What is a good AHI with ASV? Revised

palerider wrote:

Uncle Flapp wrote:with a leak rate above the acceptable threshold 28% of the time,

there's *nothing* wrong with the leak rate reported there. nothing at all.

respironics machines report *total* leak, not excess leak. if it was a resmed machine, your comment would be valid.

Oops! You're right. I forgot that was a Sleepy Head preference setting and confused it with the % of night in LL. My apologies.

- Flappy

Uncle Flapp wrote:

palerider wrote:

Uncle Flapp wrote:with a leak rate above the acceptable threshold 28% of the time,

there's *nothing* wrong with the leak rate reported there. nothing at all.

respironics machines report *total* leak, not excess leak. if it was a resmed machine, your comment would be valid.

Oops! You're right. I forgot that was a Sleepy Head preference setting and confused it with the % of night in LL. My apologies.

and mine, if I sounded terse... I just wanted to make it clear both for the OP and others in the future, the differences in how the resmed and respironics reports leaks.

30lpm is great on a respironics, terrible on a resmed as if there isn't enough confusion already

palerider wrote:

Uncle Flapp wrote:...My apologies.

and mine, if I sounded terse...

No, no - not at all. I am far from being an expert on this stuff and appreciate the clarification. In fact, it inspired me to do some research and I learned something new today.

Rarely do I have a problem with large leaks but noticed that I couldn't find any identified in Sleepy Head - so I fired-up Encore Pro. Found one session (December 12th) where the leak rate went over 90l/min for 3 minutes and was flagged as a LL (FWIW it was not flagged in SH). My initial assumption was that the number 90 should be entered into Sleepy Head as the large leak threshold but with some Googling I found this on Philip's website:

Large leak (LL) – Measures pressure versus flow and calculates total leak. LL flagged when approximately 2X expected leak is detected for more than 1.5 minutes.
Auto adjusting response: pressure is dropped by 1 cm every two minutes until LL is cleared (leak level falls below threshold for more than 1.5 minutes).

I confirmed this by looking at other sessions. There were several where pressures got high with leak rates >90l/min. They were not flagged whereas the night of the 12th, pressures were low but leak rate high thus LLs were flagged. So, it seems there is no predefined number to identify large leaks - it fluctuates based on current pressure. I learned something new.

Thanks for your help!

- Flappy

Uncle Flapp wrote:

palerider wrote:

Uncle Flapp wrote:...My apologies.

and mine, if I sounded terse...

No, no - not at all. I am far from being an expert on this stuff and appreciate the clarification. In fact, it inspired me to do some research and I learned something new today.

Rarely do I have a problem with large leaks but noticed that I couldn't find any identified in Sleepy Head - so I fired-up Encore Pro. Found one session (December 12th) where the leak rate went over 90l/min for 3 minutes and was flagged as a LL (FWIW it was not flagged in SH). My initial assumption was that the number 90 should be entered into Sleepy Head as the large leak threshold but with some Googling I found this on Philip's website:

Large leak (LL) – Measures pressure versus flow and calculates total leak. LL flagged when approximately 2X expected leak is detected for more than 1.5 minutes.
Auto adjusting response: pressure is dropped by 1 cm every two minutes until LL is cleared (leak level falls below threshold for more than 1.5 minutes).

I confirmed this by looking at other sessions. There were several where pressures got high with leak rates >90l/min. They were not flagged whereas the night of the 12th, pressures were low but leak rate high thus LLs were flagged. So, it seems there is no predefined number to identify large leaks - it fluctuates based on current pressure. I learned something new.

ohh, nice find, what I'd heard was more vague, "around 70-80lpm" but it seemed to vary, this explains why. thanks! I learned something new too (I've done the most of my research on the resmed side, and just pick up the PRS1 stuff as I go along, mostly)

as to SH not showing the LL flag, are you using the most current version (0.9.8-1)?

palerider wrote:...as to SH not showing the LL flag, are you using the most current version (0.9.8-1)?

Yeppers. 0.9.8-1. Here are some screenshots. 5.7 minutes (1%) large leak on 10/5 in Encore but nothing in SH (shows 0.00%).

it might not be long enough to actually see a mark at the top. try zooming in on the same section in the SH graph and see if theres something.

mark might need to make it more prominent in the flags section when it's a brief one like that.

palerider wrote:it might not be long enough to actually see a mark at the top. try zooming in on the same section in the SH graph and see if theres something.

mark might need to make it more prominent in the flags section when it's a brief one like that.

Nope, not there zoomed-in either. Also note in the left pane (under Cheyne Stokes Respiration) the count is 0.00%. It is not listed under the "events" tab either.

Going back through my history, the longest LL I could find was my first month on therapy - 3/29 of last year. The LL lasted 22 minutes - 17% of the two hour session. Again, SH does not pick up the flag:

On the statistics page, it shows 28.29% of time spent in large leak territory during the last month, but only 3.82% over the last 6 months. Is there something wrong with the Sleepyhead output in that it's showing something alarming that really isn't? Or are the preferences set wrong for the machine/mask/etc?

cathyf wrote:On the statistics page, it shows 28.29% of time spent in large leak territory during the last month, but only 3.82% over the last 6 months. Is there something wrong with the Sleepyhead output in that it's showing something alarming that really isn't? Or are the preferences set wrong for the machine/mask/etc?

huh?

The "% of time above Leak Rate threshold" stat really jumped out at me:



If he's really spending more than a quarter of the time over the "leak threshold" then either the threshold isn't meaningful, or the AHI probably isn't.

EMTUnit-

It might be worth getting a recording pulse oximeter to see if those hypopneas are causing any blood oxygenation problems. If they are, you may need further treatment. If your spO2 looks good, then really comes down to how you are feeling. 20 is high, but AHI is just a number. I would rather have an AHI of 20 with all hypopneas than an AHI of 1 where once an hours I quit breathing altogether for 2 minutes.

My AHI with an ASV machine is typically 0.18 - 1.

cathyf wrote:The "% of time above Leak Rate threshold" stat really jumped out at me:
If he's really spending more than a quarter of the time over the "leak threshold" then either the threshold isn't meaningful, or the AHI probably isn't.

read the next three messages after the one way back there that you're referring to, and report back.

palerider wrote:

EMTunit31 wrote:9.6

ah, well, then, time to upgrade, see below.

I have tried to update several times and it keeps telling me the downloads are not available right now. Any suggestions?

palerider wrote:

Uncle Flapp wrote:with a leak rate above the acceptable threshold 28% of the time,

there's *nothing* wrong with the leak rate reported there. nothing at all.

respironics machines report *total* leak, not excess leak. if it was a resmed machine, your comment would be valid.

That was my original confusion -- Sleepyhead is reporting that 28% of the time the machine was above leak "threshold". If there is nothing wrong with that, then there is something wrong with the logic that decided what the threshold was going to be. Or maybe the settings are set wrong and confusing the machine or Sleepyhead.

Here is the link to last nights report. It's better than the previous night.
https://www.dropbox.com/s/nkf80hwlrtqi9 ... M.jpg?dl=0