Obstructive event lasting 3¾ minutes?

According to Sleepy Head, I experience an even lasting approximately 3¾ minutes (224 seconds). According to the data, I continued on after than as if nothing had happened. I really find it hard to believe that this could have happened and I would not have been aware of it. MY AHI is generally below one (it was .84 last night), and I have never seem an obstructive event nearly so long. Also, I do not feel anything out of the ordinary.

Can the CPAP or the Sleepyhead software be wrong? Should I be concerned?

Can you get an image zoomed in on the prolonged event....just click on it and the graphs will zoom in nicely?

While the number in parentheses isn't exactly seconds on the Respironics machines (it is on ResMeds).... it isn't usually off by that much. Usually maybe 2 or 3 seconds.
Let's look at it up close.

Like you I would have a hard time believing that one myself. Unless you are in training to go dive for pearls.

This might be one of the time we say the aliens payed you a little visit last night and messed with you. That's who I blame for anything I can't logically put the blame on.

D.H. wrote:Can the CPAP or the Sleepyhead software be wrong? Should I be concerned?

Yes, an xPAP can misscore things. It can also have trouble tracking breathing in certain circumstances. And SH does still have some bugs that appear to affect the DreamStations. (Or perhaps it's Encore Pro that has the bugs.)

Both the xPAPs we use and SH are the work of human beings. As such, they're not perfect.

But should you be concerned? The data from our xPAPs is reasonably accurate, particularly for trending data over the course of many days and weeks. If a user's machine scored AHI is usually less than 1.0 over many days, weeks, and months, it's reasonable to assume the xPAP is doing an outstanding job of managing the apnea. If a user's machine scored AHI is often above a 5.0 over many days, weeks, and months, it's reasonable to assume that the xPAP is NOT doing its job of properly managing the apnea, and most likely the machine's settings need to be tweaked in some fashion.

xxyzx wrote:why didnt the cpap fix it
is your pressure fixed and too low

Regular old APAPs do NOT increase the pressure in the middle of an apnea in an attempt to "trigger" an inhalation. (ASV machines do increase the IPAP pressure when they think you may be in the middle of a central apnea, but that's a whole different can of worms.)

A regular old APAP is NOT a ventilator in the usual sense of the word. The pressure it provides helps keep the airway from collapsing, but once the airway is collapsed, additional pressure is not going magically push the collapsed airway open. The point of APAP/CPAP therapy is to keep the airway unblocked in the first place.

A regular old APAP may increase the pressure after an event is over. How much the pressure increase will be depends on whether the machine is already running at max pressure, the manufacturer's auto algorithm, and whether the event is "isolated" or not. In general, PR APAPS increase the pressure after the end of an event only if there are at least two obstructive events (OAs and Hs) that have occurred within the last 2 minutes. The PR machine will then increase the pressure by 1cm and it will wait for another minute or two before it increases the pressure again. Resmed machines are somewhat likely to increase the pressure even for isolated events, and when clusters of events are occurring, the Resmed is likely to increase the pressure by more than 1cm, and perhaps a lot more than 1cm.

xxyzx wrote:it looks like you had apnea followed by hypopnea that cleared when you snored
maybe sleepy head has problems distinguishing them when they are run together

What is this hypothesis based on? The only flow rate data provided does not show a close up of the 3.75 minute long event. And in the event table, there's an FL scored at roughly 2:20, a VS2 (snore) scored just before 2:24:53, and the OA scored at 2:24:53.

And it's important to understand that SH does not decide where the events are scored or what kinds they are; all SH does is display the data that the machine actually recorded.

Here's a zoom in on the time of the event (which is approximately 2:24 AM).

Not zoomed in enough. Do like I asked and click on the actual event in the events tab and then the graphs will be zoomed in like we need and the event in question will be right in the middle.

Here is a zoomed as Sleepy Head will go.

From what I'm seeing, it wasn't even an OA. The most I would have thought was after the deep breathing, the slower shallower section may have been tagged as a hypopnea

I'm happy to call it a fizzer.

No, if you will do as I ask the zooming in will show up in the optimal size to evaluate the event.
Got to the events tab
Click on the category
and then click on that one event.
Then do the screen shot.

Here's what you asked for.

D.H. wrote:Here's what you asked for.

Set your "view size" (left side, bottom) to something like 5 minutes. It's set at 30 minutes, now.
Looking at your Flow Rate in the zoomed image a couple above, there is no apnea at 02:24. However, there is a reduction in average Flow back to normal from the Flow just earlier. Perhaps that is what is triggering this Flag.

I don't have stats on my O2 level.

I have nine recorded snoring incidents, all for one second.

The view size is what is screwing it up. It is compressing 30 minutes and I like 4 or 5 minutes. It's cramming too much stuff together when you select just the one event. I the one really zoomed in that you did was too much zoomed in. At any rate doesn't look real to me. I vote for alien abduction.
So that the flow rate looks like this.

Here is the same posting with a five minute display window as requested.

D.H.

My guess is that what you are looking at is a mis-scored piece of sleep transition breathing, and hence of no real concern.

I base that on analyzing this shot, https://web.archive.org/web/20170720202 ... fUVzLr.jpg, which shows a one-minute interval centered on the "event" and this shot, https://web.archive.org/web/20170720201 ... M1Jmgr.jpg, which shows a ten-minute interval centered on the "event".

I apologize in advance for the technical mathy stuff that follows, but here is an explanation of why I think this OA is a false event:

First, notice the very large breaths in the 10-minute snippet at 2:23:15-2:24:00. That is probably a spontaneous arousal. Also notice that the peaks of the inhalations in the arousal are at +65 L/min and the lowest points on the exhalations during the arousal are at -130L/min. Those numbers are driving SleepyHead's choice of the vertical scale on the Flow Rate graph: The vertical distance goes from -130 L/min to +130 L/min, which is a pretty large range for a Flow Rate graph.

Next notice what looks like normal sleep breathing before the arousal in the 10-minute screen shot: Those inhalations top out somewhere around 30-35 L/min since they're about half-way up to the +65 L/min line. So it's a good bet that your normal sleep breathing inhalations often top out at around 30 L/min. Note that since the vertical axis goes all the way up to +130 L/min, that means that your normal inhalations are only going to take up 1/5 to 1/4 of the distance between the horizontal axis (0 L/min line) and the top of the graph.

When I stare at your one-minute close up centered on the OA, I can tell that these inhalations are topping out at around 25-30 L/min. A bit less than the sleep breathing before the arousal, but hardly enough of a drop to be flagged as either an OA or an H. And the overall shape of the breathing is pretty regular AND it's not really any more distorted (in the 10-minute window) than the normal sleep breathing that precedes the arousal. Hence I think all that we're looking at here is your transition back to sleep after the arousal.

So why did the machine score an event? Well those huge inhalations during the arousal most likely skewed the moving baseline average that a PR machine uses to determine events and they would skew the moving baseline upwards. So my guess is that as soon as you were drifting back to sleep after the arousal, the airflow dropped to 25 L/min and that was just barely enough to fool the machine into thinking it needed to score an event. I have no idea why the event is scored as an OA instead of an H, but I've seen this kind of thing in my own data as well. It's also worth noting that the shapes of the inhalations around when the OA is scored are pretty similar to the shapes of the inhalations before the arousal in the 10-minute screen shot. In the one-minute screen shot, there does seem to be some flattening of the inhalations, but apparently it's not enough to make the machine score them as a FL. Or perhaps the fact that the machine labeled them as an OA prevented them from also being scored as a FL. But quite frankly, I think that the flattening is exaggerated by the vertical scale in the Flow Rate graph and I personally think that the breathing after the arousal is actually decent, normal sleep breathing rather than sleep disordered breathing.

In my own data I have noticed when my PR machines have mis-scored sleep transition breathing after an arousal or a wake as sleep disordered breathing events, it will sometimes score OAs that clearly don't look like apneas (because there's plenty of airflow) and it will sometimes score Hs. And often times the "number" attached to these "false" events don't make much sense. And that's another reason I think this OA is just a "false" event that is just part of sleep transitional breathing after the arousal.

As for why the "length" of the event is pegged at 3.75 minutes (224 seconds): It's important to remember that on the PR machines, we don't actually know the full meaning of the numbers attached to each event. When we zoom in on a very clear OA or CA, it's clear that the length of the event is almost always pretty close to the number that the PR machine attaches to that event. But in my data when the event is one like this one---an event where I scratch my head and say, "I'm breathing just fine there"---the numbers are often way far off. So I've also learned to take those numbers with a grain of salt.

Finally I'd love to see how this data shows up in Encore Pro. I've recently become aware of the fact that sometimes Encore and SleepyHead display different numbers of events in their event tables. I've not had a chance to analyze my data closely to figure out how often it happens or even what kinds of events wind up being scored differently. But it's just possible that there's a SleepyHead bug that is misinterpreting some flag the PR slapped on the Flow rate.

xxyzx wrote:

D.H. wrote:Here's what you asked for.

========

i see a snore of long duration not apnea on that recent chart
that is quite likely to be possible

The snore you are looking at is a VS2 snore with a "magnitude" number of 1.

PR machine's scoring of VS2 snores is still a bit of a mystery, but they are scored in two minute intervals and the magnitude number appears to be the number of snores in the two minutes preceding the VS2 flag. The SH Snore graph, however, draws a two minute horizontal line at the magnitude number following the VS2 snore flag. So this snore graph does NOT mean D.H. had a snore lasted two minutes during the time the OA was being flagged. It most likely means that his PR machine thought that he snored once between 2:22:15 and 2:24:15. For a full discussion of what's known about how PR machines score snoring, please read this post: viewtopic.php?f=1&t=88983&start=105#p1182875 in Pugsy's Pointers.

Finally note that Resmed machines score snoring in a very different fashion. When you are looking at a SleepyHead Snore graph for a RESMED machine, the vertical axis is always between 0 and 1, and the ResScan information indicates the higher the snore number, the louder the Resmed machine thinks the snoring is. How Resmed machines determine the loudness of the snoring is anybody's guess since there is no microphone to record the snoring.