EncorePro 2 giving incorrect leak numbers

GumbyCT wrote: I know you know - can you esplain to Doug? I think the numbers are close enuff and this was all human misinterpretations. And not comparing apples to apples.

Of course unless and until we see the graphs that Doug saw we will never know. The reason he doesn't want to show them raises the eyebrows. Hey, I gave it my best WA

I am thinking that Doug's EP leak line (while reporting unintentional leak) probably looked a lot like this one below. Only it probably started just a bit closer to the 20 L/min line instead of 10. This is an Unintentional Leak report using EP but on an M series machine. When we toggle to Total Leak in EP the leak line will start at around 45 to 50 L/min. I would have to sort through a ton of emails to find one when Total leak was chosen in EP to get the exact number where it starts. Maybe later I will look. This person obviously doesn't have much leak going on here because not many blips. This is also the Quattro FX.

We all know that it doesn't make any difference what the "average" is as long as the line is pretty much straight. That is why I don't get hung up on "averages" all that much.

I am thinking that baseline leak line may simply start (on the graph) at a higher L/min number and not zero when using Full Face masks. Since this is the only report I have to look at I guess this has to be sufficient. So I am not doubting what Doug says he saw. The average unintentional leak on this report is around 10 but it is obvious that there wasn't much leak going on. I think it is just where the machine starts reporting leak after whatever magic thing it does inside to determine what it thinks baseline should be. If it starts at 20 L/min and doesn't vary much at all the overall leak (even unintentional) will show as 20 L/min because that is where it started.

Okay, this itched has been scratched and I don't want to scratch it anymore. If it bugs Doug to see the variable, then by all means use Total Leak in the report. These are all just WAGs anyway. Main thing is to watch for leak spikes that might impact therapy or sleep no matter what the average number might say. Straight leak line is good and that is all that matters to me. The machine has no clue what mask is being used much less what the vent rate is. So it gives us a close WAG. Far from perfect but it is what it is.

GumbyCT wrote:

DougVK wrote:I am positive it doesn't vary from mask to mask as it has no idea what what is being used.

Each mask will in fact vary as will each machine, believe it or not.

How could the software possibly know what mask is being used.

DougVK wrote: I posted the average leak numbers which should be close enough. The numbers I posted shows the software thinks my mask has an vent rate of around 27 when the vent rate is 38 at those pressures. This is not a human misinterpretation.

Please explain how Vent rate + Unintended leak != Total Leak. You fail to explain how the software could possibly know my mask vent rate when it isn't something I tell it.

Take a breath Doug.... I was looking for the graph not numbers. Numbers are misleading, you will learn.

The vent rate you are speaking of is a "constant" in the software. I/we don't know what it is but it is the only thing that has not changed in all of this.

You need to spend some time reading before you go off on those who are trying to help you. The good news is IF you can get your therapy right - the mind will clear and the mood swings will lessen.

Good Luck

The Vent Flow Rate and the Total Leak should be fairly close to each other.
Disregard the "Unintended Leak" data. Don't even look at it.


Den

GumbyCT wrote:

DougVK wrote: I posted the average leak numbers which should be close enough. The numbers I posted shows the software thinks my mask has an vent rate of around 27 when the vent rate is 38 at those pressures. This is not a human misinterpretation.

Please explain how Vent rate + Unintended leak != Total Leak. You fail to explain how the software could possibly know my mask vent rate when it isn't something I tell it.

Take a breath Doug.... I was looking for the graph not numbers. Numbers are misleading, you will learn.

The vent rate you are speaking of is a "constant" in the software. I/we don't know what it is but it is the only thing that has not changed in all of this.

You need to spend some time reading before you go off on those who are trying to help you. The good news is IF you can get your therapy right - the mind will clear and the mood swings will lessen.

Good Luck

I only said the software is using an incorrect vent rate for my mask and the unintended leak numbers are significantly off. I only wanted to know how I could get the vent rate in the software to be correct for my mask. You keep saying I was wrong, that I was reading the numbers wrong. I wasn't wrong the software is using a vent rate that is very different than my mask. I don't know why you cannot accept that my mask has a significantly higher vent rate than what the software uses.

it appears that EP 2 uses more information about a mask - there is a resistance number that is put in I understand

If you look at Pugsy's graph, I wouldn't want a "leak rate" (meaning flow rate) of 12 with a pressure range of 16 - 18.

Would you? I don't know how EP decides what to use for it's baseline flow rate, but at least in Pugsy's case it isn't zero like the older versions of EP / EV.

now gumby, can I please go back into hiding and occasionally lurk here? pretty please? or I will send a tornado your way ASAP

To Whom It May Concern:

I would never expect a newbie (with a FFM) to have a straight leak line (esp. w/FFM).

The average numbers will be misleading. Now some here might take that to mean they were rwong. Not rwong, just on a very steep learning curve.

DougVK wrote: I posted the average leak numbers which should be close enough. The numbers I posted shows the software thinks my mask has an vent rate of around 27 when the vent rate is 38 at those pressures. This is not a human misinterpretation.

Please explain how Vent rate + Unintended leak != Total Leak. You fail to explain how the software could possibly know my mask vent rate when it isn't something I tell it.

I too consistently have "Total leak rates" that are below the advertised expected leak rate for my mask at my pressure. And if I toggle to from "Total Leak Rate" to "Unintended Leak Rate", it's very, very clear that EncorePro is NOT just subtracting of that published number for the intended leak rate. No surprise actually, since there is NO place in either the software nor the machine's set up menu that tells the System One (or Encore Pro) what mask I'm using.

So what the software IS doing is some statistical analysis to estimate what it thinks the unintended leak rate that night was and it subtracts that number off the Total Leak Rate curve to draw the Unintended Leak Rate curve. Now, is that statistical analysis got any basis in reality? No one on this board actually knows the answer to that question because Resprionics has NOT published anything in the documentation of EncorePro about how EncorePro determines the estimated Expected Leak Rate.

My best guess (as a mathematician, but NOT a stats person) is that Encore Pro does some kind of statistical analysis for identifying the "lowest flat" parts of the Leak Data that last for a sufficiently long period of time. And it discards any data that is above a certain number of L/min higher than the data it has identified as the "lowest flat parts" before re-averaging the data that appears to be close to being leak-free because it's flat and low or close to being flat and low and then it most likely adds back in some kind of margin of error. But once Encore Pro does come up with its magical guesstimate for the intentional leak rate, it subtracts that off the Total Leak data, but sets any resulting "negative" numbers to 0.

And what's the point of doing this kind of statistical analysis? That's a very good question. And to be honest, I don't have a good answer. Other than it takes a "fuzzy flattish" Total leak curve at around n.d L/min and turns it into a very nice, ultra-flat line right at 0.0 L/min.

Me, personally? I just look at the Total leak graph, with more of an emphasis on how flat it is (and fuzzy flat is fine) rather than worrying over much about the numbers. In other words, I worry much more about "big bumps" and other irregularities in the leak data than what the value of the base of the Total leak graph seems to be.

jules wrote:it appears that EP 2 uses more information about a mask - there is a resistance number that is put in I understand

If you look at Pugsy's graph, I wouldn't want a "leak rate" (meaning flow rate) of 12 with a pressure range of 16 - 18.

Would you? I don't know how EP decides what to use for it's baseline flow rate, but at least in Pugsy's case it isn't zero like the older versions of EP / EV.

That report is not mine. It is someone else who wanted to feel "closer" to zero leak rate so we switched EP2 to report only unintentional leak. If we switch it back to normal Total leak it will show somewhere between 45 to 50 L/min. Which goes along with that 18 cm pressure. I cleaned out my downloads and I don't have easy access to similar report showing Total Leak but trust me, the report posted is perfectly normal for the way it was reported in the software. Nothing to be alarmed about. We just played with the software choices to get closer to that zero line that I have just for visual pick me up for someone struggling.

Resistance number in the PR S1 machines doesn't seem to change whatever magic number is used to create the vent rate base. For my own testing purposes the Resistance setting was at 0. In EP 2.4 there is no where to allow for any Resistance settings. They are on the machine only. I read what it did but it wasn't used in figuring leak rates itself.
I will go try to find it.

Here is part of it.Optimized pressure delivery, no matter which mask is used.
Remember also that the choices that the machine undestands are all Respironics masks.
No where does it say that this affects leak reporting. Only that it tries to compensate for resistance that each mask has and that resistance is listed next to pressure graphs.

System One Resistance Control achieves complete system comfort by enabling the device and mask to work optimally together. Through simple settings, the technology instructs the device to compensate for variable resistance characteristics related to different masks. The result is synchronous pressure delivery and the assurance patients can enjoy the full benefits of our Flex comfort technologies no matter which Philips Respironics mask is used.
http://systemoneresistancecontrol.respironics.com/

GumbyCT wrote:To Whom It May Concern:

I would never expect a newbie (with a FFM) to have a straight leak line (esp. w/FFM).

The average numbers will be misleading. Now some here might take that to mean they were rwong. Not rwong, just on a very steep learning curve.

They are wrong, I have already shown this. I know the average is slightly misleading but they shouldn't be off by over 10 L/min.

1. You insist the software knows what my mask vent rate is even though I never input it.
2. You insist I am reading the charts wrong, when you don't know how I am reading them.

I do not know why you keep insisting these points when they have been proven wrong.

If you are so sure they are correct PLEASE TELL ME HOW THE SOFTWARE KNOWS MY MASK VENT RATE.

really good posting My partner and I prefer it relatively considerably hold up the superior efforts furthermore take a look at my website if you have have a probability canadian family law

Perhaps I can give a little insight, as I spent a fair wack of time hacking away at this in my software.

It looks like Encore, when configured to show the lesser Leak value, calculates an Average Leak of the last (n) days, and subtracts this from every leak value in the days reporting.

As I think it's basically reporting deviation from the norm.

I have not been able to 100% verify this yet, because I got bored stiff hacking the various statistical models trying to make things fit - I came very damn close though..

Encore maths doesn't always add up. The Calculations screen gives a bit of insight into their madness. A lot of their rounding stuff is done off .4 and not .5, which screws up calculations - I don't even bother trying to match encore calculations any more. I prefer mine to be accurate, not another messed up clone of their mistakes. (It appears the system designers are smarter than the encore developers.. Although I have seen firmware bugs in raw data too though. I seriously hope they are not one and the same. We are putting our lives in their hands!)

It could still be using a lookup table in the list of masks and just use them. (Which would be incredibly stupid, because not everyone uses Respironics masks) <-- If that's the case I'd be stopping using the other dodgy leak mode. (This can be confirmed by changing the mask in Encore Pro.. if the leak is numbers are different, this is the case.)

jedimark wrote:It could still be using a lookup table in the list of masks and just use them. (Which would be incredibly stupid, because not everyone uses Respironics masks) <-- If that's the case I'd be stopping using the other dodgy leak mode. (This can be confirmed by changing the mask in Encore Pro.. if the leak is numbers are different, this is the case.)

I used Encore Pro and changed mask prescription for last night's use to comfort full (figured that was enough difference between the nasal pillows) no change in baseline leak for me. It still starts at zero and tells me I had unintentional leak of 1.0 L/Min. Had a couple of times where pressure went up just a little and overall average was pressure was 10.6.
So even with adding a mask to the software it doesn't seem to change the reported data. For me anyway. I don't see any other place in software to choose mask.

Hmmm.. I'm glad you checked that Pugsy. I now have a little more faith in Encore's dev's not being totally useless.

I'm starting to think my original subtracted mean theory isn't far off then. I'll let you all know if I manage to work out for sure which method they used.

Even if not, it's probably not a bad idea to add this as an display option in my software, unless anyone can think of a better method to show deviations in mask leakage.