demonstrated in the Bickler et al. and Jubran and Tobin publications was apparently not well known by medical professionals, as gauged by a lack of description of this phenomenon in major textbooks of medicine, surgery, and emergency medicine. The phenomenon is described in textbooks of anesthesiology, though the degree to which this is considered in current clinical practice is unclear. The Sjoding et al. publication, if replicated, is concerning because measurement bias was demonstrated at SpO2 levels thought to be consistent with normoxemia. Since the Sjoding publication, there has been significant work by pulse oximeter manufacturers, the United States Food and Drug Administration (FDA) and independent testing laboratories to further investigate the potential for bias due to skin tone (Personal communications). The results of this work will be forthcoming, but are not yet ready for publication.
Regulatory Response to Date
The FDA began to investigate the Sjoding et al. findings shortly after they were published and that work is ongoing. On January 25, 2021, United States Senators Warren, Wyden, and Booker requested that the FDA “conduct a review of the accuracy of pulse oximeters across racially diverse patients and consumers.”5 On February 19, 2021, the FDA issued a safety communication entitled: “Pulse Oximeter Accuracy and Limitations.” That communication emphasizes the known accuracy limitations of pulse oximeters including patients with dark skin tones stating that “if an FDA-cleared pulse oximeter reads 90%, then the true oxygen saturation in the blood is generally between 86 and 94%.⁶ It is important to note that FDA clearance of a pulse oximeter requires that 15% of test subjects with dark pigmentation, or two subjects (whichever is greater) be included in the participant pool.⁷ The FDA safety communication addresses the Sjoding publication, identifying the limitations of that retrospective analysis and recognizing the “need to further evaluate and understand the association between skin pigmentation and oximeter accuracy.”
The preponderance of evidence supports the conclusion that there is a measurement bias in pulse oximeter measurements due to skin tone such that pulse oximeter measurements may overestimate the actual oxyhemoglobin saturation in patients with dark skin tones. Laboratory data obtained under controlled conditions does not indicate that the magnitude of the bias is significant enough to influence clinical decision making until the saturation is less than 80%. Clinical performance is likely to be different from that obtained in the laboratory, and it is clear that many factors will influence the accuracy of pulse oximetry in addition to skin tone. Therefore, clinicians should not make patient care decisions such as hospital or intensive care unit discharge on the basis of a single SpO2 value.
Despite the known limitations of the pulse oximeter, APSF believes that patients are safer with continued use of pulse oximetry to estimate arterial oxygenation. It is potentially more harmful if the known bias in measurement related to skin tone resulted in a lack of confidence in pulse oximetry as a monitoring tool for patients with dark skin tones.
The findings by Sjoding et al. require verification but present at least two opportunities to improve clinical care and outcomes. First, there is an opportunity for manufacturers, regulators, and clinicians to work together to ensure that technology is developed and tested to document clinical performance in demographically and clinically diverse populations. The FDA’s requirement for inclusion of “darkly pigmented subjects” in device development warrants reconsideration. Requirements for objective measurement of skin tone should be specified. More importantly, including 15% darkly pigmented subjects in the study group may reduce the average measurement bias in that population, but not necessarily result in ideal performance for the individual patient. Closer scrutiny to minimizing measurement bias in subjects with dark skin tones is warranted, including reconsideration of the 15% threshold. Second, this is an opportunity to examine more closely how pulse oximetry is used in the clinical setting and to heighten awareness of the factors that can lead to inaccurate measurements. Like any monitoring device, the measurements obtained by a pulse oximeter are estimates of the actual physiologic condition and can be erroneous. Factors other than skin tone known to affect the accuracy of pulse oximetry include perfusion, dyshemoglobinemias, anemia, brand of oximeter, and motion. Sound clinical decision making depends upon a complete assessment of the patient, not a reliance on a single monitored parameter.
APSF supports the renewed attention to the accuracy of the pulse oximeter, which has rightly revolutionized medical care and augmented patient safety. We call on clinicians, manufacturers, and regulators to work together to ensure that this device offers equitable benefits to all the patients we serve.
- Sjoding MW, Dickson RP, Iwashyna TJ, et al. Racial bias in pulse oximetry measurement. N Engl J Med. 2020;383: 2477–2478.
- Bickler PE, Feiner JR, Severinghaus JW. Effects of skin pigmentation on pulse oximeter accuracy at low saturation. Anesthesiology. 2005;102:715–719.
- Feiner, JR, Severinghaus JW, Bickler PE. Dark skin decreases the accuracy of pulse oximeters at low oxygen saturation: the effects of oximeter probe type and gender. Anesthesia Analgesia. 2007;105:S18–S23.
- Jubran A, Tobin MJ. Reliability of pulse oximetry in titrating supplemental oxygen therapy in ventilator-dependent patients. Chest. 1990;97:1420–1425.
- 2020.01.25 Letter to FDA re bias in pulse oximetery measurements. https://www.warren.senate.gov/imo/media/doc/2020.01.25%20Letter%20to%20FDA%20re%20Bias%20in%20Pulse%20Oximetry%20Measurements.pdf. Accessed February 24, 2021.
- Pulse Oximeter Accuracy and Limitations: FDA Safety Communication. https://www.fda.gov/medical-devices/safety-communications/pulse-oximeter-accuracy-and
email&utm_source=govdelivery. Accessed February 24, 2021.
- Pulse Oximeters – Premarket Notification Submissions [510(k)s]: Guidance for Industry and Food and Drug Administration Staff. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pulse-oximeters-premarket-notification-submissions-510ks-guidance-industry-and-food-and-drug. Accessed February 24, 2021.