Understanding leak report

[SnoozyQ]

Hopefully leak issues will be eliminated with new mask...but can I get a little help with leak readings? What does the median, maximum and 95th percentile mean? What is my actual reading? It says on the graph 0.4 L/sec=24 L/ min.

My readings since Jan 12 are:

76.0
7.2
21.6
15.6
81.6
64.8

[avi123]

Hopefully leak issues will be eliminated with new mask (I doubt it) ...but can I get a little help with leak readings? What does the median, maximum and 95th percentile mean? What is my actual reading? It says on the graph 0.4 L/sec=24 L/ min.

My readings since Jan 12 are:

76.0
7.2
21.6
15.6
81.6
64.8

As I understand it, the 0.4 L/sec = 24 L/min is the maximum leak allowed (by professionals) for a "mask leak", without including the intentional leak for the specific mask (comes in a graph on paper in the same package as the mask). This intentional leak is thru those drilled holes that you suppose to clean and don't suppose to cover.

Suppose that for your Quattro mask you read in its graph that at a pressure of so and so in cm H2O (that the CPAP is set on) the intentional leak is 30 L/min. So you add the two: 24 plus 30 = 54 L/min. In such a case those readings that are above it are too hi. You need to find out their sources and eliminate them. Your Mouth Leak could be the highest leaker. However, b/c you use a full face mask it should be MUCH less than those CPAPniks using nasal masks (and thus jeopardizing their treatments).

[robysue]

Hopefully leak issues will be eliminated with new mask...but can I get a little help with leak readings? What does the median, maximum and 95th percentile mean? What is my actual reading? It says on the graph 0.4 L/sec=24 L/ min.

As a previous poster noted, the leak data on the ResMed S9 machines has the mask's intentional leak rate data subtracted out of it before it is reported in ResScan (or on the S9's LCD in the Sleep Quality Report). So the leak data represents your unintentional leak rate as detected by your S9. The S9 samples the flow and compares it to the expected flow to figure out the leak rate multiple times a minute. [I want to say that it samples the leak rate every two seconds, but I really don't remember any more since I no longer use an S9.]

Now for a brief math lesson.

The median leak rate in ResScan is the rate at which you are leaking AT or BELOW for 50% of the time the machine was running for the night. So if your median leak rate is 10 L/min and your S9 was running for 7 hours, that means you were leaking at a rate of LESS THAN or EQUAL to 10 L/min for 3 1/2 hours during the night. And that also means that you were leaking at a rate GREATER THAN or EQUAL to 10 L/min for 50% of the night as well (3 1/2 hours if you ran the machine for 7 hours.)

The 95% leak rate in ResScan is the rate at which yoiu are leaking AT or BELOW for 95% of the time the machine was running for the night. This also means that you were leaking at a rate AT or ABOVE the 95% leak rate for 5% of the time the machine was running. To get an idea of how long 5% of the night is, remember that five percent of one hour is three minutes. So if your 95% leak rate is 15 L/min and your S9 was running for 7 hours, that means you were leaking at a rate that was LESS THAN or EQUAL to 15 L/min for 6.65 hours (roughly 6 hours and 40 minutes). And you were leaking at a rate that was GREATER THAN or EQUAL to 15 L/min for 5% of the night. Now 5% of 7 hours is (3 minutes * 7 hours = 21 minutes). So you were leaking at a rate of more than 15 L/min for no more than 21 minutes all night. Now you might have had one gigantic 21 minute leak or you might have had a series of 5 minute leaks. There's no way to tell unless you look at the detailed leak graph. A final note about the 95% leak rate: This is the number that is shown in the Detailed Sleep Report on the S9's LCD.

The maximum leak rate is just that: It's the maximum leak rate of all the sampled leak rates for the night. In other words, it's the single biggest leak the S9 happened to catch. It could easily be a very short-lived leak that was caused when you broke the seal of the mask to scratch your nose and the S9 just happened to sample the leak rate at that particular point in time. You can usually ignore the maximum leak rate.

The Red Line on the detailed graph that says 0.4 L/sec = 24 L/min indicates the unintentional leak rate where the Resmed engineers believe that therapy and data collection on the S9 begins to become compromised. Short leaks above that level are one thing. But consistently leaking above that level is a real problem. That's why the Red Line is there: If you have LONG LEAKS above that line, you need to start thinking about how to address them. How long is a LONG LEAK? Well, Mr. Red Frowny Face shows up in the Short version of the Sleep Quality report on the S9's LCD if your 75% leak rate is AT or ABOVE 24 L/min. In other words, if your leak rate is AT or ABOVE 24/min for an average of 15 minutes for every hour the machine is running, then the Resmed engineers believe the leaks are serious enough that they may well be compromising your therapy and the data gathered by the machine. So if you've got a 15 minute (or longer) leak above that Red Line, that's an indication of a problem

And 0.4 L/sec is the same leak rate as 24 L/min if you do the conversion of units. L/sec stands for "liters per second" and L/min stands for "liters per minute." Since there are 60 seconds in a minute, to convert L/sec to L/min, you simply multiply the L/sec by 60 to get the same rate converted to L/min.

My readings since Jan 12 are:

76.0
7.2
21.6
15.6
81.6
64.8

Without knowing whether these are median, 95% leak rates, or maximum leak rates, it's impossible to give you any feed back on what these numbers might mean.

But I'll take a wild stab in the dark: If they are the 95% leak rates that came off the S9's LCD Sleep Quality Menu, then you've got wildly varying leaks. Some days are good enough (7.2, 15.6) and some are really awful (76.0, 81.6, 64..

But if those numbers are max leak rates, then the data for the days where the max leaks are 7.2, 15.6,, and 21.6 all indicate really good days, but the rest of the data are meaningless because those max leaks could easily have been triggered by something as simple as you needing to scratch your nose.

[SnoozyQ]

Thank you for the explanation. I will look at that tonight along with the graphs so I can translate them

Sorry, the numbers I gave were the top numbers on the graph--"Maximum".

I believe my red line is at 20.

I used my new mask last night, and I know that I didn't take it off for any reason during the night. I didn't hear or feel any leaks, but my LCD displayed 43L/min...........

[tektrek68]

I just switched to an S9 to where I can monitor my sleep and have noticed avg leak rates over 30. Suspecting mouth breathing I tried a chin strap for the first time last night and leak was down below 5. I am hoping I can retrain myself to keep my mouth closed so I dont have to wear the strap forever

[SnoozyQ]

I'm on the other end of the spectrum.....I clench my teeth when I sleep.

[robysue]

Thank you for the explanation. I will look at that tonight along with the graphs so I can translate them

Sorry, the numbers I gave were the top numbers on the graph--"Maximum".

As I stated in my first post, the Maximum leak rates are largely irrelevant because a really large one can be caused by a really short lived leak that you know about: You broke the seal to scratch your nose is an example. Or you broke the seal to re-seat the mask to take care of a smaller, but bothersome leak. Or you took the mask off before turning the S9 off. Or ....

The 95% leak rate is the one you really want to pay attention to if you're only going to look at one number. It's the middle number on the bar graphs. Even better is to look at the median and 95% rates together.

I believe my red line is at 20.

Unless ResMed has changed the placement of the line, it should be at 24L/min. But based on the vertical scale, it may be tough to see the difference between 20 L/min and 24 L/min.

I used my new mask last night, and I know that I didn't take it off for any reason during the night. I didn't hear or feel any leaks, but my LCD displayed 43L/min...........

It is CRITICAL to make sure you properly SET the mask setting on the S9 each and every time you change masks! That mask setting is what makes the S9 and ResScan set the 0 line in the leak data correctly. Remember on the S9, a leak rate of 0.0 indicates that the S9 is detecting NO LEAKS above the intended leak rate for the mask at the given pressure setting. And that's the ideal situation! But different mask types have different leak rates at different pressure settings.

Your LCD display (Leak 43 L/min) means that for 95% of the time the machine was running last night your UNINTENDED leak rate was AT or BELOW 43 L/min. And that means that for 5% of the night your UNINTENDED leak rate was AT or ABOVE 43 L/min. (And remember that 5% of one hour is three minutes. So multiply the usage time by three minutes to find out how long you were leaking at a rate AT or ABOVE 43 L/min.)

Now is a 95% leak rate of 43L/min too high? Well---that depends. Strictly from a therapy point of view and from the Resmed engineers point of view: Did Mr. Red Frowny Face show up this morning? If so, then your 75% percentile leak rate was AT or ABOVE 24 L/min and that means that the Resmed engineers believe the leaks were both LARGE enough and LONG enough to both compromise your therapy and the machine's ability to gather data. On the other hand, if Mr. Green Smiley Face showed up, your 75% leak rate was still below 24 L/min and that means that the leaks that triggered the high 95% leak rate number (43 L/min) were short enough where the total time involved was under an average of 15 minutes per hour---probably well under 15 minutes per hour, but over an average of 3 minutes per hour. And in this case, the Resmed engineers are not worried about your leaks. But should you be worried about them? Well that depends: If the leaks bothered you enough to wake you up or keep you up, then you need to worry about them and figure out a way to fix them. Otherwise, maybe you can put off worrying about them until there's nothing else you need to worry about concerning CPAP therapy.

Hope that helps

[SnoozyQ]

Wow Robysue, you have this stuff down pat! LOL Will you be my sleep doctor? LOL

By new mask, I mean replaced the old mask with the same type, just one size smaller.

I admit, I don't even check the smiley face....I go straight to the detailed data for AHI and leaks. I will try to check Mr. Face and compare the details. I'm also trying to note days I feel great and days I feel like poo, so I can go back and see if they coincide with good/poor data.I had a not so great morning as far as waking goes, but feel ok now. My AHI last night was 0.8.

[tektrek68]

I have noticed AHI can be reported innacurate if you have major leaks

[robysue]

Wow Robysue, you have this stuff down pat!

Numbers are my first love.

I admit, I don't even check the smiley face....I go straight to the detailed data for AHI and leaks. I will try to check Mr. Face and compare the details.

Just look for him when you're holding down both the info button and the double check mark button for the magic 3 seconds to get the detailed data to load. Mr. Smiley and Mr. Frowny are kind of hard to miss .... And since you seem to be chasing leaks, you really do want to note if/when Mr. Frowny shows up because even ResScan doesn't give you a NUMBER for that all important 75% percentile leak rate. The meaning of Mr. Red Frowny Face in terms of the 75% leak rate number is buried deep in the clinician's manual somewhere if I recall correctly.

I'm also trying to note days I feel great and days I feel like poo, so I can go back and see if they coincide with good/poor data.I had a not so great morning as far as waking goes, but feel ok now. My AHI last night was 0.8.

This is a critical idea in my opinion. See what seems to correlate to great days and "poo" days. See if the data that corresponds to great days can be reproduced and the data that corresponds to the poo days can be avoided.

By the way, notes on how I feel have played a critical role right from the start of my four month long slog through adjusting to xPAP therapy---both in the sense of addressing real issues in adjusting my xPAP therapy as well as in getting me to the point where I could see that by the end of December (after the switch to the BiPAP) my really lousy days were now corresponding to insomnia problems, not xPAP problems. Those notes are why I'm pretty sure my body has now physically made the adjustment to sleeping with the BiPAP every night now. But the remaining problems I have with fragmented sleep still have to be dealt with before I'll truly feel refreshed when I wake up on many (most? all?) mornings. And the sleep log my PA is having me keep is indeed helping me see patterns in what helps reign in the insomnia and what tends to feed the insomnia.