Addressing SpO₂ sensor accuracy to drive health equity

Monitoring SpO₂ is essential for assessing respiratory function and oxygenation levels in multiple care areas throughout the hospital, aiding in the early detection and management of respiratory distress or compromise.

Pulse oximeters use LEDs to emit light into tissues and a photodetector to measure the amount of transmitted or reflected light. Hemoglobin, the molecule that carries oxygen in red blood cells, will change light absorption characteristics depending on its oxygenation state. Oxygenated hemoglobin (HbO₂) will absorb more light in the infrared range, while deoxygenated hemoglobin (Hb) will absorb more in the red range. The SpO₂ sensor measures the amount of light that reaches the photodetector, and by comparing the amount of the red and infrared light using a calibration curve, the value is displayed as a percentage, representing an estimate of the oxygenated Hb in the blood.

FDA-cleared pulse oximeters have a typical accuracy, reported as Accuracy Root Mean Square (ARMS), of within 2 to 3% of arterial blood gas values. This generally means that during testing, around 66% of SpO₂ values were within 2 or 3% of blood gas values³; and 95% of SpO₂ values were within 0 to 6% of blood gas values. The difference between SpO₂ and SaO₂ is typically small at saturations above 90% and slowly becomes greater when saturations continue to drop below 90%.

Real-world accuracy of pulse oximeters can differ from lab-measured accuracy during desaturation studies. Under certain conditions, pulse oximeters have limitations and may produce inaccurate readings, potentially leading to unrecognized low oxygen saturation levels. Key factors contributing to these inaccuracies include, but are not limited to ambient light interference from external light sources; nail polish, dyes, henna, or tattoo ink; environmental factors, such as extreme temperatures; motion artifacts caused by movement, shivering, or shaking; dyshemoglobinemia or severe anemia disorders; low perfusion due to critical illness, cold temperatures, vasopressors, or shock; and skin pigmentation impacting light transmission.

Various factors can affect the accuracy of the SpO₂ measurement. However, recent academic publications have highlighted the potential for accuracy differences between people with different skin pigmentation.^4-7

We've already seen that pulse oximeters use LEDs emitting red and infrared light to measure SpO₂, using the different absorption characteristics of oxygenated and deoxygenated hemoglobin. However, the pigment responsible for skin color, melanin, also absorbs light in the red and infrared range of the spectrum. Individuals with darker skin pigmentation have higher levels of melanin.

Recent FDA communications have warned that pulse oximeters may overestimate SpO₂ among people with a higher level of skin pigmentation, leading to unrecognized ‘occult’ hypoxemia.⁸ Questions have been raised about pulse oximeter technology, given during its original development subjects with darker skin pigmentation were under-represented in the validation study populations.

FDA guidance requires the inclusion of subjects with darkly pigmented skin in validation studies. This includes at least two darkly-pigmented subjects (or 15% of the subject pool, whichever is larger).

In the future, to avoid subjective responses and ambiguous color definitions, colorimetry, the science and technology used to measure and describe physically the human color perception, will play an important role in laboratory desaturation studies to be used for pulse oximetry accuracy validation purposes.

To this end, the International Organization for Standardization (ISO 80601-2-61) is also considering clinical study designs that collect data points from a minimum of 24 participants spanning the Monk Skin Tone (MST) scale as a way to capture pigmentation diversity. The 10-shade MST scale is known to agree with a person’s self-identified skin tone better than other subjective scales⁹ and has originally been developed to replace the Fitzpatrick scale in fields such as computer vision research. It has recently been proposed as a tool to diversify study participant enrollment in pulse oximeter desaturation studies.

However, there are challenges in the consistency of the MST color scale when printed. Compared to professional printers, office printers can show a large overlap between different MST grades at the light and dark shades of the scale. Without clear printing guidelines, different color prints may thus lead to different MST classification of people with the same skin pigmentation.¹⁰

Another way to improve the assessment of skin pigmentation is by employing more objective measurement techniques. Using the subjective MST scale for screening and the objective Individual Topology Angle (ITA) measurement for skin type stratification at enrollment may be a practicable approach to ensure appropriate representation of all skin types in desaturation studies to evaluate pulse oximetry performance as well as differences that may be due to skin pigmentation.

It is clear that many current studies have their limitations, and skin pigmentation is not solely responsible for inaccuracies in SpO₂ readings.

In the words of the FDA¹¹: “An association of a variable with pulse oximeter accuracy does not always imply causation and may be observed for many reasons.”

This highlights the need to further evaluate and understand all potential causes of bias in pulse oximeter performance before drawing any firm conclusions. For example, low perfusion is associated with misdiagnosis of arterial hypoxemia by pulse oximetry in healthy subjects with dark skin pigmentation in controlled laboratory conditions.¹²

Philips, as a global leader in pulse oximetry monitoring devices, is committed to a continual program of research that enhances the performance of our portfolio. We have already taken the lead in addressing the issue of skin pigmentation and SpO₂ accuracy as part of our company-wide drive for greater health equity and inclusivity. We are actively collaborating with the ISO committee and the FDA on the new SpO₂ standard. We urge manufacturers, regulators, and clinicians to work together to ensure that technology is developed and tested across demographically diverse populations. Additionally, we advocate for adopting best practices to help maintain the accuracy of pulse oximeters in clinical settings.

All our SpO₂ sensors undergo demanding testing to make sure they meet Philips quality and ISO standards. We have also conducted multiple studies to gather data on the impact of skin pigmentation on the accuracy of SpO₂ sensors.

We believe that we can provide a guiding light on this critical issue, helping to aid accurate monitoring, confident diagnosis, and delivering equitable healthcare for all.