Getting discouraged

tisket wrote: ↑

Thu Aug 30, 2018 8:01 pm

Thank you for the analysis and the information.

I was confused by the "4" and "32" numbers because I thought it was saying that was for the entire night, I didn't realize that was just where the line cursor happened to be on the graph. Whoops.

I changed the flag threshold from 24 to 70 as you suggested but it didn't (despite a note about recalculations being complete) change any of the stats. I then tried changing it to 95 and it still didn't change. So I guess it only does this calculation on import.

You'll probalby have to go up to 'data' on the menu, and rebuild your cpap data from the backups that SH makes on import.

Hmm. Loaded data this morning from last night. Flow Rate is missing, as are a number of other graphs. They are not options in the dropdown on the bottom right either.

They are still there for other days. Just missing for last night.

I tried rebuilding under the Data menu but SH crashes each time I try to do this. Seems to be a separate problem.

Was the SD card completely and correctly inserted in the machine last night?
Some graphs and the flow rate is the primary graph...have to go directly onto the SD card during the night to have the data available.
Some stuff is stored in the machine's internal memory and will go onto the card if you put a card into the machine that is even blank but you will be missing some of the details.

How about just create a new profile...and fix the red line threshold first before importing to that new profile?
See what happens. It hurts nothing to have multiple user profiles available. I have 3 in my SleepyHead that I have created for various reasons.

Ah ok, the card was not in the machine last night. I thought it would write all data to the card when I put it back in, didn't realize it only kept some data types in its non-card memory.

I created a new profile and imported and realized I had not changed the 24l/min. I went in to change it and noticed it still had the ventilation numbers I had put in for my other profile (for 4 and 20 cmH2O). Odd. So I purged all the imported data, changed the flag threshold to 70, and imported again.

Nothing looks different. As before, the only large leaks are that one time it went almost to 90. Oddly it is not flagging times when it went over 70, which I would expect it to under either a 24 or a 70 standard.

I figured maybe it is using the expected leak settings for 4 and 20 cmH2O after all. So I unchecked that "calculate unintentional leaks when not present" box, purged the data, and imported it again. Still no change. Leaks at 84 are being flagged as large leak, leaks at 75 are not despite the threshold being 70. Now I am wondering what it takes to make it flag something as a large leak.

I am not sure what we were trying to accomplish with this. Even when the flag threshold was 24 l/min, it was not flagging leaks at 75 l/min as large leaks. Shouldn't it? Maybe I don't understand what the threshold is for.

Thanks

As a test I made sure the "calculate unintentional leaks" was unchecked, changed the leak threshold to 5 l/min, purged all data, and reimported.

No difference. It seems to be totally ignoring both of these settings.

tisket wrote: ↑

Fri Aug 31, 2018 6:23 pm

I am not sure what we were trying to accomplish with this. Even when the flag threshold was 24 l/min, it was not flagging leaks at 75 l/min as large leaks. Shouldn't it? Maybe I don't understand what the threshold is for.

tisket wrote: ↑

Fri Aug 31, 2018 6:23 pm

Leaks at 84 are being flagged as large leak, leaks at 75 are not despite the threshold being 70.

What is "it"....SleepyHead or the machine?

The machine trumps the SH statistics...if you don't see a flag on the Events graph...you didn't have a large leak per the machine.
To heck with what SH might say.

The 70 L/min number...just a ball park number because Respironics doesn't ever use a fixed set number for large leak territory.
It's a moving threshold depending on the pressure used and the mask used.
You might be better off just turning that particular statistic off and just relying on the Events graph for large leak notification.

I told you that the 70 L/min was probably a conservative number and that the real threshold is probably higher.
Quit worrying about the leak statistics....just watch the Events graph...that data comes from the machine itself and not some SH fancy calculations that we can't really get 100% accurate because of the way the threshold will move as the pressure moves.

OK. I am confused I guess about SH. That 24 L/min number is marked "Flag leaks over threshold." I thought that meant it would show up in the Large Leaks event flag in SH. Sounds like you're saying it doesn't, that the machine itself creates those flags. I don't understand what SH does with that 24 L/min number then. Is it used to create the percentage statistics on the left and that's all?

Looking at my breathing waveforms, it looks suspiciously like I am doing mouth exhalation the entire night.

Just was looking at this thread

viewtopic.php?f=1&t=112758&p=1085906#p1085906

particularly this quote

"When a mouth leak occurs, pressurized air does not reach the lungs and does not contribute to ventilation, thereby rendering the treatment less effective or ineffective. "

Which makes me wonder again if the AHI scores I am getting really mean anything if I might be mouth-exhaling the entire night.

Is it safe to rely on AHI reports and leak reports if I might be mouth exhaling all the time? Or does that even matter?

tisket wrote: ↑

Fri Aug 31, 2018 7:08 pm

OK. I am confused I guess about SH. That 24 L/min number is marked "Flag leaks over threshold."

That 'flag leaks over threshold' is for Resmed... you should just turn that off... it's bad programming that it's on..

tisket wrote: ↑

Fri Aug 31, 2018 7:08 pm

Is it safe to rely on AHI reports and leak reports if I might be mouth exhaling all the time? Or does that even matter?

Yes...these machines can and will compensate for leaks no matter where they come from. If it's not a large leak..the machine can deal with it and even if it is a large leak the machine can deal with a lot before it throws up its little hands and says "I quit".
With a Respironics machine...somewhere around 110 L/min total leak is when the machine runs into big trouble. You are no where near that total leak.

Here's a little report where I had a chance to actually see what known mouth breathing looked like. I woke up one morning and I was mouth breathing...a very gentle mouth breathing with no tornado out my mouth or anything like that. Just a gentle breathing.
I wanted to see just what it looked like on the leak graph so I lay there for about 15 minutes wide awake just breathing normally through my mouth with mask and machine on. Finally had to abort the experiment...bladder was about to explode.

Below is the leak graph and I circled the known mouth breathing. Barely a blip on the leak line...and proof that not all mouth breathing equals massive leaks.

Thanks for your patience and for the information!