I've had this happen, and I decided that it was caused by mouth breathing while wearing a nasal mask, and that whatever was happening represented a complete meltdown of the AirSense 10 software. I'm wondering if your takeaway should be that you can't successfully use a nasal mask without mouth breathing if you are intoxicated.
My takeaway is that I can't successfully use a nasal mask without mouth breathing even if I'm cold sober!
The key is that the machine shuts itself off, and then when it turns back on it's still at the old high (actually pegged-to-maximum) pressure that it was at when it turned off. What I think I see happening is that the machine is underestimating the leak, I pretty much entirely stop inhaling through my nose, the machine uses inspiration to calculate what's going on, even though it's measuring and logging exhales, it logs it as not breathing. You really need to look at the whole parallel set of the graphs at once to figure out what it's thinking -- so those periods where the exhales are very clearly there and the inhales aren't, the machine stops EPR and logs the respiration rate as zero. Even though the expirations on the flow rate curve are there.
The key piece of data is that when the machine returns after the break, does the pressure reset to your minimums, or does it start at the same pressure that it left off at? If it powers on at minimum, that's a proper shutoff. If it guiltily sneaks back on at the same pressure that it was running when it turned off, that's because it's realized that it shouldn't have shut off. (A note about how a ResMed machine comes back on after a power flicker -- it comes back at the pressure that it was running when the power dropped. You can test this by unplugging the power cord and plugging it back in.)
We had another brand-new cpap user (midwestguy) whose machine was doing this. In all of his data, the machine didn't record itself detecting ANY leaks. Even when he was awake and deliberately pulling the mask away from his face he would look at the data later and no leaks would be shown. It became obvious from looking at his data that he was mouth breathing, the machine's leak sensor(s) was not operational, and the machine couldn't figure out what to do. His DME replaced the machine.
Here's my pictures from a particularly ugly night a year ago.
First the whole night:
![[Image: overview.png?raw=1]](https://www.dropbox.com/s/oecsvapil1ziwh9/overview.png?raw=1)
A series of shutoffs over 72 minutes:
![[Image: Sequence1.png?raw=1]](https://www.dropbox.com/s/7c7x18d30juzwzv/Sequence1.png?raw=1)
![[Image: Sequence2.png?raw=1]](https://www.dropbox.com/s/a6wyaewm2eklpqy/Sequence2.png?raw=1)
![[Image: Sequence3.png?raw=1]](https://www.dropbox.com/s/241kmjqjrujxjun/Sequence3.png?raw=1)
Notice that the last break is real -- that's me turning the machine off, which is why it reset to minimum pressure.
Finally, here's a closer view of the period leading up to a shutoff:
![[Image: Closeup.png?raw=1]](https://www.dropbox.com/s/9u76di8tpdjwds9/Closeup.png?raw=1)
Notice how the respiration rate is logged as zero when I'm clearly measurably breathing. And the leak rate is below redline for all but that brief period.
What's absolutely terrifying is that the machine is logging data which the algorithms obviously have no freaking clue how to interpret and how to react in response. And while the machine is faithfully
recording this, it's not
reporting any of it. This represents a catastrophic failure of the machine to deliver therapy. We all know from our experience that the medical "professionals" who are supposed to be monitoring our therapy only look at the summary reports that the manufacturers' software spit out. If the manufacturer's software reports that the manufacturer's product is "successfully controlling your apnea" then the "professionals" are completely uninterested in the fact that the machine recorded terrible failure.
My interpretation of this data is that I have to use a full-face mask.
I had a sleep clinic appointment two weeks after this trainwreck of a night. I brought my card, and a "professional" analyzed the data, for which the clinic billed $144, and my insurance allowed a charge of $82.6
6. This is what "expert analysis" looks like:
Code:
Today's CPAP download shows an apnea-hypopnea index of 1.3 events/hr, minimal mask leak, and excellent compliance, wearing CPAP 100% of the time for an average of 6.5 hours per night.
CPAP Download:
CPAP Pressure (cm):............................. 7-8
Machine Type:.......................... AirSense S10
Days of use?..................................... 95
Percent (%) of days used?....................... 100
Hours of use on days used?........................ 6
Minutes of use on days used?..................... 34
Average daily hours of usage, all days?........... 6
Average daily minutes of usage, all days?........ 34
Three months later they wrote me a prescription for a new machine, which included a prescription for nasal pillows. (Fortunately Apria ignored that and sent me the full-face mask I asked for.)
![[Image: vautoShutoff.png?raw=1]](https://www.dropbox.com/s/nlk53ibzrl93x1s/vautoShutoff.png?raw=1)
![[Image: vautoWideCloseup.png?raw=1]](https://www.dropbox.com/s/m35xbkjfocxcz8n/vautoWideCloseup.png?raw=1)
