Need help understanding my Sleepyhead Stats

[bwexler]

I just read much of RobySue's blog and the definitions in the help section of Sleepyhead.
I am now smarter than I was 2 hours ago.
I still have questions.

Here is a snapshot of my Sleepyhead Stats:


SleepyHead v0.9.2

This is a beta software and some functionality may not work as intended yet.
Please report any bugs you find to SleepyHead's SourceForge page.
CPAP Statistics as of Thursday, February 16, 2012
437 days of CPAP Data, between 12/5/2010 and 2/16/2012
Details Most Recent Last 7 Days Last 30 Days Last 6 months Last Year
AHI 0.65 1.02 0.79 0.97 0.97
Hours per Night 07:39 07:51 08:01 15:36 13:57
Average Pressure 7.75 5.95 6.51 2.47 1.16
95% Pressure 10.20 9.68 10.32 10.50 10.40
Average Leaks 17.97 11.38 8.45 2.28 1.07
95% Leaks 32.40 32.40 32.40 32.40 32.40

Changes to Prescription Settings
First Last Days AHI Machine Mode Pr. Rel. Min Pres. Max Pres.
11/24/2011 2/16/2012 84 0.91 ResMed Auto EPR x3 5.00 20.00
11/11/2011 11/23/2011 13 2.09 ResMed Auto EPR x3 6.00 20.00
12/5/2010 11/10/2011 340 0.94 ResMed Auto EPR x3 4.00 20.00
The above has a threshold which excludes day counts less than 5 from the best/worst highlighting

Machine Information
Brand Model Serial First Use Last Use
ResMed S9 AutoSet 22101691401 12/5/2010 2/16/2012

The formatting did not copy well, but hopefully you experts can decipher it.
Most Recent Last 7 Days Last 30 Days info seems correct, but

Last 6 months Last Year data does NOT seem correct.
I never used my CPAP 13+ hours per night and my minimum pressure was never below 4

[robysue]

bwexler,

First, let me clean up that data table to make it easier for folks to understand.

Code: Select all

This is a beta software and some functionality may not work as intended yet.
Please report any bugs you find to SleepyHead's SourceForge page.

CPAP Statistics as of Thursday, February 16, 2012
437 days of CPAP Data, between 12/5/2010 and 2/16/2012

Details           Most Recent      Last 7 Days	       Last 30 Days     Last 6 months    Last Year
AHI                  0.65             1.02                  0.79             0.97            0.97
Hours per Night     07:39            07:51                 08:01            15:36           13:57
Average Pressure     7.75             5.95                  6.51             2.47           1.16
95% Pressure        10.20             9.68                 10.32            10.50           10.40
Average Leaks       17.97            11.38                  8.45             2.28            1.07
95% Leaks           32.40            32.40                 32.40            32.40           32.40

The numbers you highlighted in blue---the Last 6 Months and Last Year figures for Hours per Night and Average Pressure are software glitches and you should report them to JediMark through the Sleepy Head page at SourceForge. Remember SleepyHead is still in beta testing and bugs are still around. Why your particular numbers for these statistics are so obviously wrong is not something I can ask. But since SleepyHead is still in beta testing, you have to look at whether the numbers make sense. If they do, then SH is usually pretty close to the "right" number. If not, report the bug so that it will eventually get fixed.

My guess is that the AHI numbers are all pretty accurate since they're consistent and most folks AHI does settle down. Long term, it looks like your machine is recording about one event per hour of run time. That's excellent for the How well is my machine controlling the apnea? question. Seems like your 95% pressure is pretty much around 10.5 cm. Since you are still running with a pretty wide open APAP range of 5-20 cm, I think you could tighten that range if you wanted to---there's no good reason for leaving the top setting at 20cm.

Either the Last 6 months and Last Year Average Leak data is inaccurate OR your leaks are growing worse with time. They're not yet alarmingly large in the sense of the average being above 24 L/min (ResMed's Large Leak definition), but they are large enough that when I look at them in context of the 95% Leak Rates, I have to wonder: Are you sometimes seeing Mr. Red Frowny Face in the morning on the short form of the LCD's Sleep Quality Report?

The 95% Leak numbers also look suspicious: I'd expect some variation after the decimal place. If this were my data, I'd do a quick step through of lots of individual days of data and see what the individual 95% Leak Rates look like. If they are often above 30 L/min, then these figures might be a coincidence. But if you see lots of days in a row with lower 95% Leak Rates, then I'd guess that there's a bug in how SH is computing these numbers too.

So I recommend wandering over to the SourceForge site and reporting these summary statistics as likely being incorrect as the result of some kind of a bug.

For what it's worth, I've found that SH does a great job (>95% of the time) with my PR System One Bipap's data. But there seem to be some serious problems with the roughly 90 days of S9 AutoSet data that I have from Fall 2010. At some point I need to put together a bug report for JediMark, but since that data is so old, it's no longer particularly relevant to me. And I've been too busy to sit down and do the bug report.

[bwexler]

Thank you RobySue.
That helps.
I just changed masks to the Sleepweaver about a week ago and experienced many leaks.
I just received the new flexible 18" hose extension yesterday.
It felt like the leak was better but I still had the frowny face this morning.
I feel like I can begin to work with the new mask now and learn to reduce the leaks.

[tekitte]

I just started using Sleepyhead. I notice that the screen with the little pie chart on it is saying that I slept only a few minutes and was awake the majority of the time. Does anyone know whether this is a bug?

Terri in CA

[DavidT]

The wake number is the time you stopped using the machine, I believe...

[robysue]

The wake number is the time you stopped using the machine, I believe...

Yep, that's correct.