- Pulse oximetry systematically overestimated blood oxygenation by one percentage point in Black patients, leading to delays in the treatment of COVID-19, according to an analysis of health record data.
- The study linked the overestimation of blood oxygen to a 4.5 hour delay in receiving supplemental oxygen treatment, and a 3.1 percentage point decline in the probability of hospital admission.
- Stephanie Brown, an author of the paper, called for the Food and Drug Administration to “immediately evaluate its policies around approving pulse oximeters” and for manufacturers to “go back to the drawing board to provide clinicians with a tool that is free from bias.”
Studies have found Asian, Black and Hispanic patients have higher blood oxygen readings than their white counterparts when tested using pulse oximeters and linked the differences to disparities in the provision of care. The findings led the FDA to schedule a meeting of experts to discuss the topic.
The latest study, which was published in the American Journal of Epidemiology, builds on the earlier work by quantifying the impact of blood oxygen overestimation on the treatment of COVID-19. Because blood oxygen levels informed whether a patient was treated for mild or moderate-to-severe COVID-19, the 1% overestimation in Black patients led to delayed care.
Researchers at Sutter Health, a non-profit health care system in northern California, looked at two cohorts: one made up of 43,753 concurrent measurements of blood oxygen using arterial blood gas and pulse oximetry; and a second that featured 8,735 adults who went to the emergency department with COVID-19.
The analysis linked the 1% overestimation of blood oxygen to delayed initiation of supplemental oxygen by 4.5 hours, and dexamethasone treatment by 37 minutes. The researchers also found Black patients had a lower probability of being admitted to hospital or of receiving dexamethasone by 3.1 percentage points, and a 4.2 percentage point decrease in the probability of receiving supplemental oxygen.
Knowing that the pulse oximeters could give biased results also affected patient care in other ways, as Brown, an emergency medicine doctor at Sutter who co-authored the paper, explained on a call with the media to discuss the results.
“That did result in many more Black patients getting the confirmatory test of the gold standard, the arterial blood gas test to really determine is this patient too sick to go home? Do they qualify for more treatment? Brown said.
“And so again, in real time in the emergency department, not being able to trust the devices that we have, the tools that we have to practice medicine, regardless of our intent, and our intent is never to do harm. That was a scary feeling,” Brown added.
Brown wants to see changes in light of the data, calling for the FDA to immediately evaluate its policies and for manufacturers to rethink the design of pulse oximeters. In the meantime, the physician is advising her peers to “maintain a heightened suspicion,” and question whether the devices are giving accurate data.