Real vs. erroneous SPO2 readings

[spottymaldoon]

I've been doing some thinking about oximetry and apnea/hypopnea. I notice that many pulse oximeters tend to throw out extremely low readings every now and then - readings in the 80% range and even lower. If these are to be believed then the person would have to be in critical condition. I think many of these sporadic low readings aren't necessarily real and may be either due to particular involuntary body movements or just software glitches.

If you have a low SPO2 episode, it seems reasonable that this would be accompanied by a change in heart-rate - an increase in fact. I therefore scanned through my records for those nights when I wasn't using the CPAP machine and I found several episodes where there was a mild drop of SPO2 accompanied by an increase in heart-rate. Here is a graph showing several of these:

https://drive.google.com/file/d/0B1ZlKa1...sp=sharing

I take these to be real minor apneas.

However, on the same night I got some much lower SPO2 readings with no change in heart rate - like this example with an SPO2 of 74%:

https://drive.google.com/file/d/0B1ZlKa1...sp=sharing

Note also that the SPO2 readings 30 seconds before and 30 seconds after that low one were 97% and 96% - I can't believe the body would recover so fast.

So far, whenever I have taken SPO2 readings with the CPAP machine I have seen no SPO2/heart-rate changes taking place together. Even when the machine itself has registered an apnea (or hypopnea) I haven't yet seen any simultaneous SPO2 drop.

I know a lot of people here love to work with data so I thought I'd submit my findings and my theory and ask you whether this makes sense to you?

[bwexler]

What are you using to check your SPO2?
There are various ways to attach the sensors to your finger. Some are more prone than others to be temporarily dislodged by movement.
Does your unit provide a profusion index. That will provide an indication of the quality/accuracy of the O2 reading.

[eseedhouse]

I've been doing some thinking about oximetry and apnea/hypopnea. I notice that many pulse oximeters tend to throw out extremely low readings every now and then - readings in the 80% range and even lower. If these are to be believed then the person would have to be in critical condition.

Depends. Climbers on Mt. Everest who don't use supplemental oxygen have been recorded with O2 levels below 50% and were able to make the summit and return. Of course these are superbly conditioned people.

[spottymaldoon]

"What are you using to check your SPO2?"

Bwexler, I'm using an old BCI Fingerprint .. but the point I wanted to emphasize is that if the pulse remains serene and stable then a sudden SPO2 drop probably shouldn't indicate an apnea.

"Climbers on Mt. Everest who don't use supplemental oxygen have been recorded with O2 levels below 50% and were able to make the summit and return"

Thanks Ed.
That's interesting but don't forget that the overall pressure at the summit of Mt Everest is less than one third of what it is at sea level - the oxygen atmospheric partial pressure correspondingly reduced. Under those conditions blood SPO2 as normally measured might be unreliable. What I hoped to hear was of somebody who is in good health but who has apnea being able to give a reliable value for a reduced SPO2 during a confirmed episode - and what the heart rate did during that time.

[archangle]

Looks like you have only one reading per minute at that scale. That's not often enough to be of any use in telling whether the readings are real or bogus. Can your software do better than that? I'd want at least 10 readings per minute to be ready to call the readings false.

Yeah, it's probably a bogus reading, but your data isn't good enough to really tell.

As for pulse rate, I think it usually goes up, but not always. An increase in BPM would help confirm the SpO2 reading, but it doesn't prove it wrong.

SpO2 can drop quite a bit during a single apnea. There's also about a 20 second delay between the onset of apnea and SpO2 dropping in your finger.

[spottymaldoon]

I've been doing some thinking about oximetry and apnea/hypopnea. I notice that many pulse oximeters tend to throw out extremely low readings every now and then - readings in the 80% range and even lower. If these are to be believed then the person would have to be in critical condition.

Depends. Climbers on Mt. Everest who don't use supplemental oxygen have been recorded with O2 levels below 50% and were able to make the summit and return. Of course these are superbly conditioned people.

Looks like you have only one reading per minute at that scale. That's not often enough to be of any use in telling whether the readings are real or bogus. Can your software do better than that? I'd want at least 10 readings per minute to be ready to call the readings false.

Yeah, it's probably a bogus reading, but your data isn't good enough to really tell.

As for pulse rate, I think it usually goes up, but not always. An increase in BPM would help confirm the SpO2 reading, but it doesn't prove it wrong.

SpO2 can drop quite a bit during a single apnea. There's also about a 20 second delay between the onset of apnea and SpO2 dropping in your finger.

Every 30 seconds actually - yes I wish my software/machine recorded more frequent data. Thanks for the insight however - in your opinion, can I simulate the SPO2 profile for an apnea by holding my breath? So, in other words, if I hold it for the 20 seconds you mention, should I start to see the SPO2 going down? I'll try in any case.

I just had the chance to try somebody's new machine and the Sleepyhead software - it is light years ahead of my combination.

The person who did Sleepyhead was just brilliant.

[archangle]

Every 30 seconds actually - yes I wish my software/machine recorded more frequent data. Thanks for the insight however - in your opinion, can I simulate the SPO2 profile for an apnea by holding my breath? So, in other words, if I hold it for the 20 seconds you mention, should I start to see the SPO2 going down? I'll try in any case.

It looks like every minute in those graphs.

I found it really hard to show an SpO2 drop by holding my breath. The respiratory drive is just too high, I usually had to breathe before my readings would drop. One trick is to exhale and then hold your breath. Otherwise, the O2 in your lungs continues to oxygenate your blood while you're holding your breath.

I was looking at some PSG graphs and the delay between apneas and SpO2 drops is really confusing. It's a complex system with several delays in the system.

Trying to hold your breath and drop SpO2 makes you really wonder how you can have long apneas and not wake up immediately.

[spottymaldoon]

Every 30 seconds actually - yes I wish my software/machine recorded more frequent data. Thanks for the insight however - in your opinion, can I simulate the SPO2 profile for an apnea by holding my breath? So, in other words, if I hold it for the 20 seconds you mention, should I start to see the SPO2 going down? I'll try in any case.

It looks like every minute in those graphs.

I found it really hard to show an SpO2 drop by holding my breath. The respiratory drive is just too high, I usually had to breathe before my readings would drop. One trick is to exhale and then hold your breath. Otherwise, the O2 in your lungs continues to oxygenate your blood while you're holding your breath.

I was looking at some PSG graphs and the delay between apneas and SpO2 drops is really confusing. It's a complex system with several delays in the system.

Trying to hold your breath and drop SpO2 makes you really wonder how you can have long apneas and not wake up immediately.

I just had a go - I exhaled and held my breath for 30 seconds: the SPO2 stayed at 98 and actually dropped to 97 only once I had resumed breathing. Then I inhaled and held my breath for one minute (which was actually easier than the 30 secs when exhaled). Again there was no change in the SPO2 which remained at 98.

The refresh rate on this old BCI Fingerprint machine is pretty quick in real time - it record mode it's every 30 seconds.

And I agree, how could anybody withhold breathing during sleep long enough to produce some of the low SPO2s we see.

[archangle]

And I agree, how could anybody withhold breathing during sleep long enough to produce some of the low SPO2s we see.

Just goes to emphasize the difference between sleep and wake. Your body works different, your autonomic nervous system is different, your respiratory system is different, and your brain is different.

I played with this a bit today. Sit down, relax, and then experiment with a few deep breaths, then exhale all the way and hold. Still darn hard to see a drop in SpO2.