Skin Pigmentation Bias in Pulse Oximetry Measurements

Pulse oximeters have revolutionized modern medicine, enabling continuous, accurate, and non-invasive monitoring of functional oxygen saturation of hemoglobin (SpO₂) (Chan et al. 2013). Taken as a fifth vital sign, the device’s use is routine in emergency care, general ward, intermediate/intensive care unit, the operating room, and in general practices (Bosch et al. 2024). However, with its prevalence in patient care and recent availability over the counter, the accuracy of the device has received critical attention, driven by the need to identify occult hypoxemia; a condition where a patient’s blood oxygen level is dangerously low, but a pulse oximeter reading incorrectly indicates it is within the normal range of greater than or equal to 92%(Cabanas et al. 2023). A critical factor affecting device accuracy is skin pigmentation, with large-scale studies and clinical reports documenting inaccuracies and overestimations in patients with higher skin melanin concentrations (Cabanas et al. 2023). These findings expose a deeper issue in healthcare: medical technologies and clinical practices designed and calibrated on lighter-skinned patients can perpetuate racial disparities, putting marginalized patients at higher risk of misdiagnoses, delayed treatment, and poorer healthcare outcomes.

A standard pulse oximeter determines oxygen saturation (SpO₂) by shining light at two or more wavelengths from a light-emitting diose through the tissue (typically the finger, and analyzing the resulting optical signals via a photodetector (Figure 1) (Cabanas et al. 2023; Chan et al. 2013).

When the device illuminates a tissue site, the photodetector measures how much light is absorbed, generating two or more photoplethysmography signals. These signals fluctuate with the cardiac cycle: during systole, increased blood volume reduces the detected light, and during diastole, decreased blood volume increases detected light (Chan et al. 2013).

Wavelengths function based on the principle that oxyhemoglobin absorbs a greater amount of near-infrared (IR) light and lower amounts of red light than deoxyhemoglobin(Chan et al. 2013). This is why oxygenated blood appears bright red to the eye, as it transmits more red light, whereas deoxyhemoglobin absorbs more red light, appearing less red. Exploiting this difference in absorption patterns allows pulse oximeters to emit two wavelengths of light, red at 660nm and near-IR at 940nm. The transmitted light is then detected by a photodiode. By comparing the ratios of absorbed light at these two wavelengths, the pulse oximeter calculates the relative proportion of oxygenated hemoglobin in the blood, providing a continuous, real-time measurement of SpO₂.

However, because of the high light absorption capacity of melanin, the transmitted light wavelength is reduced, making the oximeter’s light unable to penetrate the subcutaneous tissue of patients with darker skin tones (Cabanas et al. 2023). This leads to systematic overestimation of SpO₂ in patients with higher melanin levels, as confirmed by clinical reports linking skin tone to diagnostic errors. For example, Sjoding et al. (2020) analyzed over 10,000 paired pulse oximetry and arterial blood gas measurements and found that Black patients were nearly three times more likely than white patients to have undetected hypoxemia (12% vs. 4%) despite “normal” SpO₂ readings. Similarly, Bickler et al. (2005) and Feiner et al. (2007) demonstrated that at true SpO₂ values below 90%, pulse oximeters overestimated oxygen saturation by up to 3–4% in individuals with darker pigmentation. These discrepancies can lead to missed diagnoses of hypoxemia in critical settings such as COVID-19 management, anesthesia, and neonatal care, where treatment thresholds are based on pulse oximetry readings (Fawzy et al. 2022). Emerging engineering solutions include advanced calibration with diverse skin tones, device redesign utilizing multi-wavelength sensing, and/or new light techniques like radially polarized light to reduce melanin’s interference with readings.

Overall, this disparity, which puts marginalized patients at risk of occult hypoxemia and delayed care, demands immediate, multi-faceted intervention to recalibrate devices and enforce equitable patient monitoring standards across all skin tones.

References

Bickler, Philip E., John R. Feiner, and John W. Severinghaus. 2005. “Effects of Skin Pigmentation on Pulse Oximeter Accuracy at Low Saturation.” Anesthesiology 102 (4): 715–19. https://doi.org/10.1097/00000542-200504000-00004

Bosch, Nicholas A., Anica C. Law, Ashraf Fawzy, and Theodore J. Iwashyna. 2024. “Prevalence of Inpatient Pulse Oximetry in Operative and Nonoperative Settings: A Multicenter Study.” CHEST Critical Care 2 (4): 100104. https://doi.org/10.1016/j.chstcc.2024.100104

Cabanas, Ana M., Pilar Martín-Escudero, and Kirk H. Shelley. 2023. “Improving Pulse Oximetry Accuracy in Dark-Skinned Patients: Technical Aspects and Current Regulations.” British Journal of Anaesthesia 131 (4): 640–44. https://doi.org/10.1016/j.bja.2023.07.005

Chan, Edward D., Michael M. Chan, and Mallory M. Chan. 2013. “Pulse Oximetry: Understanding Its Basic Principles Facilitates Appreciation of Its Limitations.” Respiratory Medicine 107 (6): 789–99. https://doi.org/10.1016/j.rmed.2013.02.004

Chandraguptha, Dilshan. 2021. “Pulse Oximetry.” FOS Media Students’ Blog, September 4. https://fos.cmb.ac.lk/blog/pulse-oximetry/.

Fawzy, Ashraf, Tianshi David Wu, Kunbo Wang, et al. 2022. “Racial and Ethnic Discrepancy in Pulse Oximetry and Delayed Identification of Treatment Eligibility Among Patients With COVID-19.” JAMA Internal Medicine 182 (7): 730. https://doi.org/10.1001/jamainternmed.2022.1906

Feiner, John R., John W. Severinghaus, and Philip E. Bickler. 2007. “Dark Skin Decreases the Accuracy of Pulse Oximeters at Low Oxygen Saturation: The Effects of Oximeter Probe Type and Gender.” Anesthesia & Analgesia 105 (6): S18–23. https://doi.org/10.1213/01.ane.0000285988.35174.d9

Sjoding, Michael W., Robert P. Dickson, Theodore J. Iwashyna, Steven E. Gay, and Thomas S. Valley. 2020. “Racial Bias in Pulse Oximetry Measurement.” New England Journal of Medicine 383 (25): 2477–78. https://doi.org/10.1056/NEJMc2029240