Swift LT pillow system leak data

[PaulyWally]

Hi. I'm trying to determine what leak rate is acceptable for the Mirage Swift LT pillow system. Can anyone point me to this data?

Thanks

[feeling_better]

Hi. I'm trying to determine what leak rate is acceptable for the Mirage Swift LT pillow system. Can anyone point me to this data?

Thanks

If you have the manual, it is on page 9 for English.
Pressure 4: 20lpm
8:28 lpm
12:33 lpm
16:38 lpm
20:42 lpm

[PaulyWally]

I believe these are the vent flow rates which is not what I was looking for. My understanding is that every mask/pillow/etc
have what might be referred to as an acceptable leak rate. Obviously you want zero leakage but a certain leak rate on a full face mask has a different significance than the same leak rate on nasal pillows. However, I'll be the first to admit that I dreamt this up when I managed to fall asleep.

[jules]

Post your pressure, the mask that is in the "menu", and leak rate off the LED.

For the Resmed machines you have to do a conversion of "excess" leak in l/sec to actually leak in l/min.

[PaulyWally]

Post your pressure, the mask that is in the "menu", and leak rate off the LED.

For the Resmed machines you have to do a conversion of "excess" leak in l/sec to actually leak in l/min.

Pressure is 8cm, mask (via menu) is Swift (but really Swift LT), and leak rate has varied but has
averaged about 0.3 l/s

Thanks

[jules]

So for the swift (menu) at 8 cm the machine is using a built in leak rate of 29 l/min (I used page 11 of this pdf - note the mask flow are +/- 6 l/min

http://www.resmed.com/en-us/products/se ... _multi.pdf)

your excess over that you stated was 0.3 l/s - this needs to be converted to l/min and added on

so to determine your total flow you take 29 + 0.3 * 60 or around 47 l/min.

you expected flow according to the post above by feeling_better is 28 l/min

So I would say you have a big leak problem as your average flow is about 67 % higher than it should be.

To be around only 20 % higher than you should be you need to have your LED leak number to be around 0.08

All of these calculations are crude but there is the problem, you got leaks!!!!!!!!!!!!! An actual graph might provide more insight.

[PaulyWally]

So for the swift (menu) at 8 cm the machine is using a built in leak rate of 29 l/min (I used page 11 of this pdf - note the mask flow are +/- 6 l/min
http://www.resmed.com/en-us/products/se ... _multi.pdf)
your excess over that you stated was 0.3 l/s - this needs to be converted to l/min and added on
so to determine your total flow you take 29 + 0.3 * 60 or around 47 l/min.
you expected flow according to the post above by feeling_better is 28 l/min
So I would say you have a big leak problem as your average flow is about 67 % higher than it should be.
To be around only 20 % higher than you should be you need to have your LED leak number to be around 0.08
All of these calculations are crude but there is the problem, you got leaks!!!!!!!!!!!!! An actual graph might provide more insight.

I understand your calculations but it appears that the built in leak rate you've quoted of 29 l/min (from page 11) is already 1 l/m above the rate of 28 l/min posted by feeling_better (above). I think your number is for the Swift II whereas feeling_better is the rate for the Swift LT.

[jules]

different masks have different leak rates - you said you were using the LT but I bet the built in mask menu uses the old swift I or swift II

The 1 L/min isn't the big problem - it is the 18 L/min you are getting from that LED leak rate of 0.3 that is the problem. On average during the night your flow rate is over 60 % higher than it should be. You got problems and it might be mouth breathing or might be the mask is coming loose and not sealing well.

[bigk]

The instructions for the Swift LT say to use Swift as the mask setting on a Resmed machine.

On the first night I unbelievable got a leak rate of 0.00 l/s. I have had leak rates as high as 0.06l/s with this mask with most days being 0.02l/s (last 2 weeks I have had it)

A leak of 0.3 to me indicates you need to try other size pillows.

I also found that tightening the straps a little (the rear one) helped a lt as I originally had it too loose.

[feeling_better]

The instructions for the Swift LT say to use Swift as the mask setting on a Resmed machine.

That is an approximation. When a new mask, such as Swift LT, becomes available, the exact leak data for that one will not be in a previously released machine. So the manufacturer of the new mask will pick another model which is already in the data tables, which is close in leak value. In this case the actual leak of LT at 8cm from their specification is 28lpm. What that machine is using 29lpm, so it will under report the leak by 1 lpm. For example suppose the total flow was 30 lpm, meaning there was a leak of 30-28 = 2 lpm. But the machine only knows the 29 number, so it would report a leak of 30-29=1lpm. (Of course your machine reports leaks in another unit, lps).

As jules said, this 1 lpm error is almost negligible, that is why Resmed recommends using the Swift as the reference mask name.

I do not have experience with your machine, but my Respironics M-series (with add on software) gives a graph of the leak over the entire night. And the variations in there is the best way to see the leak. If there was no leaks at all ( seldom achieved), the graph would be very close to a straight line.

[feeling_better]

One additional comment on leak reporting. It is possible and likely, if you are using c-flex, a-flex or other pressure relief during exhale, the actual leak reported will be less than the real leak. This is because the leak numbers used by most machines for the expected (intended) leak (the normal out port flow) is based on the pressure setting. Let us say your pressure for example is 8cm. The average pressure under pressure relief methods will be a bit lower than 8cm, and hence the intended flow should also be a little less than the value of leak from the table for 8cm. For a very detailed careful analysis of this, please see my thread:

viewtopic.php?t=33829&highlight=

[bigk]

0.00 again last night and you're right - the leak graph is basically a straight line. Even if it is under-reporting leak a little it's still so small as to be negligable.

[Abbi-Normal]

I am having trouble understanding all of this and how the 'conversion' works. I am also unsure about the pressure relief during exhale. Where do I find that data on my machine, or is it even there?

My L/S over the last 9 days has ranged from .2 to 1.1 with it typically being .4 or .5
Are you saying that those readings indicate major leaks?

I am really trying to understand how to intrepret the data but it can be overwhelming especially when you are still trying to learn all the codes and abbreviations. The one code I DO know is CRS

[RiverDave]

Which conversion are you referring to?

A leak rate greater than about 0.3 is pretty high. But that number can be misleading. Is the machine only on when you are hooked up to it? If yes, then you have a bad leak problem. If no, the number is suspect because while you were unhooked from the machine and it was running, it was recording massive leaks. The number you see is the leak rate you were at or below for 95% of the time the machine was on.

[jules]

Which conversion are you referring to?

A leak rate greater than about 0.3 is pretty high. But that number can be misleading. Is the machine only on when you are hooked up to it? If yes, then you have a bad leak problem. If no, the number is suspect because while you were unhooked from the machine and it was running, it was recording massive leaks. The number you see is the leak rate you were at or below for 95% of the time the machine was on.

I don't have a Resmed. I have a Tank APAP and run EP 1.4 (yeah I know obsolete but it works and I can even move the database between computers with no hassles).

I know the LED pressure information may include both a 95 %ile and the average. However, I was under the understanding, perhaps a mistake on my part, that the leak number was an "average" not the 95 %ile value.