Evaluating Pulse Oximeter Bias Across a Range of Skin Pigment in Critically Ill Adults - EquiOx

Mar 30, 2025
Font Size
A
A
A
Author/Summarized by Author:
Demilade Adedinsewo, MD, FACC
Summary Reviewer:
Fred M. Kusumoto, MD, FACC
Trial Sponsor:
U.S. Food and Drug Administration
Date Presented:
03/30/2025
Date Updated:
03/30/2025
Original Posted Date:
03/30/2025
References

Contribution To Literature:

The EquiOx study showed that in critically ill patients, oxygen saturation assessed by pulse oximetry (SpO2) was often lower than arterial oxygen saturation (SaO2) and the degree of measurement error varied by skin pigmentation; bias was less negative in darker skin categories.

Description:

The goal of this real-world study was to evaluate the relationship between skin pigmentation and pulse oximeter bias in critically ill patients.

Study Design

Patients who received treatment in the intensive care unit (ICU) at Zuckerberg San Francisco General Hospital between 2022 and 2024, received ≥2 oxygen saturation readings by both a pulse oximeter and blood gas analysis, performed simultaneously. Skin pigmentation was measured using both the subjective Monk Skin Tone Scale and objective measurements of melanin content with a spectrophotometer.

  • Total number of enrollees: 631 patients (1,760 observations)
  • Mean patient age: 62 years
  • Race: 25% White, 21% Black, 20% Hispanic, 20% Asian
  • Skin pigmentation classification: 33% light, 53% medium, 14% dark

Inclusion criteria:

  • Adult ICU patients at level 1 trauma center
  • Directly observed simultaneous SpO2 and SaO2 measurements

Exclusion criteria:

  • No observed arterial blood gas sample
  • Missing skin pigment data

Principal Findings:

The primary outcome was pulse oximeter bias, defined as the difference between SpO2 (pulse oximeter peripheral oxygen saturation) and SaO2 (arterial oxygen saturation), calculated as SpO2 – SaO2. The median pulse oximeter bias observed was negative in all skin pigment categories (measured SpO2 was lower than measured SaO2) with an overall median difference of -1.7% (IQR -2.8%, -0.5%).

  • The Accuracy Root Mean Square error (ARMS) was 3.9%.
  • A less negative bias was observed among individuals with darker skin pigment.
  • The adjusted bias attributable to skin pigment (dark vs. light) based on individual typology angle colorimetry, an objective measure of skin pigmentation, was 0.5% (p<0.04).
  • 20% of the observations had a positive bias. Factors associated with a higher likelihood of a positive bias included Black race, dark skin pigmentation, diabetes, hypertension, peripheral vascular disease, tobacco use, hypoxemia, hypothermia, and a low perfusion index.

Interpretation:

In this study of critically ill patients, oxygen saturation assessed by pulse oximetry (SpO2) was often lower than arterial oxygen saturation (SaO2) and the degree of measurement error varied by skin pigmentation. Notably, within the subgroup of observations where SpO2 values were greater than SaO2 values, a positive association was seen with darker skin pigmentation and individuals of Black race.

References:

Presented by Dr. Carolyn Hendrickson at the American College of Cardiology Annual Scientific Session (ACC.25), Chicago, IL, March 30, 2025.

Clinical Topics: Vascular Medicine

Keywords: ACC25, ACC Annual Scientific Session, Health Care Technology, Heart Rate, Race and Ethnicity, Vascular Diseases


< Back to Listings