Racial bias in pulse oximetry present in patients with acute hypoxemic respiratory failure

NASHVILLE, Tenn. — In a new study, racial bias in the use of pulse oximetry was consistently present even with high levels of oxygen supplementation in patients with acute hypoxemic respiratory failure, researchers reported.

Julia Baranda Balmaceda

For a study presented at the CHEST Annual Meeting, Julia Baranda Balmaceda, BS, an MD candidate (2024) at University of Kansas School of Medicine, and colleagues aimed to investigate any discrepancies based on race in pulse oximetry devices used in the ICU and their institution.

“We reproduced findings of a race-based discrepancy of pulse oximetry to patients’ true oxygen saturation based on arterial blood gas. Moreover, we see this difference persists in a critically ill population on supplemental oxygen,” Balmaceda told Healio.

Balmaceda and colleagues conducted a retrospective analysis comparing direct arterial blood gas and pulse oximeter readings of 112 self-identified white and 32 self-identified Black patients with acute hypoxemic respiratory failure. The researchers evaluated data on arterial oxygen saturation (SaO₂) and the associated pulse oximetry level (SpO₂).

According to data presented at the meeting, there was no difference in average SaO₂ between Black and white patients, at 92.6% and 92.1%, respectively. However, the researchers noted SpO₂ was on average significantly higher among Black patients compared with white patients (P = .041). The researchers concluded that race was a significant predictor of SpO₂ in regression analyses of SaO₂. The SpO₂ pulse oximeter readings overestimated oxygen saturation by 0.814% in Black patients compared with white patients, according to the results.

Balmaceda and colleagues also evaluated the clinical management of these patients. They found that the maximum flow rate was significantly lower among Black patients. But, in regression analyses of flow rate including SaO₂, SpO₂, partial pressure of oxygen (PaO₂) and race, the only significant predictor was SpO₂ (P = .024), according to the results. When the researchers analyzed for fraction of inspired oxygen (FiO₂) including SaO₂, SpO₂, PaO₂ and race, they found that SaO₂ (P = .0003) and PaO₂ (P = .023) were significant predictors.

“In comparison to white patients, Black patients had pulse oximetry values with a statistically significant discrepancy from their arterial blood gas oxygen saturation. However, race was not a statistically significant predictor in the treatment management based on FiO₂ and max flow rate of oxygen,” Balmaceda told Healio.

The researchers are finishing a larger dataset that will potentially provide more information on what this race-based discrepancy could mean from a practical standpoint.

“I’d like to see the companies that produce these devices take into account skin color when calibrating pulse oximeters,” Balmaceda said.

“Any tool we use that affects patient care warrants being investigated. The question of ‘clinically relevant difference’ should not stop the medical world from elucidating any bias in the health care system in regard to race, sex, socioeconomic status, etc.”

Reference:

Balmaceda J, et al. Chest. 2022;doi:10.1016/j.chest.2022.08.930.