Pulse oximeters may overestimate blood oxygen saturation for people with high levels of skin pigmentation in hospital settings compared with gold standard measures
During the COVID-19 pandemic, there have been concerns regarding potential bias in pulse oximetry measurements for people with high levels of skin pigmentation. In November 2021 the UK Health Secretary ordered a review into racial bias in medical devices including pulse oximeters.
Researchers from the National Institute for Health and Care Research (NIHR) Applied Research Collaboration Greater Manchester (ARC-GM) and The University of Manchester have undertaken a review to investigate the accuracy of pulse oximetry in measuring blood oxygen saturations by levels of skin pigmentation. The review found that, compared with the gold standard measure for blood oxygen saturation, hospital-based pulse oximetry may overestimate oxygen saturation by around 1% (on average) in people with high levels of skin pigmentation and people whose ethnicity is reported as Black/African American.
This review included all of the available evidence up to December 2021.
Dr Chunhu Shi, Research Fellow at The University of Manchester and NIHR ARC-GM Evaluation Theme, led this review:
Our estimate of the average bias of 1% in people with high levels of skin pigmentation is new knowledge. This estimate, for the first time, reflects how large the extent of the bias would be in pulse oximetry measurements for people with high levels of skin pigmentation.
The findings from this work have substantial implications. The overestimation identified at threshold values for diagnosis of hypoxaemia could lead to clinically important hypoxaemia remaining undetected and untreated.
For example, when using 92% as the threshold of diagnosing hypoxaemia, clinicians could consider a pulse oximetry reading of 93% as normal for a patient with high levels of skin pigmentation if the overestimate is disregarded. However, when the overestimate of 1% for this level of skin pigmentation is considered, the patient’s true oxygen saturation could be around 92% and suggest a possible hypoxaemia.
Despite the clinical implications of the overestimation, the bias estimates met internationally recommended thresholds and UK standards. We considered that the currently recommended thresholds may need re-evaluation, and the use of more conservative criterion may have merit.
Our findings support calls for the use of better calibrating algorithms within oximeter device software to address possible measurement bias. This review results offer some insights into the possible amount of bias to consider
.
The full findings of this work have been published in the BMC Medicine:
This review has been funded by the NIHR ARC-GM and ARC North West Coast and supported by NIHR, and the Accelerated Access Collaborative at NHS England and NHS Improvement.