Let's play a game and maybe learn something while playing

[Pugsy]

So...here's the game.

Tell me if you think I was asleep or awake when the grouping of OAs got flagged at the beginning of the night.
Tell me why you think whatever it is that you think.
I included the main overview as well as a zoomed in section.

[Okie bipap]

The first two minutes you appear to be awake as indicated by the irregular breathing pattern. The next two minutes, the breathing shows a very steady rhythm which would be consistent with sleeping. If you can fall asleep within two minutes of putting the mask on, you are doing much better than I do. It usually takes me around five to ten minutes to drift off on a good night. Many nights it takes longer.

[Pugsy]

Hint....everyone please note the break in therapy where I turned the machine off and right back on at 22:20:30.

[Pugsy]

Answer at 9 PM CST. Give you all a little more time to take the quiz.

Don't be shy....

another hint....the machine is easily fooled and it will likely surprise you how much it got fooled.

[Pugsy]

So....I wasn't asleep at all during that first session I zoomed in on. Totally awake and no where near asleep. I am 100% certain I never fell asleep.
I know it looks like asleep breathing but that was just me making a concerted effort to breathe nice and slow and even.
The OAs were my attempt to see what the machine would flag if I held my breath. I held my breath for 15 to 25 seconds...I counted under my breath.
I thought it might flag centrals but it didn't. It does seem like I see a couple of the flagged events with the FOT thing going on. Don't ask me why....I dunno.

These machines can be fooled and fairly easily.

So....false positives aren't limited to just central apnea flagging....here's proof that it can and will flag OAs as well,
I have also seen false positive hyponeas as well but they are rare.

Note the pressure increase in response to those flagged OAs and notice it got quite high (for me) but the pressure increases didn't stop the flagged OAs at all...and since I was awake and holding my breath I wouldn't expect it to help out with those false positive OAs.
Funny thing though...when the pressure was going up I didn't notice it increasing. I was a bit surprised that I couldn't tell any change in the pressure.

The break in therapy at 00:33 was me getting up because Toby had got in the bathroom and accidentally shut the door and couldn't get out. Last time he did that he destroyed the door trim and made an awful mess so I didn't dilly dally around when I heard him scratching. I immediately got up and let him out and went right back to bed and from the looks of things I think I fell back asleep fairly quickly. Obviously if my known awake breathing is so regular it looks like I am sleeping (that first session with the false positives I know for a fact I was awake) it isn't always so easy to spot the transition to sleep.

Moral of this story....these machines are good but they aren't perfect and they aren't always right. I have known this for a long time and why I don't get all worked up when I spot the machine doing something that seems a bit off.
It's why I don't go changing things willy nilly based on some weird outlier results.

Side note...the machine shows only 2 sessions for the night and it should show 3. The brief 8 minutes first session somehow got included in the first session as it shows a little over 2 hours but doesn't show me a session of just 8 minutes.
Not sure why....unless maybe my turn off and right back on was too rapid. Tonight if I am awake enough I plan to test that part again.

Some false positive flagging is very obvious but as you can see in my case here....it's not always so easy to spot.
Don't go telling me I was asleep....I wasn't. I was awake and holding my breath for each of those 4 OAs that look so very real.

[galeforcewinds]

Pugsy--

Newbie here with two questions regarding the flagging of CAs and OAs:

1) Do OAs ever get misflagged as CAs? I've seen what I thought were CAs being flagged as OAs in my own flow charts, but never the reverse (but then I don't always look at every CA). Moreover, I haven't seen much commentary on this possibility. Is it unlikely that an OA would be misflagged as a CA?

2) I understand that REM breathing can also be erratic and variable. Most of my CAs seem to be post-arousal CAs judging from the erratic breathing before they occur, but could I also be looking at a CA during REM?

The reason I ask is that, right now, my CAs outnumber OAs by a fair amount. I have the OAs down to what I would consider a quite acceptable number--at most 1 or 2 a night. And based on my reading, I've decided not to fuss too much about my CAs because they seem to be primarily post-arousal and could be the result of being relatively new to CPAP (just finished my first month.) Plus my oxygen monitor is showing few, if any, desaturations. I just want to make sure that I am not having more OAs than I think.

Thanks.

[Pugsy]

1) Do OAs ever get misflagged as CAs? I've seen what I thought were CAs being flagged as OAs in my own flow charts, but never the reverse (but then I don't always look at every CA). Moreover, I haven't seen much commentary on this possibility. Is it unlikely that an OA would be misflagged as a CA?

You know I don't have a clear cut answer for those questions. Never thought about the reverse happening and some CA/centrals being mislabeled as an OA. I don't think we have the level of detail data needed available to be able to manually tell the difference between the 2 types of flagged events because we don't have any way to measure effort to breathe.
The machine only measures air flow....not effort...all it knows is we weren't breathing.
My gut says that the chances of reversal mislabeling like you are asking about....unlikely but I have no proof.
I once asked a sleep tech if a ResMed called an event as being a central was it likely a central and not OA ....his response was "yes". I didn't ask the converse....didn't think about it.


Why the worry about centrals?

Did you know that they are normal and expected and they are only a problem if they cause desats or if sleep onset centrals keep a person from falling asleep or if there is a huge number of them?
Did you know that it is normal to wake up after the completion of a REM sleep stage? Google "sleep stages" and look at the hypnograms. We may or may not remember those awakenings but they happen it causes the chance for a false positive post arousal apnea event flag.

2) I understand that REM breathing can also be erratic and variable. Most of my CAs seem to be post-arousal CAs judging from the erratic breathing before they occur, but could I also be looking at a CA during REM?

During REM??? unlikely.
Now after REM chances are you see a central that is post the normal to have arousal after REM is over with.

[Pugsy]

So....remember my side note about the sessions not matching up like they should???

I never paid attention until now so I don't know how long this has been going on and don't have the time or energy to go back and see if I spot anything.

Last night it shows one sleep session yet I have a noticeable break in therapy about 1 minute into therapy and for a little over a minute...and I don't remember what happened.
Then when I woke up early this morning I turned the machine off twice for about 15 seconds...and the break in therapy didn't show up at all for those 2 times.

I should have more than one sleep session reported....

So....I don't know what to think. Never had this sort of issue with any of my past machines.
I unplugged the machine for a few minutes forcing a reboot once I plugged it back in.
Didn't know what else to try.
This machine has close to 12K hours on it. Silent and all that and everything else seems to function properly.

I will see what shows up tomorrow but it looks like I need to break out my new machine sooner than later.
I need to use it anyway because it has that 2 year warranty I don't want to waste.

Any other ideas????

[loggerhead12]

Any other ideas????

Maybe the big software update they pushed is ignoring breaks under xx seconds. I've seen quite a few people post that they stop and restart their machines if they wake up during the night to kick off the ramp again.

[Pugsy]

Any other ideas????

Maybe the big software update they pushed is ignoring breaks under xx seconds. I've seen quite a few people post that they stop and restart their machines if they wake up during the night to kick off the ramp again.

Well the reboot didn't help.

I am in airplane mode and have been for years. I don't think I ever got any sort of software update anyway.
Something in OSCAR somewhere???? I no longer have ResCan available to double check results.

I had 3 breaks in therapy and only the one at the end of my night shows up in the statistics.

This break in therapy shows as a session break in the statistics. It is close to 2 minutes in duration.


But these 2 breaks in therapy don't show up in the first session at all. They are more along the lines of a minute in duration. Maybe I need to wait longer but that is kinda hard to do sometimes.

Got some stuff to do today. Will think on this a bit more later this evening.

[kteague]

Thanks for the discussion. Since I'm nearly clueless about the data I love learning from those in the know.

[Pugsy]

Soooo.....last night I was doing some thinking and I came up with this idea about my missing sessions from the statistics.

I was thinking I never noticed this before but then again I have to admit that I never really looked.
Doing the turn off and right back on thing gave me what I was looking for and that was a break in the therapy so I could better evaluate the flow rate/breathing when I zoomed in on that part of the flow rate graph. So I got what I wanted and it didn't really matter how many machine "sessions" the machine reported....so I just never noticed because I wasn't looking.

It appears from my limited experiments that I need somewhere between 1 minute (where I don't get a separate session marked) and 2 minutes (where the session does show up) for there to be a separate session. Probably has to do with data sampling time frame. I rarely have more than one sleep session recorded anyway ....unless I am doing an experiment like I was doing that caused me to start this topic.

I don't think anything is wrong with my machine and even if there was...nothing I can do about it anyway.
For all I know the machine is working exactly like it is supposed to work within the software of the machine itself.

I am going to shrug my shoulders and just move on and not worry about it.

[lazarus]

Do OAs ever get misflagged as CAs?

To quote my friend Monk, "here's the thing":

My understanding is that for some patients with OSA, most all pauses in breathing, even natural ones, quickly result in a closed, or mostly closed, airway, and those can get labeled as obstructive by a home machine. In an NPSG, those either get called centrals or are ignored as non-events, since the effort belts give useful added info in that setting. Of course, if the airway remains closed, a true obstructive event results once the struggle to breathe sets in. But in those situations, the pause caused, or at least facilitated, the obstruction.

And when an OSA patient also has a significant central aspect as part of their sleep issues, then an obstruction, even a very brief one, or a relatively minor narrowing of the airway, may not meet the criteria for getting labeled an obstructive apnea by a home machine but may nevertheless cause significant instability in breathing that then leads quickly to open-airway pauses in breathing that then get labeled centrals until the breathing settles into a more normal pattern. That is why many sleep researchers consider a patient's having centrals in a sleep study to be as much an indication of likely OSA rather than CSA. Simple CPAP can often lead to fixing both kinds of events, given time.

It is extremely useful that home machines differentiate between open-airway and closed-airway "events," however the accuracy cannot be considered perfect and the significance of the seeming patterns may differ from patient to patient. Therefore, I consider home-machine labeling of events to be useful hints instead of being definitive pronouncements.

Similarly, I believe that SWJ (sleep/wake junk) can occur when someone is mostly awake or is mostly asleep in a very bumpy stage one of sleep, which as a stage is mostly a transition state from wake to sleep anyway, depending on how you look at it. Subjectively, a patient may perceive those periods as either wake or sleep, which is one of the reasons why self-estimates of the time asleep are so inherently unreliable.

I am truly lousy at deciphering squiggly lines, myself, though. That is just one of my many limitations. I appreciate and admire those with a talented eye for that.

[Ralph437]

I have also seen false positive hyponeas as well but they are rare.

Pugsy, would this be an example of a false positive hypopnea?
The event that occurred just prior to the hyponea segment was a single cough.

[Pugsy]

I have also seen false positive hyponeas as well but they are rare.

Pugsy, would this be an example of a false positive hypopnea?
The event that occurred just prior to the hyponea segment was a single cough.

I have no idea.

Please don't go putting someone else's questions in this thread. It's not even your own data....it belongs to someone else's wife.

I am going to lock this thread because I already asked you not to post in my thread and you still want to ignore my request.