SpO2 vs CPAP data in Sleepyhead

[Albercook]

There seems to be little correlation between my PulseOximeter readings and my Resmed output. 

Here are some examples. 

There may be a correlation in for this event. The vertical red line shows what the CPAP calls an RE event and the Oximeter calls a drop in SpO2. But the next two images show nothing in the CPAP data but clear SpO2 drops appear in the Oximeter data.

I know that there are lots of leaks. I'm working on that. They don't obscure the data in the vast majority of cases.

[url=][/url]

[url=][/url]

[url=][/url]

Finally, you can see in the top panel (event flags) that very few of the CPAP events CA, OA, UA, H, RE correspond to the Oximeter data SD.

[url=][/url]

The Oximeter data looks real. The shape is reasonable. After all, it is a direct measure of my blood oxygen level. It seems that maybe it would pick up more events than the CPAP because the software is using indirect measurements. 

But it seems like the CPAP is identifying events when there is no evidence from the Pulse Oximeter.

Any thoughts?

Thanks in advance for any input that you can provide.

Regards,
George

[Sleeprider]

I think your sync might be off by about 20 to 30 seconds with the SpO2 events showing a little later than the CPAP events. All of your views are zoomed in pretty far, showing only about 6-8 minutes of data. It might be interesting to look at a little longer time segment with more events. Around 05:15 looks like a good candidate.

I think you need to keep in perspective why SpO2 measure might be useful. It's not so much to correlate to individual events, although that is an interesting thing to do. I think Oximetry needs to be looked at in a broader sense to see if CPAP therapy alone is effective, or if the number and extent of desaturation events are such that supplement oxygen should be considered.

In your case, you have a moderately high number of events, and optimization of your therapy should be the first priority. At 7-20 cm pressure and no EPR, it seems your pressure settings might be better at 12-16. The leaks might be affecting the CA event rate. Have you tried EPR in the past and found it increases CA?

[Albercook]

Thanks the pressure is now set to 12-20. Doc changes it last night. 

I thought that I followed the wiki guidelines for the image posting. What did I get wrong so that there are two images?

Here is the section around 5:15.

[url=][/url]

I have read that the bodies carbon dioxide chemosensors are far more sensitive than its' oxygen sensors. 

That would seem to imply that sleep could be disrupted by increased CO2 before O2 has a chance to drop. 

I assumed that the change in O2 and CO2 were linked in time and magnitude. If that is the case, that would explain how the CPAP detects changes in breathing before O2 drops. And the response may even prevent a measurable drop in O2.

But if they are that tightly linked how can O2 drop without the CPAP detecting my bodies' response?

I guess there could be some sort of hysteresis that could explain it.

Still confused.

[Albercook]

I don't see a consistent delay between the SpO2 data and the CPAP data. There are a few places where you can see that but it does not do a very good job of explaining the differences.

[Albercook]

I think you need to keep in perspective why SpO2 measure might be useful. It's not so much to correlate to individual events, although that is an interesting thing to do.  I think Oximetry needs to be looked at in a broader sense to see if CPAP therapy alone is effective, or if the number and extent of desaturation events are such that supplement oxygen should be considered.

I don't understand this? SpO2 is the only direct measure of a physiological parameter.

[Sleeprider]

It's been a long time since we engaged in this conversation, but your SpO2 data does not seem to show a problem, and it's very difficult to correlate apnea or hypopnea events with desaturation. If such a correlation exists, I can't identify it.

[sheepless]

"...it's very difficult to correlate apnea or hypopnea events with desaturation. If such a correlation exists, I can't identify it."


Sleeprider, are you saying this 'in general' or with respect to Albercook's charts only? (the former is counter-intuitive and would be news to me.)

[Sleeprider]

"...it's very difficult to correlate apnea or hypopnea events with desaturation. If such a correlation exists, I can't identify it."


Sleeprider, are you saying this 'in general' or with respect to Albercook's charts only?  (the former is counter-intuitive and would be news to me.)

Specifically with regard to the graphs above.  This post was actually made in April, so I'm a bit out of the loop on this case.  Looking at the chart, there is a large cluster of events at 5:15. The SpO2 data shows a minor dip around 5:45, so maybe that is the tie offset. What I'm saying is I can't identify a relationship between events and SpO2. Also, my original statement that was quoted, means to say that there are not any serious denaturation issues here.  Pulse and SpO2 are all in a narrow and acceptable range.

[Albercook]

I have new data and will post it ASAP.

Since April I found the "cause" of my leak. I was opening my mouth. A chin strap solved that.

I'm still not sure how well synchronized the data are. Even though both machines are within a couple seconds of each other there is still a problem. They don't appear to be accounting for time zone the same way.

I'm glad that you don't think that they are uncorrelated in general. That makes more sense to me.

I'm not ready to say that they are uncorrelated in my case until I have faith in the time synchronization.

When I use the PulseOx import wizard I get two choices
I started my PulseOX at (or near) the same time as my CPAP
I want to use the time reported by my Oxmeter's build in clock
Or I can manually adjust the time

I have been manually adjusting the time because of the time zone error

Without a clear sync it is difficult to figure out any relationship. There are so many events according to my PulseOx it is hard to figure out which if any correlate with CPAP events. By definition, most of the PulseOx events must not correlate since there are so many.

Even if we ignore the correlation question.

1. What does it mean to have so many PulseOx events
2. What does it mean to have so many more PulseOx events than CPAP events?

Correlation almost does not matter. Either the PulseOx is giving false positives or the CPAP is giving false negatives.

Why believe the CPAP over the PulseOx.

Considering that I still want a nap most days I'm inclined to think that the CPAP is missing events.

It is not uncommon for me to have over 10 SPo2Ox events/hour with an average duration of 30 sec. and 30 events per hour. The few PulseOx charts that I have seen from other people don't look like that.