If the leaks below in my example disturbed my sleep...I would want to fix them but since they don't and I rarely have a leak wake me up...I just shrug my shoulders and move on.
Leaks....and that ResMed 95% number or from a Respironics machine and using Encore 90% number...all that means is that for 95/90 percent of the time the leaks were at OR below that number. People tend to overlook the OR BELOW part of the definition.
It's just a number that gives us a sort of idea how the leaks went that night. If it's low...we know that leaks were minimal but if it is elevated it may or may not be statistically important...remember the OR BELOW part of the definition.
Past experience has shown me that the 95/90 percent numbers are easily skewed by a short time in larger number (either pressure or leak numbers) so the % numbers may not be as significant as first thought.
Now I am not saying they are never significant...I am just saying they aren't always the holy grail of numbers to look at.
It's just a number and a person needs to evaluate how that number might be significant (or might not) by looking at the overall graphs.
Here's an example from last month. 95% number was a whopping 27.6 L/min and one might think the whole night was trash due to leak if looking at that number and forgetting the OR BELOW part of things.

But when I go look at the actual leak graph it isn't nearly as horrible looking. I barely went above 24 L/min for maybe an hour or so and had a short time with some leak spikes. I suspect my mouth was open part of the time. I don't know for sure what happened because I normally sleep through stuff like this. This report is a prime example of how the 95% number could possible cause alarm because on the surface it might point to a horrible leak night when in reality the bulk of the night was really quite acceptable and often zero. If I was mouth breathing....it wasn't enough to cause the entire night to go to hell.
