How accurate is oximeter on the Apple Watch? I did a comparison with my medical grade oximeter (Beurer) and the difference is huge: 97% on the watch and 90% on my oximeter!!!
Take a look at the screenshots.
Is it supposed to work like that, or I have a faulty sensor, or the fancy algorithms need some more work?
Apple Watch Series 6, watchOS 7
As far as we know there is no calibration function for the blood oxygen display in the Apple Watch. It’s a mass produced device, clearly marketed as “non-medical” so I would not be surprised to see larger discrepancies than an officially approved device. However, I have a few thoughts you might want to consider.
Both devices are displaying integer (whole number) results. That means you should consider the value with an error margin ±1 simply as a result of converting an analog value into a digital display.
The Watch seems to show some correlation between repeated measurements at the same or very close times. In other words there is a hint that the results might be repeatable but not always consistent over longer periods. This might be confirmed or negated by a longer period of comparison than it was possible to show here.
The trend (estimated by eye) is that the Apple Watch is averaging about 3-4% higher than the approved device. Again this needs a longer review and more results to get a good comparison.
The approved device sensor is probably a closer fit on your skin than the Apple Watch sensor. This might make a difference to individual readings and variation in the watch, but it is probably unavoidable for comfortable wear.
Both sensors use a light beam, so they might be vulnerable to things like skin tone or tattoos. We know the heartbeat monitor in Apple Watch has some limitations from these factors.
Even something as simple as warmth or variable moisture under the watch could make a difference. If you are warm or cold makes a difference to blood flow near the skin, and a film of sweat or water could change the characteristics of lighter transmission particularly as the watch is unlikely to have such consistent skin contact.
I’m not saying any or all of these thoughts are relevant, but they seem to be potential causes of error. All I can suggest is keep monitoring and develop your own idea of how your watch behaves and the correction you need to make. Meanwhile continue to rely on the approved monitor because your medics will be annoyed if you don’t!
As far as we know there is no calibration function for the blood oxygen display in the Apple Watch. It’s a mass produced device, clearly marketed as “non-medical” so I would not be surprised to see larger discrepancies than an officially approved device. However, I have a few thoughts you might want to consider.
Both devices are displaying integer (whole number) results. That means you should consider the value with an error margin ±1 simply as a result of converting an analog value into a digital display.
The Watch seems to show some correlation between repeated measurements at the same or very close times. In other words there is a hint that the results might be repeatable but not always consistent over longer periods. This might be confirmed or negated by a longer period of comparison than it was possible to show here.
The trend (estimated by eye) is that the Apple Watch is averaging about 3-4% higher than the approved device. Again this needs a longer review and more results to get a good comparison.
The approved device sensor is probably a closer fit on your skin than the Apple Watch sensor. This might make a difference to individual readings and variation in the watch, but it is probably unavoidable for comfortable wear.
Both sensors use a light beam, so they might be vulnerable to things like skin tone or tattoos. We know the heartbeat monitor in Apple Watch has some limitations from these factors.
Even something as simple as warmth or variable moisture under the watch could make a difference. If you are warm or cold makes a difference to blood flow near the skin, and a film of sweat or water could change the characteristics of lighter transmission particularly as the watch is unlikely to have such consistent skin contact.
I’m not saying any or all of these thoughts are relevant, but they seem to be potential causes of error. All I can suggest is keep monitoring and develop your own idea of how your watch behaves and the correction you need to make. Meanwhile continue to rely on the approved monitor because your medics will be annoyed if you don’t!
I would doubt the 90% answer from your medical oximeter, that would be a serious medical emergency if it was correct. The healthy normal range is usually accepted to be >95%. Either it needs cleaning or calibration check if that answer is repeated and you are still feeling well.
Thank you for the reply I wasn’t expecting this time of the night!
I have a lung issue and that is why the %SpO2 is low. I am on 24/7 oxygen therapy, usually is above 90% but now I have a cold and is affected. For me the medical emergency is when the oxygen saturation goes below 85%, so let me worry about it.
I am well, a minor cold only, and I did the test again just now, my oximeter showed 90% and the watch showed a 93%.
I will clean it with waterwipes, but if it needs calibration how can I do that?
Thank you in advance
An update...
I have good news the results are looking better, not good but better, after cleaning the sensor on the watch, although there was nothing to clean just wipe it with waterwipes and then dry it.
I would like to find out about calibration, I guess that is internal to Apple since there is nothing on the web.
btw the values from my medical oximeter are an average because it measures continuously the oxygen saturation.
thanks again for the tip
update
I think it needs calibration...
Oximeter on the Apple Watch
