Pulse Oximeter Performance and Skin Pigment: Comparison of 34 Oximeters Using Current and Emerging Regulatory...

Area of Science:

• Medical Devices
• Biomedical Engineering
• Health Equity

Background:

• Regulatory bodies like the International Organization for Standardization (ISO) and US Food and Drug Administration (FDA) are revising pulse oximeter standards to address performance disparities related to skin pigmentation.
• Previous assessments indicated potential biases in pulse oximeter readings based on skin tone.

Purpose of the Study:

• To evaluate the performance of common pulse oximeters against current and proposed regulatory standards, specifically examining bias related to skin pigmentation.
• To test the hypothesis that not all pulse oximeters exhibit increased positive bias in individuals with darker skin pigmentation compared to those with lighter skin.
• To determine the proportion of pulse oximeters that would meet anticipated FDA regulatory criteria.

Main Methods:

• A controlled desaturation protocol was employed to test 34 pulse oximeters in healthy adults across a range of arterial oxygen saturations (SaO2) from 70% to 100%.
• Participant skin pigmentation was characterized using the Monk Skin Tone (MST) scale and the individual typology angle (ITA).
• Key performance metrics included accuracy root mean square error (ARMS), bias, and skin pigment differential bias, with Monte Carlo simulations used to assess the impact of participant selection on device compliance.

Main Results:

• Under current ISO and FDA guidelines, a significant majority of devices met accuracy standards. However, only 3% of devices passed the anticipated FDA criteria.
• Eleven oximeters demonstrated increased positive bias in participants with darker skin pigmentation compared to lighter skin.
• Device performance assessment varied significantly based on cohort selection, with 18 devices potentially passing or failing depending on the participants included in the analysis.

Conclusions:

• Pulse oximeter performance is inconsistent across different manufacturers and models.
• While not universal, some devices exhibit greater positive bias in individuals with darker skin.
• Anticipated ISO and FDA regulatory updates present differing assessment outcomes, highlighting the need for refined cohort size and differential bias criteria to ensure improved real-world performance and mitigate health disparities.