Where and how are pulse oximeters used, and how has this been impacted since the COVID-19 pandemic?
Finger pulse oximeters are widely used to detect low oxygen levels in the blood (hypoxaemia). Their original role was as “trend” monitors; during anaesthesia and intensive care to monitor for reduction in oxygen (SpO2) for the individual patient.
However, oximeters have become more widespread as “threshold” monitors over the last 15 years; with multiple guidelines quoting SpO2 threshold levels at which to act (including BTS/NICE/SIGN Asthma guidelines, NEWS2 & PEWS scores, SEPSIS and approximately 12 other NICE guidelines).
During the COVID-19 pandemic a plethora of constantly updating guidelines for triage and management of COVID-19 gave pulse oximetry a pivotal role in both primary and secondary care. For example, NHSE guidelines currently state that people with COVID-19 with pulse oximetry levels of <=92% should be considered for immediate emergency transfer to hospital.
What are the general issues with accuracy in these devices, and how can this be addressed?
Pulse oximeters can pass EU regulatory approval with an average accuracy of up to +/- 4 percentage points (i.e. pulse oximeter SpO2 of 94% when the arterial O2 (SaO2) is as low as 90%). The standards allow oximeters to be tested with only 10 healthy young volunteers and a total of 200 data points across all 10 subjects. The FDA calculates only 2/3rds of readings will be within this range, with 1 in 20 being more than twice as inaccurate, i.e. > +/- 8% (i.e. SaO2 <86% despite oximeter SpO2 of 94%).
This means that seemingly reassuring pulse oximeter SpO2 readings can mask dangerous hypoxaemia (SaO2).
The FDA allows accuracy up to +/- 3 percentage points; with 1 in 20 being more than twice as inaccurate, i.e. > +/- 6% (i.e. <88% despite oximeter SpO2 of 94%).
Current pulse oximeter standards (FDA/ ISO) do not appear suitably robust to ensure that pulse oximeters can be relied on as “threshold measures” where +/- 1% can be enough to change management, eg altering the NEWS2 score.
As these devices have been shown to have higher levels of inaccuracy in those with black or brown skin – what is the impact on care on these patients?
This widespread reliance on pulse oximetry during COVID-19 highlighted the racial bias of oximeters as they rely on the light absorption/ reflection through dark skin. The pivotal NEJM letter in 2020 identified that Black patients in the ICU with COVID-19 were 3 times more likely to have missed hypoxaemia than White patients, when using pulse oximeters.
The racial differences in mortality from COVID-19 are well documented, and some of this difference has been attributed to the use of pulse oximeters, guidelines based on SpO2 thresholds, and this racial bias. This may have delayed or even prevented some people from gaining appropriate treatment for severe COVID-19 due to false reassurance based on skin colour.
The UCSF OpenOximetry.org project is systematically and objectively testing pulse oximeters to FDA standards and is a great resource to check for accuracy and racial bias for the oximeters that are being used in clinical care.
To combat the racial inequality in these devices and across respiratory treatment, what steps need to be taken by clinicians to ensure People of Colour (POC) have access to the best possible care?
Education with clear warnings regarding four essential problems:
- General inaccuracy of many finger oximeters (for any skin colour)
- Further inaccuracy of many oximeters for people with brown or black skin
- Possible gender bias for women (poorly publicised/researched)
- Risks of clinicians and patients using non-medical grade oximeters (“lifestyle” devices have CE marks for electrical conformity. A CE mark does not mean something is a regulated medical device).
This awareness will allow clinicians to critically assess what diagnostic “weight” they should put on a SpO2 reading and check the accuracy of the device they are using (OpenOximetry.org).
Current oximeter use (including in non-specialist & primary care) appears to represent major patient safety risks. This should be addressed through UK-wide and global awareness campaigns, regulatory changes, labelling and guideline updates that promote safety; by manufacturers, regulators (MHRA, FDA, etc), clinical authorities including NICE, NHSE, guideline/policy makers and professional bodies.
Conflict of Interest declaration
Dr Nick Gomptz is CEO and founder of EarSwitch Ltd; an organisation currently working to developing more accurate and equitable pulse oximeters.