Racial/Ethnic Pulse Oximetry Variance Linked to Hidden Hypoxemia, Mortality

Close-up of unrecognizable black woman using a pulse oximeter to check her blood oxygen saturation level and heart rate
A study examines the accuracy of pulse oximetry measures for various racial and ethnic subgroups, and how variability among subgroups affects outcomes.

High variability exists among all racial and ethnic groups with respect to arterial oxygen saturation (SaO2) for fixed pulse oximetry (SpO2) levels, with the greatest discrepancy observed among patients who self-report as Black, Asian, and Hispanic, compared with those who self-report as White. A multicenter, retrospective, cross-sectional study was conducted that included 5 databases. Results of the analysis were published in JAMA Network Open.

The investigators sought to examine ethnic and racial disparities between SaO2 and SpO2 measures and their relationships with clinical outcomes. They used 3 publicly available US electronic health record databases (ie, the Electronic Intensive Care Unit-Clinical Research Database [eICU-CRD], the Medical Information Mart for Intensive Care III [MMIC-III], and the Medical Information Mart for Intensive Care IV [MMIC-IV]), along with the databases of 2 healthcare organizations in the Atlanta, GA, area: Emory Healthcare (from 2014 to 2021; 277 units, including 26 intensive care units [ICUs] across 4 hospitals) and Grady Memorial (from 2014 to 2020; 73 units, including 9 ICUs, in 1 hospital).

The main outcome measures of the study included SaO2 for each SpO2; hidden hypoxemia prevalence; initial demographic features (ie, age, sex); clinical outcomes (ie, length of hospital stay, in-hospital mortality); laboratory values (ie, lactate and creatinine levels); and organ dysfunction by scores (ie, Sequential Organ Failure Assessment [SOFA]) prior to and 24 hours following the arterial blood gas [ABG] measurement).

Among a total of 141,600 hospital encounters with recorded arterial ABG measurements, a total of 87,971 participants with an initial ABG measurement and an SpO2 of at least 88% within 5 minutes prior to the ABG test were included in the analysis. Overall, 37,713 of these participants were women; the mean patient age was 62.2±17.0 years; 1919 patients were Asian, 26,032 were Black, 2397 were Hispanic, and 57,632 were White. Hidden hypoxemia was reported in a total of 4859 participants with varying incidence — Black patients: n=1785; Hispanic patients: n=160; Asian patients: n=92; and White patients: n=2822 (P <.001).

These rates of hidden hypoxemia were associated with greater organ dysfunction at 24 hours after the ABG measurement was obtained, as observed with higher mean SOFA scores and significantly higher in-hospital mortality in Black patients compared with White patients (P <.001). Further, patients with hidden hypoxemia had significantly higher mean serum creatinine levels and serum lactate levels (P <.001 for both) prior to the ABG measurement than did those without hypoxemia.

The investigators concluded that “In this study, discrepancies in pulse oximetry accuracy among racial and ethnic subgroups were associated with higher rates of hidden hypoxemia, mortality, and organ dysfunction,” with findings demonstrating “greater variability in oxygen saturation levels for a given SpO2 level among those who self-identified as Black, followed by those who self-identified as Hispanic, Asian, and White.” Given these results, they added, “Validation of all health technologies, including pulse oximetry, must be carried out across a wider range of patient populations, to prevent the perpetuation of harm from miscalibration.”

Disclosure: Some of the study authors have declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of disclosures.


Wong AI, Charpignon M, Kim H, et al. Analysis of discrepancies between pulse oximetry and arterial oxygen saturation measurements by race and ethnicity and association with organ dysfunction and mortalityJAMA Netw Open. 2021;4(11):e2131674. doi:10.1001/jamanetworkopen.2021.31674