Accuracy of fingertip pulse oximeters: a device accuracy study

AimTo assess the accuracy of commercially available pulse oximeters at various price points, and to investigate how long is needed to obtain a reliable measurement. MethodsAll 75 participants had pulse oximetry with 6 test oximeters and a reference (gold standard) oximeter. Where blood gases were clinically indicated, these results were also recorded. ResultsCompared to the reference oximeter, average errors in oxygen saturation ranged from 0.6% to 2.6% overestimation. All devices appeared to require some time to "settle" to an accurate reading, but accuracy at 30 seconds from application was similar to that at 2 minutes. There was no clear association between device cost and accuracy. ConclusionsFingertip pulse oximeters tended to overestimate oxygen saturation by up to 2.5% on average. We recommend that pulse oximeters are not read immediately after application, but are allowed at least 30 seconds to settle to an accurate reading. KEY MESSAGESO_LIFingertip pulse oximeters are widely used in clinical care, particularly in the primary care and pre-hospital environment, and increasingly for self-monitoring by patients. C_LIO_LIWe found that pulse oximeter accuracy varies for both oxygen saturation and heart rate, with 2.5% overestimation of oxygen saturation on average. C_LIO_LIPrice is not necessarily an indicator of accuracy in pulse oximeters. C_LIO_LIAccuracy can be improved by waiting at least 30 seconds before recording oxygen saturation. C_LI PLAIN LANGUAGE SUMMARYPulse oximeters are non-invasive fingertip medical devices that measure the amount of oxygen in the blood ("oxygen saturations" or "sats"), which in healthy people should be between 96% and 100%. They also measure heart rate. Some pulse oximeters are available to buy by the public quite cheaply, and their use for home monitoring has become more common during and since the COVID-19 pandemic. We wanted to understand if cheap oximeters are still accurate. We bought 6 different pulse oximeters and tested them on 26 healthy people and 49 patients who were attending a hospital outpatient clinic. We compared the oxygen saturation measurements from the 6 cheap oximeters with measurements from a hospital quality pulse oximeter as a "gold standard". We compared the heart rate measurements from the pulse oximeters to a measurement of heart rate made by hand by a nurse. We also measured how long it took for the pulse oximeter to give an accurate measurement. We found that the cheap pulse oximeters generally gave an oxygen saturation value that was between 0.6% and 2.6% higher than the hospital pulse oximeter. However, sometimes there was a larger difference of up to 7.4%. The cheap pulse oximeters also generally gave a heart rate that was between 0.8 and 2.7 beats per minute higher than the rate measured by a nurse. After 30 seconds, the pulse oximeter measurements were about as accurate as they were after 2 minutes. We found that very cheap pulse oximeters were just as accurate as less cheap pulse oximeters. Our results suggest that cheap pulse oximeters tend to overestimate oxygen saturations and heart rate, but usually by a small amount which would not make much difference to the decisions a doctor would make. However, sometimes there may be a bigger difference so we would recommend repeating with a different device if there was concern about the patient.