Call for Update: S9 Data Skew Issue

[-SWS]

Are there any updates to the S9 data skew investigation that people would like to share in this thread?

Those people that have a Skew issue choose to either ignore it or reformat the card every once in a while.

I have an S9 AutoSet with about 2 weeks worth of sessions and no data skew. However, these were the steps I took when I set my machine up before my first sleep session:

1) placed SD card in computer
2) used ResScan 3.10 to connect to the blank card
3) programmed all AutoSet settings onto that blank card---including machine type, pressure values, etc.
4) placed that computer-programmed SD card into my S9 unit before my 1st sleep session
(thereby allowing the computer-programmed SD card to program/initialize the S9 machine before my first use)

I don't suspect the data skew is completely a computer-side problem---if at all. I asked another S9 owner to email their unprocessed SD files. I then manually placed those unprocessed SD files onto a blank SD card and introduced the data into my ResScan. I observed the data skew on THAT person's S9 data sessions----while using the very same ResScan software & computer environment that works without data skew for my own S9 sessions.

Any more information or ideas floating around out there?

[rested gal]

For new readers who might wonder what is the S9 data skew?

"S9" refers to the new model line of ResMed's CPAP and autopap machines.
"ResScan" is the software used to download therapy data from ResMed machines.

Here are some background links:

viewtopic.php?p=464957#p464957
S9 flow skew bug... Anyone else seeing this? - topic started by DreamDiver

viewtopic.php?p=466232#p466232
S9 missing 2x Hyopopneas - in detailed view [bug?] - topic started by dave21

viewtopic.php?p=466620#p466620
ResMed S9 / ResScan lost an hours worth of sleep data - topic started by dave21

viewtopic.php?p=472503#p472503
S9 - Still Checking SKEW ??? - topic started by Nord

viewtopic.php?p=468936#p468936
S9 Skew investigation - your help needed! - POLL by dave21

[-SWS]

Many thanks for those links, rested gal. Also thanks to DreamDiver for all of his excellent S9 graphs to date, including the graph below highlighting the data skew:




Above, the 12s, 19s, and 32s central apnea flags are right-shifted from the corresponding flat-line apneas by about a minute. I have additionally noticed the bug can skew the top turquoise vertical markers with respect to the graph frames below. The turquoise markers are supposed to reference or frame the entire time-magnified graphs below (5-minute window panes in the above case).

In my above post I wondered if my S9 machine's lack of data skew had to do with my first having used ResScan to program the blank SD card----which, in turn, programmed/initialized the S9's NVRAM. Is that a better first-time S9 data initialization sequence?

[DreamOn]

I have used my new S9 Autoset for 9 nights. No skew yet. I didn't format the ResMed SD card. Just used it right out of the box as-is.

Early users had reported that their S9 software version (shown on the Clinical Info Menu on the S9's LCD screen) is SX474-0601. They have apparently updated that, because my version is SX474-0602. It'd be interesting to know what changes were made!

This is the procedure that I've been following since Day 1:

• Wake up and get coffee
• After making coffee, etc. (takes at least 10 minutes), remove card from S9 machine
• Hold down On/Off button for 3 seconds to turn machine off
• Switch off machine at the power strip
• Insert SD card into computer's built-in card reader
• Make a back-up of the card in a special folder labeled by date (just in case...)
• Open ResScan
• Download Data (rejecting duplicate data), Save, and Exit ResScan
• Eject SD card by right-clicking on drive in “My Computer”
• Replace card into the switched-off S9
• When ready for bed, turn on the machine

I'll report here if the skew does show up for me. I do check the data every day, including 5- and 10-minute detailed views, as shown in the graph posted here by -SWS. Everything has lined up perfectly, so far.

[-SWS]

I have used my new S9 Autoset for 9 nights. No skew yet. I didn't format the ResMed SD card. Just used it right out of the box as-is.

Do you happen to know which of two setup methods was initially used to program your correct pressure : 1) machine settings were initially programmed directly into the S9's LCD console, or 2) settings were initially programed onto the SD card using the ResScan software?

Early users had reported that their S9 software version (shown on the Clinical Info Menu on the S9's LCD screen) is SX474-0601. They have apparently updated that, because my version is SX474-0602.

I also have firmware version SX474-0602 without data skew. Thanks for the response.

[oscar98]

I've been wondering if there was any sort of a solution. About 2 weeks into using my S9 the skew showed up. It is now horrible. I've read the previous threads and I'm hoping there is a solution.

[DreamOn]

Do you happen to know which of two setup methods was initially used to program your correct pressure : 1) machine settings were initially programmed directly into the S9's LCD console, or 2) settings were initially programed onto the SD card using the ResScan software?

Yes, I set everything up myself. I programmed all machine settings directly into the S9's LCD console.

About 2 weeks into using my S9 the skew showed up. It is now horrible.

Oscar, which machine software version is shown on your S9's LCD screen (via Clinical Info Menu) -- SX474-0601, SX474-0602, or another?

[-SWS]

I programmed all machine settings directly into the S9's LCD console.

I'll discard that theory... Thanks!

[DreamOn]

I programmed all machine settings directly into the S9's LCD console.

I'll discard that theory... Thanks!

I've only used the machine for 9 nights, so it may just be too early for skew to show up for me (if it's going to). We'll see!

[Nord]

Are there any updates to the S9 data skew investigation that people would like to share in this thread?

Those people that have a Skew issue choose to either ignore it or reformat the card every once in a while.

I have an S9 AutoSet with about 2 weeks worth of sessions and no data skew. However, these were the steps I took when I set my machine up before my first sleep session:

1) placed SD card in computer
2) used ResScan 3.10 to connect to the blank card
3) programmed all AutoSet settings onto that blank card---including machine type, pressure values, etc.
4) placed that computer-programmed SD card into my S9 unit before my 1st sleep session
(thereby allowing the computer-programmed SD card to program/initialize the S9 machine before my first use)

I don't suspect the data skew is completely a computer-side problem---if at all. I asked another S9 owner to email their unprocessed SD files. I then manually placed those unprocessed SD files onto a blank SD card and introduced the data into my ResScan. I observed the data skew on THAT person's S9 data sessions----while using the very same ResScan software & computer environment that works without data skew for my own S9 sessions.

Any more information or ideas floating around out there?

Hi SWS, DreamOn and Oscar...

The skew "issue" begins at about 4 or 5 mask sessions after formatting and gets progressively worse.

If you set the Detailed session in ResScan to compare looking the starting point for Flow and then Pressure ( or any other data line in the detailed area)... use 10 second viewing.
It will show up beginning at 1st and second mask session with no difference then by 4th or 5th mask session will show 2 or 3 seconds difference.
By the time you get to 12th session the difference will be 15 to 20 seconds and sometimes you will be missing 1 or 2 minutes of Flow Data or your events will be displaced by 1 to 2 minutes.
I only go to about 20 masks sessions at most... then I re-format and we start over.

I have tried all variations that I can think of w/o success at changing the skew.

ResMed have not been helpful to solutions yet. Customer Service agreed with the problem then engineers denied there is any problem. I am still in discussion with them but...

BTW: SWS and DreamOn... perhaps this new firmware has a solution. Let me know if you see the same skew with 10 second viewing...
I'm guessing you still have it Oscar...

Nord

[-SWS]

The skew "issue" begins at about 4 or 5 mask sessions after formatting and gets progressively worse.

Nord, by "mask sessions" do you refer to mask-fit sessions? If so, that might explain why I see no skew. I only ran the mask-fit session once during initial machine setup. I don't bother with that part of the routine.

perhaps this new firmware has a solution. Let me know if you see the same skew with 10 second viewing...

No skew in 10-second view... But I only have 13 nights worth of data---and so far I don't run mask-fit sessions. Thanks for the input, Nord.

Can others pin the data skew down to mask-fit sessions? If so, can you experimentally avoid the skew by refraining from mask-fit sessions?

[DreamDiver]

I've backed off the subject here because there seems to be a 'Resmed-can-do-no-wrong' contingent on this forum that those of us with the skew issue just got tired of fighting. It isn't worth the energy. I've tried just about every method, including having the software set up the card. I've tried four different cards - all directly supplied by ResMed - all with the same result.

I suspect that either shortcuts were made, using substitute or substandard integrated circuits or chips to meet a deadline, or they simply made a batch with a firmware skew, regardless of what ostensible version the firmware says it is on the machine. After showing conclusively that skew exists because of a firmware or hardware issue, ResMed has basically decided not to communicate with me any longer.

I spent thousands of dollars on two entirely inconclusive sleep studies because the S9 says I have nothing but centrals. My baseline study suggested I have OSA. My titration didn't have a single apnea to say one way or another because they used a crappy Respironics machine that kept waking me up with their <insert-your-favorite-potty-mouth-words-here> 'clear airway apnea' detection system. So, I could have CompSAS, but the only thing they know for sure is that I need some sort of xPAP.

My doctor admits he is a mere clinician and that he cannot help me figure out exactly whether I have centrals or not. He relies exclusively on what his techs write up and signs it. Frankly, that's no way to run a doctor's office. I contest the results of both studies, because the lab techs didn't even know their own omnilab machines were administering 2cm pulses, and waking their paitents. I'm too tired to fight, and I'm out of money.

That's pretty much why I haven't been around lately.

I continue having to start with a freshly initialized card every few days to remove the skew.

At this point, I would like for ResMed to step in and explain why their central apnea detection really is any better than the lab method, and I'd like for them to replace my machine with one that actually doesn't skew. I'm also angry that months of data showing nothing but centrals doesn't trump two inconclusive lab studies.

Every time I go to a doctor - any doctor - I trust them less. Today's MD more about marketing than medicine. Either the doctor is taking kickbacks from big pharma, or they sell European remedies out of their own offices that aren't covered by insurance and cost out the nose. It's all about sucking your wallet dry and disappearing when your money is gone.

This whole thing has left a bitter after-taste and a hole in my pocket.

[-SWS]

DD, can you find any correlation between your machine's data skew and number of mask-fit sessions?

Sorry to hear about your frustrating experience with sleep studies.

[Nord]

The skew "issue" begins at about 4 or 5 mask sessions after formatting and gets progressively worse.

Nord, if mask sessions refer to mask-fit sessions, then that might explain why I see no skew. I only ran the mask-fit session once during initial machine setup. I don't bother with that part of the routine.

perhaps this new firmware has a solution. Let me know if you see the same skew with 10 second viewing...

No skew in 10-second view. But I only have 13 nights worth of data---and so far don't run mask-fit sessions. Thanks for the input, Nord.

Hi SWS...

No I'm using my own terminology for "mask sessions" .
Since I have separate naps or I will get up during the night and shut off CPAP... then go back to bed an hour later.
Those are all masks sessions of which I may have between 1 and 7 during any 24 hour reporting period.
Since each of these sessions are designated differently and use a separate container by the S9... each adds to the skew factor.

I have not found that the mask-fit adds or changes the skew since it is 3 minutes long and can run continuously through to sleep.

I use the 10 second viewing to see the skew by the second markers as it is easier to see exact numbers.

BTW: OT How do you track your Slow Wave Sleep differences or do you do it by "feel" ??

Nord

[-SWS]

I have not found that the mask-fit adds or changes the skew since it is 3 minutes long and can run continuously through to sleep.

Okay, thanks for the clarification.

BTW: OT How do you track your Slow Wave Sleep differences or do you do it by "feel" ??

I don't think there's any way to track Slow Wave Sleep on an ongoing basis at home---short of expensive home-based PSG diagnostic equipment with EEG channels. So I don't really track mine at home.