Confirm CPAP reports with secondary sleep tracking

[Gus Wah]

Since I began CPAP therapy I've learned a lot about what my AirSense nightly reports tell me.  And in my opinion the data is questionable.

First, I've learned to totally ignore ResMed's mobile app that provides a pat on the back with generous scores (often 100 or close).  For me, these eye-candy numbers are fraught with inaccuracy.  And in recent months the software is so messed up that the app won't even connect unless I reboot both the AirSense and the mobile phone every single day.  No loss, really.

Then there is OSCAR, which is enormously better but involves a procedure -- transferring data from the SD card to computer and a time consuming daily ritual of examining all the great stuff it offers.  If you have time for this, fine.  But OSCAR still relies on data from the "observations" of my AirSense 10.  And that's what I find dubious.

What I really need to know on a daily basis is simple.  I want an accurate reporting of my AHI, when I had periods of REM, my overnight oxygenation, and for how long I actually slept, as opposed to laying in bed and thinking about the day ahead.  ResMed has not been able to do this.  If I turn in my sleep and flatten one of the air tubes, a breach in the airflow increases pressures elsewhere and leaks find their way out.  Prolonged leakage while I'm sleeping seems to give the AirSense misinformation.  On such nights my AHI can jump from 0.5 to as much as 7.x.  Have a couple of long stretches, typically these involve breath-holding, and I've just told the AirSense that I've had apneas or hypopneas.  If I awaken at 5am and my index is higher than I'd like, no problem.  I just lay there until 8am and presto -- ResMed averages it out to a lower number that will make me feel good.

For me, at least, this is not accurate sleep monitoring, not by any means.

For this reason I have been looking at wearable sleep trackers, the kind that provide all the statistics on REM, oxygen levels, AHI, heart monitoring, and whatever else to authenticate the results of my efforts to improve my sleep.  I'm curious if others on this board have tried a secondary device -- I'm quite sure that they have -- and if someone will comment with information that might be helpful.

[PurchanceToDream]

In terms of tracking your breaths, I think resmed data is pretty darn accurate. Leaks flush out more CO2 which can change your breathing response and might account for some of the AHI difference.

However, you are right about the inane pat-on-the-back app and the overall AHI. The former is so condescendingly useless it makes me want to gag. The latter cannot distinguish when you are actually asleep.

Something like a fitbit can give you a better sense of when are asleep, which you can correlate with the OSCAR data to get a better sense of AHI. You can get an oximeter easily enough and port that data into OSCAR as well. And you can get a device like Dreem 2 to much more accurately track actual sleep and sleep stages than a fitbit (or similar).

I'm pretty sure there are no simple all-in-one sleep & breathing trackers than can just give you an accurate AHI each morning.

I think the single most important measure of whether treatment is working is: how do you feel in the mornings?

[Gus Wah]

I watched the Viennese fellow's videos about the Dreem 2 a few weeks ago. The deal breaker was that it appears to be a headgear type of device, and for me that's an immediate conflict for CPAP users since all masks also require headgear.

But this could be a great marketing opportunity for CPAP mask producers who should be building the technology directly into their headgear. The possibilities would be endless.

[kappa]

Sleep stage tracking via wrist wearables is not really that reliable - with Rob ter Horst's comparisons ('The Quantified Scientist' - Dutchman living in Vienna - population sample size of 1!) the comparisons between various sleep tracking devices is all over the place. Actually, sleep stage tracking by any means seems to be more art than science and even PSG based staging involves subjective decisions by the person involved (and automated systems are trained based on that manually created staging data).

This paper - http://dx.doi.org/10.3390/s21051562 - provides an interesting summary of various papers comparing different sleep tracking devices. Most of the movement based trackers have a kappa value of 0.5 or less, which puts them in the "weak" agreement range.

A pulse oximiter such as a SleepU or O2Ring or one of the Contec devices is probably the best device for an alternate view of sleep data. These will provide a reasonable frequency view of any Oxygen desaturation (due to apnea) or heart rate acceleration (due to arousal).