How do I read leak data in Sleepyhead?

[amyela]

Hi! So, I'm brand new to the forums, so I apologize if this has been addressed before...

I'm new-ish to CPAP therapy and have no idea how to understand leak data in sleepyhead. Last night I tried a new mask for the first time. I really like the mask, but I checked my data in sleepyhead this morning and my leak rate seemed high, to my completely untrained/uninformed eye. Can you help me understand how to interpret the leak graph? Shouldn't there ideally be no leaks?? (But if my AHI is low then does it even matter anyway?)

It didn't fit into the screenshot, but also here are the leak statistics in number form:

Leak Rate: min 0.00 - med 3.00 - 95% 8.00 - Max 13.00
Total Leaks: min 21.00 - med 33.00 - 95% 38.00 - max 38.00

Thanks in advance for your help!!

[palerider]

Shouldn't there ideally be no leaks?? (But if my AHI is low then does it even matter anyway?)

It didn't fit into the screenshot, but also here are the leak statistics in number form:

Leak Rate: min 0.00 - med 3.00 - 95% 8.00 - Max 13.00
Total Leaks: min 21.00 - med 33.00 - 95% 38.00 - max 38.00

your mask has a built in leak (vent) rate, which varies based on pressure.
yes, it matters, because too high a leak can give false readings on ahi.
'leak rate' is sleepyheads *GUESS* as to what your excess leak is.
total leak is how much, in total, excess and built in leak there is.
and, it'd fit on the screen shot if you'd turn off the useless calendar by hitting the ^ before the date.

[Pugsy]

If you go to the top thread in the Announcements section at the top of the main forum page you will see where I have written some stuff about understanding what SleepyHead is showing you.

I did cover leak explanation for the Respironics machines. Don't confuse it with the ResMed explanation because they report leaks a bit differently and what I say about ResMed leaks doesn't pertain to Respironic leak numbers to avoid.

On this one report you posted...look on the events graph on the top right. See the LL on the left side? That's for Large Leak and if you hit Large Leak territory it will put a flag in that line. I don't see any on this report and based on what the top leak line in the leak graph shows...you never got close. Large Leak territory will vary with the pressure used (so if the pressure varies the large leak territory borderline will vary) and the mask used so we can't give you an exact large leak line in the sand. We can get close though and it probably is up around 70 L/min.

Based this report anyway....you aren't experiencing leaks that reach large leak territory. Your machine can compensate for any excess leaks (above the expected vent/intentional leak rate) up to around 80 to 100 L/min.
Stay below the 70 to 80 and if you don't see any large leak flags and whatever leak you see will be well within the machine's ability to deal with.

Now any leak that wakes you up, even if it is a tiny leak, needs to be dealt with from the aspect of it messing with sleep quality. So it might be small and not reach large leak territory and impact the therapy itself....if it messes with your sleep it is unwanted.

Are you waking often during the night?

[Phil_in_CA]

Thank you, amyela, for posting and for the replies.

I have been looking at Sleepyhead with my PR CPAP data again, recently, and wondered the same thing. Maybe the the leaks corresponding to sleep events (RERAs, etc.) are just coincidental... but if I see LLs, I know I need to look at my mask cushion or something m,ask related, to find out what is leaking...

Thanks, Pugsy, for sharing that the smaller peaks (that don't flag LLs) are within the machine's ability to cover with increased pressure. I know where to look now! Thanks!

[Pugsy]

Thanks, Pugsy, for sharing that the smaller peaks (that don't flag LLs) are within the machine's ability to cover with increased pressure. I know where to look now! Thanks!

You misunderstood.
These machines don't/won't try to compensate for leaks with more pressure...in fact they often reduce the pressure in an effort to get leaks under better control.
What I meant was with increased pressure the vent rate (which is the intentional leak rate) that all masks have will increase as the pressure increases because the vent rate increases with more pressure.
The vent rate at 10 cm pressure is going to be less than the vent rate at 15 cm pressure even if there are zero excess leaks.

This is why it is important to understand how Respironics reports leaks...the total leak is the sum of the vent rate plus any excess leak. So it's possible to have a wild looking leak line and not have any excess leaks at all if the pressure itself is varying quite a bit.
Won't be so much with this OP since the pressures are primarily fixed after ramp has ended but could be quite prominent in someone who uses auto adjusting pressures.

Also SleepyHead will often make mountains out of mole hills when a person isn't having much of a higher vent rate due to the scale of the graph. It can look like there is a lot more change in the leak than is really going on of any importance. So it can look ugly when in reality it is quite decent.

[palerider]

Thanks, Pugsy, for sharing that the smaller peaks( leaks) (that don't flag LLs) are within the machine's ability to cover with increased pressure.

yours is a common misconception.

Pressure is the result of a resistance to flow.

Your cpap doesn't create "pressure" it creates "flow" (which is why you'll sometimes see them referred to as 'flow generators').

for any given amount of flow, if you decrease the amount of resistance (say with a mask leak) the pressure will drop, the machine will then increase the flow in order to maintain your pressure, if the leak is bad enough, the machine won't be able to increase the flow any more, and pressure will drop. for instance, if you take the mask off, the flow generator will ramp up to full speed, with a LOT of flow, but there will be negligible pressure.

as Pugsy mentions, the machines are programmed to reduce flow in order to reduce the pressure in the face of large leaks in order to try and get the mask to seal, before going back to normal operation.

[GEG]

Palerider, that was a great explanation for me. I wasn't worried about it at this point, but it was also something I didn't understand completely. Thanks for taking the time to explain it in detail like that!

[palerider]

Palerider, that was a great explanation for me. I wasn't worried about it at this point, but it was also something I didn't understand completely. Thanks for taking the time to explain it in detail like that!

I figured that Pugsy's answer was enough for this particular situation, but I wanted to expand a bit on the whole 'how it all works' bit.

another way to think about it is the garden hose analogy, if you turn garden hose on full, there's lots of water coming out, but it doesn't go very far.. (lots of flow, low pressure) now, if you put your thumb over the end, you're resisting the flow, which causes the pressure to go up, and that higher pressure causes the water to shoot further, because your resisting the flow has increased the pressure at the end of the hose.

[Uncle_Bob]

Thanks, Pugsy, for sharing that the smaller peaks( leaks) (that don't flag LLs) are within the machine's ability to cover with increased pressure.

yours is a common misconception.

Pressure is the result of a resistance to flow.

Your cpap doesn't create "pressure" it creates "flow" (which is why you'll sometimes see them referred to as 'flow generators').

for any given amount of flow, if you decrease the amount of resistance (say with a mask leak) the pressure will drop, the machine will then increase the flow in order to maintain your pressure, if the leak is bad enough, the machine won't be able to increase the flow any more, and pressure will drop. for instance, if you take the mask off, the flow generator will ramp up to full speed, with a LOT of flow, but there will be negligible pressure.

as Pugsy mentions, the machines are programmed to reduce flow in order to reduce the pressure in the face of large leaks in order to try and get the mask to seal, before going back to normal operation.

That was a very useful (quality) post. I need to look into Pugsy's post regarding the PR ONE leak stats in Sleepyhead since I see duplicate sleep lines my graph at different levels. I don't see that with the Resmed.

[Pugsy]

That was a very useful (quality) post. I need to look into Pugsy's post regarding the PR ONE leak stats in Sleepyhead since I see duplicate sleep lines my graph at different levels. I don't see that with the Resmed.

With SleepyHead and the Respironics machine the top leak line is the "total" leak which is the vent rate plus any excess leak.
The bottom leak line is SleepyHead's stab and excess only leak reporting (which is what the ResMed one leak line is because ResMeds report only excess leak after subtracting a generic vent rate first) but it isn't as accurate as the ResMed excess only leak line because Mark found it extremely difficult to accommodate the changing vent rates with changing pressures and come up with calculations that matched what ResMed offers.
He gets close though. Doesn't really matter though because Respironics never gives us an exact line the sand like ResMed does with its 24 L/min line to avoid (which I think is a bit over conservative anyway). Large leak territory will continually change with pressure changes and mask differences and it's too difficult to compute an exact always 100% line in the sand. Best we can do is come close and trust the machine...if there are no large leak flags spit out on the events graph...then they didn't happen no matter what that top leak line says.

Also Respironics users need to change the default red line (Preferences/CPAP tab) to something other than the default to ResMed's 24 L/min or the leak statistics will tell you that you are having massive large leaks and you aren't having any large leaks at all. Either turn off that particular statistic (remove the check mark) or set it something closer to where large leak territory is on the Respironics machine and allow a bit of wiggle room for safety. Something like 70 L/min is probably close and still allow some room for error.

[cands]

He gets close though. Doesn't really matter though because Respironics never gives us an exact line the sand like ResMed does with its 24 L/min line to avoid (which I think is a bit over conservative anyway).

I think I remember reading in some technical document that Resmed machines start to respond differently when the leak is 30 l/minute or more. When I have leaks at this level apnoeas are described as UA - Unexplained Apnoeas, and the pressure stays lower until the leak rate drops.

[Pugsy]

I think I remember reading in some technical document that Resmed machines start to respond differently when the leak is 30 l/minute or more. When I have leaks at this level apnoeas are described as UA - Unexplained Apnoeas, and the pressure stays lower until the leak rate drops.

Wouldn't surprise me.
I know from my own personal experience with leaks at various stages above 24 L/min that the machine seems to do it job well until I hit around 30 L/min and then I also saw the "unknown" apnea flags. The machine knew something happened but it's sensing ability was compromised due to the leak. I figure if the sensing ability was in question then the responding ability also comes into question.
Over 35 L/min and I had blank spots where there should have been apnea events of some sort flagged. So I know things really get compromised at 35 L/min or more.

I don't get all in a tizzy with some time spent between 24 and 30 L/min leak. The machine is still able to do a decent job. It's when it gets over 30 L/min that I might take a second look to see how far it went and how long I was up there.