CPAPtalk.com

Physician wrote: . . . several certified sleep docs . . . agreed ResMed should add MEAN LEAK to the LCD Sleep Report . . . .

The screen does provide weekly, monthly, six-month, and yearly medians of the nightly leak 95th centiles (which I find very valuable), as explained in the clinical manual for the Auto 25:

"A leak rate of more than 24 L/min (0.4 L/s) affects the accuracy of other measurements and is associated with patient discomfort, disturbed sleep, and reduced efficacy of treatment. The leak reported in the Efficacy Data submenu is the 95th centile value for mask-on time for each session, or the median 95th centile value for a selected time interval (last week, last month, last six months, or last year)."

(I know you know that. And that wasn't the point you were making. That was for the gallery.)

I have seen leak charts where the software-collected median for the night looked great, but the chart looked like an 8.0 on the Richter in Arousal City. The 95th centile value seemed to me to be the most useful number (if the choices are maximum, 95th centile, or median, that is) that I would have personally chosen to put on the screen for a heads-up to the patient or clinician in those instances. Maximum would be falsely alarming. Median could be falsely comforting. 95th is juuuuuuust right, IMO, for saying "Dude, hook me up to check the charts, would ya?!"

But what do I know? I still think ASVs should include effort belts.

jnk wrote: I still think ASVs should include effort belts.

I think ASVs should include oximeters. But who cares what we think?

OutaSync wrote: . . . who cares what we think?

We do!

Although I admit that some days I don't even care what I think.

ozij wrote:

Physician wrote:Speaking of leaks, just downloaded all my S9 data into ResScan. In the past week when that 7.2 leak repeatedly appeared every night and created concern, the printout shows a MEDIAN LEAK of 1.2 L/min. and a MAX of 120 L/min. and AHI = 1.0

I wonder if you were trying to check the data immediately after turning the machine off. That processor has quit a bit of data reading and caltulating to do immediately after the machine is turned off, it needs time. I saw at 7.2 leak rate one time -- when I pressed the review buttons immdieately.

As for not "needing" graphs: I disagree. A statistic of central tendency -- be it average, median or mode -- will never give you the information you need for troubleshooting. Only a graph shows you what happened when.

If you wish to disagree with five sleep experts, that is your decision. But I repeat: We should NOT have to rely on ResScan to get a good assessment of progress, efficacy, and problems. That 95th %ile by itself is misleading and incomplete.

I think sleep docs often leave the setting up of home machines, the collection of data from home machines, and much of the interpretation of the data from home-machines to the RTs, if the existence of home-machine data even crosses their minds. At least, although I consider my sleep doc to be an excellent one, I don't think he would know my machine from a kitchen blender. He certainly isn't versed on the differences between brands.

My understanding is that it is patients who have learned the importance of leak data on their own, based on observations of their own sleep. One day the sleep docs may catch up on that stuff. But I wouldn't put any money down on that notion.

I could be very wrong, though.

Has anyone out there ever had his or her sleep doc initiate a conversation about (1) the serious problems with mouth leak or (2) the importance of leak data?

I don't know how a doctor could even have that conversation, since, sadly, he or she often has no idea whether the patient is going to receive a machine that gives leak data!

Asking a "sleep expert" about home-machine data can sometimes be like asking the CEO how many scoops of coffee to put in the coffeemaker. He/she would likely have no idea. He/She has people who do that sort of thing for him/her.

jnk wrote:

Physician wrote: . . . several certified sleep docs . . . agreed ResMed should add MEAN LEAK to the LCD Sleep Report . . . .

I have seen leak charts where the software-collected median for the night looked great, but the chart looked like an 8.0 on the Richter in Arousal City. The 95th centile value seemed to me to be the most useful number (if the choices are maximum, 95th centile, or median, that is) that I would have personally chosen to put on the screen for a heads-up to the patient or clinician in those instances. Maximum would be falsely alarming. Median could be falsely comforting. 95th is juuuuuuust right, IMO, for saying "Dude, hook me up to check the charts, would ya?!"

95% can be falsely alarming OR falsely comforting.

If the 95% is, say, 15L/min, it's easy to assume that since that's well below the Red Line, you couldn't possibly have a leak problem even though you were leaking at 13L/min all night long. I personally think leaking all night long even at something that's only half the Red Line value is likely to affect therapy or patient comfort or both. In that case, the 95% Leak Rate is falsely comforting.

And many newbies look at that 95% number when it's say 25 L/min and conclude (falsely) that they've got a major leak problem going on without ever realizing until they look at the Leak Rate graph that 90+ percent of the time, they have no detectable leak and that one 25 minute leak accounts for the alarming 95% leak rate figure. In this case, the 95% leak rate is falsely alarming.

And the fact that the Leak Rate number on the LCD is not labeled as the 95% number also leads to confusion in my opinion---particularly on the part of newbies who are very worried about whether they have leak issues: I was one of those worried newbies not too long ago. But because I had other issues, I decided to put "leaks" on the back burner for a while. Then once I started looking at the graphs, I realized I really didn't need to worry about leaks since my leak rate is typically 0.0 for well over 80% of the time even on a "bad" night where I had two or three 15 minute detectable leaks at most.

Me personally, I think the only way to provide meaningful data on the LCD is to provide and label both the 95% leak rate and either the mean or median leak rate. It's only by looking at the two numbers together that you can determine (1) if you are having short-term, but really big leaks and (2) if you are having middle-sized leaks for most of the night. In either case, I think you'll be more comfortable and get better therapy if the leak issues are addressed instead of ignored.

I also think that since ResMed will not or cannot make the software directly accessible to all S9 users, they have some responsibility to make sure the numbers on the LCD actually provide meaningful information. In my case, I know the 95% leak rate number is largely meaningless because it seldom reflects the reality of what really happened in the night as far as my leaks are concerned.

robysue wrote:

jnk wrote:

Physician wrote: . . . several certified sleep docs . . . agreed ResMed should add MEAN LEAK to the LCD Sleep Report . . . .

I have seen leak charts where the software-collected median for the night looked great, but the chart looked like an 8.0 on the Richter in Arousal City. The 95th centile value seemed to me to be the most useful number (if the choices are maximum, 95th centile, or median, that is) that I would have personally chosen to put on the screen for a heads-up to the patient or clinician in those instances. Maximum would be falsely alarming. Median could be falsely comforting. 95th is juuuuuuust right, IMO, for saying "Dude, hook me up to check the charts, would ya?!"

95% can be falsely alarming OR falsely comforting.

If the 95% is, say, 15L/min, it's easy to assume that since that's well below the Red Line, you couldn't possibly have a leak problem even though you were leaking at 13L/min all night long. I personally think leaking all night long even at something that's only half the Red Line value is likely to affect therapy or patient comfort or both. In that case, the 95% Leak Rate is falsely comforting.

And many newbies look at that 95% number when it's say 25 L/min and conclude (falsely) that they've got a major leak problem going on without ever realizing until they look at the Leak Rate graph that 90+ percent of the time, they have no detectable leak and that one 25 minute leak accounts for the alarming 95% leak rate figure. In this case, the 95% leak rate is falsely alarming.

And the fact that the Leak Rate number on the LCD is not labeled as the 95% number also leads to confusion in my opinion---particularly on the part of newbies who are very worried about whether they have leak issues: I was one of those worried newbies not too long ago. But because I had other issues, I decided to put "leaks" on the back burner for a while. Then once I started looking at the graphs, I realized I really didn't need to worry about leaks since my leak rate is typically 0.0 for well over 80% of the time even on a "bad" night where I had two or three 15 minute detectable leaks at most.

Me personally, I think the only way to provide meaningful data on the LCD is to provide and label both the 95% leak rate and either the mean or median leak rate. It's only by looking at the two numbers together that you can determine (1) if you are having short-term, but really big leaks and (2) if you are having middle-sized leaks for most of the night. In either case, I think you'll be more comfortable and get better therapy if the leak issues are addressed instead of ignored.

I also think that since ResMed will not or cannot make the software directly accessible to all S9 users, they have some responsibility to make sure the numbers on the LCD actually provide meaningful information. In my case, I know the 95% leak rate number is largely meaningless because it seldom reflects the reality of what really happened in the night as far as my leaks are concerned.

Good points. Well said. Thanks.

I do find it ironic, though, that ResMed often gets critiqued (by me too) on which nightly data it provides, when it is the company in the forefront of providing good data easily accessed from the screen. They deserve a lot of credit for that. Who else provides a menu of data on the machine that is specifically for the patient, a menu in which the patient need not worry about changing any settings?

My fear is that one day the only info from the screen will be smiley faces and frowny faces!

jnk wrote:

Physician wrote: . . . several certified sleep docs . . . agreed ResMed should add MEAN LEAK to the LCD Sleep Report . . . .

The screen does provide weekly, monthly, six-month, and yearly medians of the nightly leak 95th centiles (which I find very valuable), as explained in the clinical manual for the Auto 25:

"A leak rate of more than 24 L/min (0.4 L/s) affects the accuracy of other measurements and is associated with patient discomfort, disturbed sleep, and reduced efficacy of treatment. The leak reported in the Efficacy Data submenu is the 95th centile value for mask-on time for each session, or the median 95th centile value for a selected time interval (last week, last month, last six months, or last year)."

(I know you know that. And that wasn't the point you were making. That was for the gallery.)

I have seen leak charts where the software-collected median for the night looked great, but the chart looked like an 8.0 on the Richter in Arousal City. The 95th centile value seemed to me to be the most useful number (if the choices are maximum, 95th centile, or median, that is) that I would have personally chosen to put on the screen for a heads-up to the patient or clinician in those instances. Maximum would be falsely alarming. Median could be falsely comforting. 95th is juuuuuuust right, IMO, for saying "Dude, hook me up to check the charts, would ya?!"

But what do I know? I still think ASVs should include effort belts.

I'm not suggesting eliminating the 95th%ile value from the LCD screen report, but ADDING the mean or average value to the LCD report. Simple !

Physician wrote:

jnk wrote:

Physician wrote: . . . several certified sleep docs . . . agreed ResMed should add MEAN LEAK to the LCD Sleep Report . . . .

The screen does provide weekly, monthly, six-month, and yearly medians of the nightly leak 95th centiles (which I find very valuable), as explained in the clinical manual for the Auto 25:

"A leak rate of more than 24 L/min (0.4 L/s) affects the accuracy of other measurements and is associated with patient discomfort, disturbed sleep, and reduced efficacy of treatment. The leak reported in the Efficacy Data submenu is the 95th centile value for mask-on time for each session, or the median 95th centile value for a selected time interval (last week, last month, last six months, or last year)."

(I know you know that. And that wasn't the point you were making. That was for the gallery.)

I have seen leak charts where the software-collected median for the night looked great, but the chart looked like an 8.0 on the Richter in Arousal City. The 95th centile value seemed to me to be the most useful number (if the choices are maximum, 95th centile, or median, that is) that I would have personally chosen to put on the screen for a heads-up to the patient or clinician in those instances. Maximum would be falsely alarming. Median could be falsely comforting. 95th is juuuuuuust right, IMO, for saying "Dude, hook me up to check the charts, would ya?!"

But what do I know? I still think ASVs should include effort belts.

I'm not suggesting eliminating the 95th%ile value from the LCD screen report, but ADDING the mean or average value to the LCD report. Simple !

Fair enough.

But if they're gonna do that, why not just have them build ResScan into every machine and have a 2" full-color lcd screen show the full leak, pressure, and event charts every morning? After all, why should my $1,000 PAP machine be dumber than my $200 unlocked smartphone? Let's go all the way with it!

jnk wrote:

Physician wrote:
Fair enough.

But if they're gonna do that, why not just have them build ResScan into every machine and have a 2" full-color lcd screen show the full leak, pressure, and event charts every morning? After all, why should my $1,000 PAP machine be dumber than my $200 unlocked smartphone? Let's go all the way with it!

Okay. Will arrange that, but give me 48 hours.