Racial bias and reproducibility in pulse oximetry among medical and surgical inpatients in general care in the Veterans Health Administration 2013-19: multicenter, retrospective cohort study

Abstract

Objectives: To evaluate measurement discrepancies by race between pulse oximetry and arterial oxygen saturation (as measured in arterial blood gas) among inpatients not in intensive care.

Design: Multicenter, retrospective cohort study using electronic medical records from general care medical and surgical inpatients.

Setting: Veteran Health Administration, a national and racially diverse integrated health system in the United States, from 2013 to 2019.

Participants: Adult inpatients in general care (medical and surgical), in Veteran Health Administration medical centers.

Main outcomes measures: Occult hypoxemia (defined as arterial blood oxygen saturation (SaO₂) of <88% despite a pulse oximetry (SpO₂) reading of ≥92%), and whether rates of occult hypoxemia varied by race and ethnic origin.

Results: A total of 30 039 pairs of SpO₂-SaO₂ readings made within 10 minutes of each other were identified during the study. These pairs were predominantly among non-Hispanic white (21 918 (73.0%)) patients; non-Hispanic black patients and Hispanic or Latino patients accounted for 6498 (21.6%) and 1623 (5.4%) pairs in the sample, respectively. Among SpO₂ values greater or equal to 92%, unadjusted probabilities of occult hypoxemia were 15.6% (95% confidence interval 15.0% to 16.1%) in white patients, 19.6% (18.6% to 20.6%) in black patients (P<0.001 v white patients, with similar P values in adjusted models), and 16.2% (14.4% to 18.1%) in Hispanic or Latino patients (P=0.53 v white patients, P<0.05 in adjusted models). This result was consistent in SpO₂-SaO₂ pairs restricted to occur within 5 minutes and 2 minutes. In white patients, an initial SpO₂-SaO₂ pair with little difference in saturation was associated with a 2.7% (95% confidence interval -0.1% to 5.5%) probability of SaO₂ <88% on a later paired SpO₂-SaO₂ reading showing an SpO₂ of 92%, but black patients had a higher probability (12.9% (-3.3% to 29.0%)).

Conclusions: In general care inpatient settings across the Veterans Health Administration where paired readings of arterial blood gas (SaO₂) and pulse oximetry (SpO₂) were obtained, black patients had higher odds than white patients of having occult hypoxemia noted on arterial blood gas but not detected by pulse oximetry. This difference could limit access to supplemental oxygen and other more intensive support and treatments for black patients.

Fig 1

Cohort flow diagram. SaO₂=arterial blood gas reading; SpO₂=pulse oximetry reading; VHA=Veterans Health Administration; ICU=intensive care unit; black=non-Hispanic black; white=non-Hispanic white; Hispanic=Hispanic or Latino

Fig 2

Adjusted rate differences in probability of occult hypoxemia (arterial oxygen saturation SaO₂ <88% when pulse oximetry SpO₂ ≥92%) in study population by pulse oximetry reading and by race, from logistic regression model. Model adjusts for age, male sex, comorbidities, and diagnoses, and run only for pulse oximetry (SpO₂) ≥92% and allowing for non-linear interactions between race and pulse oximetry. Top row of graphs shows estimated predictive margins by race; bottom row of graphs shows the differences between the groups; shaded areas are 95% confidence intervals. Inclusion criteria within columns of graphs is the maximum difference between the time stamp on SaO₂ collection and the recorded SpO₂ time. Moving graphs from left to right, analyses included 5305 SpO₂-SaO₂ pairs with SpO₂ readings of ≥92% measured up to 2 minutes apart (median time difference 1.0 minute (interquartile range 0.2-1.5)); 12 603 pairs measured up to 5 minutes apart (2.6 minutes (1.0-4.0)); and 24 009 pairs measured up to 10 minutes apart (5.0 minutes (2.4-7.7)); these numbers of pairs are lower than all possible SpO₂-SaO₂ pairs because of restricting SpO₂ readings to those 92% and over

Fig 3

Probability of occult hypoxemia (arterial oxygen saturation SaO₂ <88% when pulse oximetry SpO₂ ≥92%) on second paired SpO₂-SaO₂ measurements, by race. Probability is based on the SpO₂-SaO₂ difference from a first pair of readings measured earlier that same day, race of the patient, and pulse oximetry reading at the time of measurement of the second pair of SpO₂-SaO₂ readings. Adjusted probabilities were calculated from a regression stratified by race, and presented using www.iconarray.com visualization recommendations. SpO₂-SaO₂ differences were divided into groups by tertiles from the first reading of the day; these groups ranged from having the lowest SpO₂-SaO₂ difference (SaO₂ 0.1 percentage points lower than or any amount higher than SpO₂) to having the largest SpO₂-SaO₂ difference (SaO₂2 at least 2.5 percentage points lower than SpO₂). Probabilities of occult hypoxemia on the second pair of SpO₂-SaO₂ readings depended on the magnitude of the difference on the first pair, and as a function of the SpO₂ at the second pair and the patient’s race