"Mask Fit" available on Dreammapper, but not Sleepyhead

It seems that the Dreammapper software gives this statistic, but I can't find a comparable one on Sleepyhead. It's towards the right hand side of the screen. Somehow, it can distinguish between poor mask fit and other leaks such as holes in the hose. I'm not sure how it would handle a well fitting nasal mask with the mouth open.

Am I missing something on Sleepyhead, or is this stat really not available?

. wrote:It seems that the Dreammapper software gives this statistic, but I can't find a comparable one on Sleepyhead. It's towards the right hand side of the screen. Somehow, it can distinguish between poor mask fit and other leaks such as holes in the hose. I'm not sure how it would handle a well fitting nasal mask with the mouth open.

Am I missing something on Sleepyhead, or is this stat really not available?

with sleepyhead, you have to *think* .... whereas it's dumbed down for you in dreammapper.

for you, stick with dreammapper.

D.H. wrote:It seems that the Dreammapper software gives this statistic, but I can't find a comparable one on Sleepyhead. It's towards the right hand side of the screen. Somehow, it can distinguish between poor mask fit and other leaks such as holes in the hose. I'm not sure how it would handle a well fitting nasal mask with the mouth open.

Am I missing something on Sleepyhead, or is this stat really not available?

The Check mask fit / Mask fit is done by measuring the amount of leak in the patient circuit. As stated in your manual. Nowhere does it state that it separates leaks. SleepyHead has a much more detailed look at your leaks.

All that 100% means is no large leak flags at all.
Since the events graph will show large leak flags (if you have them) then a person can easily see for themselves what their "mask fit" might be.

If you see 98 % for "mask fit" on DreamMapper all it means is that for 2 % of the night you had some time in large leak territory and the other 98% of the time you were below large leak territory. Now how far below....for that you have to look at the software reports.

palerider wrote:

. wrote:It seems that the Dreammapper software gives this statistic, but I can't find a comparable one on Sleepyhead. It's towards the right hand side of the screen. Somehow, it can distinguish between poor mask fit and other leaks such as holes in the hose. I'm not sure how it would handle a well fitting nasal mask with the mouth open.

Am I missing something on Sleepyhead, or is this stat really not available?

with sleepyhead, you have to *think* .... whereas it's dumbed down for you in dreammapper.

for you, stick with dreammapper.

Yeah but DreamMapper gives you a little trophy when you achieve 75% Mask Fit.

And everybody gets an award.
Sound familiar?

sc0ttt wrote:Yeah but DreamMapper gives you a little trophy when you achieve 75% Mask Fit.

here's your trophy:

Pugsy wrote:All that 100% means is no large leak flags at all.
Since the events graph will show large leak flags (if you have them) then a person can easily see for themselves what their "mask fit" might be.

If you see 98 % for "mask fit" on DreamMapper all it means is that for 2 % of the night you had some time in large leak territory and the other 98% of the time you were below large leak territory. Now how far below....for that you have to look at the software reports.

That's strange, because I'm seeing 100% mask fit on Dreammapper, but I'm not seeing 0% leak on Sleepyhead. In fact I'm seeing 6% there! Dreammapper is showing 0.5%, so at least that is within 1% of matching. Somebody said that this is a known issue with Sleepyhead, perhaps that's the discrepancy.

D.H. wrote:I'm not seeing 0% leak on Sleepyhead. In fact I'm seeing 6% there!

6% in the statistics column??
Are you seeing any large leak flags over in the Events graph?
Did you ever change the default from the ResMed 24 L/min red line?
SleepyHead defaults to the way ResMed reports leaks and uses the excess only 24 L/min as the red line. Respironics machines report total leak and that 24 L/min is way too low for how the leak is reported. It needs to be up around probably 70 to 80 L/min or just turn it off.
Go to Preferences/CPAP tab...change the 24 to something closer to where large leak territory is for total leak on a Respironics machine...there's no set number because total leak varies with the pressure and vent rate but 70 to 80 will get you close.
Or just turn that statistic off by removing the check mark.

Pugsy wrote:

D.H. wrote:I'm not seeing 0% leak on Sleepyhead. In fact I'm seeing 6% there!

6% in the statistics column??
Are you seeing any large leak flags over in the Events graph?
Did you ever change the default from the ResMed 24 L/min red line?
SleepyHead defaults to the way ResMed reports leaks and uses the excess only 24 L/min as the red line. Respironics machines report total leak and that 24 L/min is way too low for how the leak is reported. It needs to be up around probably 70 to 80 L/min or just turn it off.
Go to Preferences/CPAP tab...change the 24 to something closer to where large leak territory is for total leak on a Respironics machine...there's no set number because total leak varies with the pressure and vent rate but 70 to 80 will get you close.
Or just turn that statistic off by removing the check mark.

Yes the machine reports total leaks, but when it's reading data from one of those machines Sleepyhead calculates the "time over redline" percentage using the calculated unintentional leaks (which is calculated from the mask leak model as set in preferences). I've verified this many times with my data - even a quick eyeball of the unintentional leak curve shows this. Only when the unintentional leaks (bottom line) go above the redline does it add to that percentage. Now whether a 24L/min redline leak is too much for the PR machines to handle, I don't know..but I'd have no reason to expect it would be drastically different from that of the Resmed machines and I see no evidence that it has any trouble at those levels (and beyond).

If this wasn't true, my numbers would show 100% over redline every night, because my total leaks are always above 24L/min for virtually any mask, at any pressure. It doesn't. It shows exactly what I would expect - percentage unintentional leak over the redline of 24L/min. It just does so with the calculated number, which is every bit as good (and quite likely better) than what the Resmed machines would be able to do, given they don't know what mask you have, and its leak curve.

linuxman wrote:Now whether a 24L/min redline leak is too much for the PR machines to handle, I don't know..but I'd have no reason to expect it would be drastically different from that of the Resmed machines and I see no evidence that it has any trouble at those levels (and beyond).

Respironics has never given us any sort of red line number. We are given total leak and we can figure out vent rate and we can come close to figuring out where our own personal "red line" might be.
My personal experience with several various Respironics machines (and ResMed) is that the red line excess number for a Respironics machine is going to be up around 60 to 70 with my mask and my usual pressures and total leak has to hit around 90. Now I have seen large leak flags on a few people's reports where total leak was 75 to 85.

That's why I tell people who want to use that red line statistic on their SH software and they are using a Respironics machine and SH is telling them that they are spending X amount of time in large leak AND there's not any large leak flags over on the Events graph...to change the red line because it isn't accurate and usually it has scared the bejesus out of someone. "OMG SH says I am leaking like crazy but I get 100% for mask fit and the graphs don't show it".
I have had SH tell me I spent 20% of the night in large leak when using my PR S1 with zero large leak flags on the events graph. The 20% statistic is in error because it is based on a too low red line number.

The large leak flags on the Events graphs trumphs the red line % data point.
DH is seeing the machine report minimal large leak...or no large leak and I assume no large leak flags but SH says 6% of the time in large leak. The 6 % is in error because it is based on an incorrect base line.

Most of the time people aren't going to go past 24 L/min excess leak on their Respironics machines and that's why for the most part the % over red line is very small or 0% when no large leak flags are present.
The problem is this...if a Respironics user happens to hit 25 or 26 or even 36 L/min excess leak then SH is going to give them a "you went over the red line of 24 L/min this amount of time" message and it can cause undue panic or stress. They never went into Respironics large leak territory...just ResMed's large leak territory.
Since we are never given an exact red line number from Respironics we have to come up with our own if we want to have any sort of accuracy in that particular red line % statistic.

Based on my own personal experience with Encore (and Encore will do a total leak graph or an excess only leak graph) and SH and too many Respironics machines to count and the thousands of reports from other Respironics users I have seen...I am confident that Respironics machines can handle 40 to 50 L/min of excess leak and maybe more before they will flag a large leak.
I also know that the machine doesn't really have a huge issues until total leaks exceed 100 L/min...once someone goes past 100 L/min (which is probably 60ish excess L/min) things go to hell pretty fast. At 110 total leak I have seen apaps lose auto adjusting capabilities and people think their machines are broken.

So...for the most part most Respironics users don't go past the ResMed red line of 24 excess leak that SH uses and those people will see 0% if the time in large leak and never know they are using a baseline number that is actually incorrect for their machine. For the few that happen to go above 24...unless they get a large leak flag from their machine that shows up in the events graph that % is incorrect.

DH was wondering why the difference of opinion on the statistic reported and the machine reported leak.
I offered a potential explanation for the discrepancy.

And no..the Respironics machine doesn't have a clue what mask we are using so it really is most likely using a generic vent rate number with whatever pressure is being used to get in the large leak ball park figuring.

Pugsy wrote:

linuxman wrote:Now whether a 24L/min redline leak is too much for the PR machines to handle, I don't know..but I'd have no reason to expect it would be drastically different from that of the Resmed machines and I see no evidence that it has any trouble at those levels (and beyond).

Respironics has never given us any sort of red line number. We are given total leak and we can figure out vent rate and we can come close to figuring out where our own personal "red line" might be.
My personal experience with several various Respironics machines (and ResMed) is that the red line excess number for a Respironics machine is going to be up around 60 to 70 with my mask and my usual pressures and total leak has to hit around 90. Now I have seen large leak flags on a few people's reports where total leak was 75 to 85.

Sure, it's possible that the Philips machines can handle higher flow (which would mean higher loss tolerance overall) or lower flow, and I really can't say one way or the other. In fact, I find it unlikely that 24L/min is really a valid cutoff for the redline for the Resmed machines. I seriously doubt that there's any significant compromise in performance at that level given the overall flow rates we're talking about here, and the ease with which the breathing signal can be filtered from the overall flow. I can't knock it though, since apparently that's what Resmed picked. They're probably just trying to pick a conservative number.

Agreed, on the large leaks flag and its ambiguity. I guess I just don't care about that. I'm an engineer. I just want the numbers for actual leaks so I can make my own determinations about how I''m doing on mask fit, mouth breathing, etc. With total leaks provided by the machine, I can get that, and tailored specifically to the mask in use. I'm practically never going to be in large leak territory. For others that want a "go/no-go" type outcome... I can understand why they have the flags, though.

linuxman wrote:I find it unlikely that 24L/min is really a valid cutoff for the redline for the Resmed machines.

I totally agree. I think that ResMed decided on a number and a really conservative number to have as a red line to avoid.

I once had a night with some substantial leaks (new type of mask) where I was able to watch the progressions of leaks from none to fairly big in a manner that made it real easy to see what happened to the machine in terms of reporting and responding all in one nice graph. This was with one of my ResMed machines.

What I found was that excess leaks between 24 to 30 L/min pretty much nothing changed...events were flagged appropriately and the machine pressure responded as expected. With leaks 30 to 35 L/min range I started seeing some "unknown apnea" flagging..meaning the machine could sense something was going on but had trouble deciding what to call it...but the pressure line still appeared to be responding as I would expect.
Once I hit over 35 L/min and it went up to 45...I stopped seeing any event flags where right before that leak I was seeing unknown apneas. I have to assume that the machine couldn't sense accurately and the pressure line also pretty much just stayed fixed because the machine didn't know if it need to change the pressure around or not.

ResMed has decided to use the 24 L/min red line in the sand. It's probably an ultra conservative number in the sand and it's also probably good that they did that number (since they wanted to do some sort of number) because people seem to always want to push the limits with any line in the sand.

Pugsy wrote:

linuxman wrote:I find it unlikely that 24L/min is really a valid cutoff for the redline for the Resmed machines.

I totally agree. I think that ResMed decided on a number and a really conservative number to have as a red line to avoid.

I once had a night with some substantial leaks (new type of mask) where I was able to watch the progressions of leaks from none to fairly big in a manner that made it real easy to see what happened to the machine in terms of reporting and responding all in one nice graph. This was with one of my ResMed machines.

What I found was that excess leaks between 24 to 30 L/min pretty much nothing changed...events were flagged appropriately and the machine pressure responded as expected. With leaks 30 to 35 L/min range I started seeing some "unknown apnea" flagging..meaning the machine could sense something was going on but had trouble deciding what to call it...but the pressure line still appeared to be responding as I would expect.
Once I hit over 35 L/min and it went up to 45...I stopped seeing any event flags where right before that leak I was seeing unknown apneas. I have to assume that the machine couldn't sense accurately and the pressure line also pretty much just stayed fixed because the machine didn't know if it need to change the pressure around or not.

ResMed has decided to use the 24 L/min red line in the sand. It's probably an ultra conservative number in the sand and it's also probably good that they did that number (since they wanted to do some sort of number) because people seem to always want to push the limits with any line in the sand.

Yeah I did same with a really leaky night I had a while back. I think it was a FFM where I'd lost seal around my chin and was blowing around 50L/min excess. Somehow it didn't wake me up . I looked very carefully at every breath over the whole duration and no sign of any missed events. It's not too surprising, really. I know the freecpapadvise.com guy did a video showing successful pressure maintenance up well over a 100 L/min, I think using a PR System One. So even if detection gets crappy, it's improbable that there are any actual issues with therapy for straightforward mask leaks. The overall range of measurement here is not so large that it's likely to max out the A/D converter on the MAF sensor (assuming they use one of those..could also just be motor rpm as that's a proxy for air flow rate). I'd think a bigger issue would be when there is rapidly *changing* leak rate. Those could be hard to distinguish from breaths in some case. Fortunately changes in leaks should be rare, and thus is constant in the short term and easily filtered with a simple low pass filter.

linuxman wrote: MAF sensor (assuming they one of those..could also just be motor rpm as that's a proxy for air flow rate).

what??? no, sensirion 1-100544-01 flow sensor

linuxman wrote: I'd think a bigger issue would be when there is rapidly *changing* leak rate. Those could be hard to distinguish from breaths in some case.

one would think not, leaks never push back like exhalations.