Sleepyhead question. In event flags, when you hover over an

[tejen]

event, it will show the event and a number in parenthesis. Fot example in Event Flag Strip it willl show an OA event at 01:00 am. If you hover over the event it will show OA (16). What does the number signify? Is it seconds on that event?

Thanks

[Pugsy]

Yep, event duration in seconds when using a ResMed machine.

[TedVPAP]

Yep, event duration in seconds when using a ResMed machine.

What about Respironics?

[Pugsy]

What about Respironics?

It's close but not necessarily exact because Respironics machines don't actually give event duration.
Respironics SH reports will show a lot of single digit numbers in parentheses in the Events tab and we all know that OAs, centrals and hyponeas have to be at least 10 seconds to earn a flag...so if we see an 8 in parentheses next to an OA (or hover over the event flag) it can't be 8 seconds.
Respironics uses some sort of "marker" and Mark when he did SH couldn't ever figure out exactly what those markers did. He knew it wasn't seconds but that's all that was available for SH to draw upon for event duration. I suspect the "markers" had something to do with how Encore used to report total time in apnea (which it quit doing some time ago with a version change) but the markers remained.

It's very close to event duration in seconds though...close enough for government work.
I spent the biggest part of a day with SH and manually zooming in on those flagged events and counting the seconds and it was real close.
Like if I saw an 8 in parentheses...the event would barely make the 10 second count. If it showed 30 in parentheses by my manual count it might have been 33...or 28...but almost always with a 2 or 3 plus or minus and sometimes spot on.

So...it's not official though...because Respironics doesn't report event duration and ResMed does...and you can see it on ResScan.
It's close though...very close.

[TedVPAP]

What about Respironics?

It's close but not necessarily exact because Respironics machines don't actually give event duration.
Respironics SH reports will show a lot of single digit numbers in parentheses in the Events tab and we all know that OAs, centrals and hyponeas have to be at least 10 seconds to earn a flag...so if we see an 8 in parentheses next to an OA (or hover over the event flag) it can't be 8 seconds.
Respironics uses some sort of "marker" and Mark when he did SH couldn't ever figure out exactly what those markers did. He knew it wasn't seconds but that's all that was available for SH to draw upon for event duration. I suspect the "markers" had something to do with how Encore used to report total time in apnea (which it quit doing some time ago with a version change) but the markers remained.

It's very close to event duration in seconds though...close enough for government work.
I spent the biggest part of a day with SH and manually zooming in on those flagged events and counting the seconds and it was real close.
Like if I saw an 8 in parentheses...the event would barely make the 10 second count. If it showed 30 in parentheses by my manual count it might have been 33...or 28...but almost always with a 2 or 3 plus or minus and sometimes spot on.

So...it's not official though...because Respironics doesn't report event duration and ResMed does...and you can see it on ResScan.
It's close though...very close.

Thanks. When hovering, SH places a box around the event time period. The box is placed upsteam of the marker which doesn't seem right. I would have thought that the box should be centered around the event marker. Are the event markers placed in the correct location? I can post pictures if you would lie. Thanks again.

[Pugsy]

Are the event markers placed in the correct location?

I have no idea what the "markers" were..Mark never told me.
Nor where they were placed. I was not a part of the SH development. I was just told "markers"...whatever that meant.
All I know is that a long time ago on my old Encore reports I used to get a "total time in apnea" amount in minutes and seconds at the end of a daily summary report. Now the machine must have had some sort of method of adding up time in apnea to come up with this data point but we were never able to determine duration of each event. As the years went by this data point for "total time in apnea" seemed to fall off the Encore reports.
I never knew or noticed that it happened until a while back I went looking for my hypothesis that those "markers" had something to do with it but by then the Encore report no longer offered total time in apnea so I couldn't compare SH and Encore and see if the markers Mark found had anything to do with it or not. I had nothing to experiment with.
Me....I am good with my day spent counting hundreds of events manually and never finding more than a 2 or 3 difference from what I manually counted (which isn't easy because at that level of zooming it I found it very hard to spot the beginning and ending of the reduction in air flow).
It's close enough to get me in the ball park.

Remember though..they have to be at least 10 seconds to earn a flag and those single digit events you see...have to be at least 10 seconds or they wouldn't be flagged at all so the single digit isn't 100% accurate.
10 seconds...OAs, centrals and hyponeas.
Now the other stuff (FLs or whatever) that SH puts in parentheses ...I have no idea how SH came up with those numbers...Mark never told me.

[palerider]

The box is placed upsteam of the marker which doesn't seem right. I would have thought that the box should be centered around the event marker.

Why?

Are the event markers placed in the correct location? I can post pictures if you would lie. Thanks again.

Compare it to encore.

[TedVPAP]

Are the event markers placed in the correct location?

I have no idea what the "markers" were..Mark never told me.
Nor where they were placed. I was not a part of the SH development. I was just told "markers"...whatever that meant.
All I know is that a long time ago on my old Encore reports I used to get a "total time in apnea" amount in minutes and seconds at the end of a daily summary report. Now the machine must have had some sort of method of adding up time in apnea to come up with this data point but we were never able to determine duration of each event. As the years went by this data point for "total time in apnea" seemed to fall off the Encore reports.
I never knew or noticed that it happened until a while back I went looking for my hypothesis that those "markers" had something to do with it but by then the Encore report no longer offered total time in apnea so I couldn't compare SH and Encore and see if the markers Mark found had anything to do with it or not. I had nothing to experiment with.
Me....I am good with my day spent counting hundreds of events manually and never finding more than a 2 or 3 difference from what I manually counted (which isn't easy because at that level of zooming it I found it very hard to spot the beginning and ending of the reduction in air flow).
It's close enough to get me in the ball park.

Remember though..they have to be at least 10 seconds to earn a flag and those single digit events you see...have to be at least 10 seconds or they wouldn't be flagged at all so the single digit isn't 100% accurate.
10 seconds...OAs, centrals and hyponeas.
Now the other stuff (FLs or whatever) that SH puts in parentheses ...I have no idea how SH came up with those numbers...Mark never told me.

Thanks

[TedVPAP]

The box is placed upsteam of the marker which doesn't seem right. I would have thought that the box should be centered around the event marker.

Why?

Why not??

Are the event markers placed in the correct location? I can post pictures if you would lie. Thanks again.

Compare it to encore.

Don't have encore. They are certainly close enough.

[Pugsy]

Comparing the actual event flag placement...pretty much the same in both software products. Pretty much at the end of whatever event (OA,CA, hyponea).
The "markers" that Mark was talking about are something apart from the event flags and I never knew exactly what the "markers" were.
They do seem to correlate rather loosely with event duration though and why they ended up being used as a duration data point.
It's close enough and the only thing available beyond manual inspection under the microscope of each event which is a lot of work just to find a maybe 2 second variation if that. I got better things to do

ResMed is the only brand that I know of that actually documents event duration.

[TedVPAP]

Here is an example. I think the event duration and the location of the flag/marker both look reasonable. Looks to me like the box just needs to be centered. Truly a nit; just curious as to what the known issues are with SH accuracy when using PR machines. More interesting is the non-event towards the left side. A guess it just snuck under the threshold. I do love PAP and sleepyhead.

[Pugsy]

Event flags and "markers" that I referred to for event duration calculation aren't the same thing. So just wanted to mention that.
Remember that SH only reports what the machine is recording with a few exceptions....like the assumption numbers in parentheses or if someone starts messing with altering the User Identified flagging by changing the defaults.

So when SH reports an event with a flag...the machine is doing the analysis...not SH.
So in terms of accuracy for the big stuff...SH matches Encore (I have used both and compared) with the exception of maybe rounding up or down minor numerical differences.

That non event thing that didn't earn a flag...wouldn't have been flagged in Encore either because it is the machine doing the flagging and not SH.

Those times where we see stuff that might look worse than what got a flag...and I have seen it myself. It's not SH but instead the machine doing the non flagging and probably because there was something in the criteria for earning a flag that wasn't met. Something that we can't visually see most likely.
I like to use this example...assuming that criteria for a hyponea is a minimum of 40% air flow reduction
Suppose the air flow reduction was 39% and lasted a minute...no flag because it didn't meet the 40% minimum criteria wasn't met despite being prolonged.
Suppose the air flow reduction was 75% but only lasted 9 seconds...no flag because of the 10 sec minimum criteria not being met.

The duration is fairly easy for us to evaluate with our eyes but the air flow reduction isn't so easy to always evaluate with our eyes and what is shown on the graphs. These machines were never really designed to be used for this minute under the microscope level of evaluation. They were designed more for the big picture type of thing for trend and pattern evaluation for therapy optimization.
There has to be some line somewhere in terms of earning a flag or not so we live with the ones that "come close but don't earn a flag".
Would having enough of those "come close" things maybe impact how we feel or sleep...of course it could but the line in the sand is what it is and we deal with it. It's not a perfect system no matter how much we might wish it was.

just curious as to what the known issues are with SH accuracy when using PR machines

SH just reports what the machine reports with the exception of what I mentioned above. So it's not SH accuracy in question but the accuracy of the machine that may or may not be the issue. Encore software doesn't show everything that SH shows and it doesn't let us zoom in and put things under the microscope either but SH is going to match up to Encore pretty much unless there is a minor rounding up or down of something because Respironics goes to only 1 decimal point and SH goes to 2 decimal points.
So for the big important (remember important in Respironics eyes) stuff...SH is going to be just as accurate as Encore because it is going to report the same stuff that Encore reports which is what the machine is recording.

Now there is the thing with SH and the AHI statistics when the SD card wasn't in the ResMed machine...that's a different thing.
That is known to be off (but not horribly off) but that is because SH needs some data from the detailed files that aren't available if the SD card doesn't have them for SH to do the AHI thing.

For your Respironics machine and SH accuracy...I think you can rest easy that SH is accurate in what it reports because all it does is report what the machine records.
The duration in parentheses thing...it's close even if not spot on sometimes and certainly better than nothing.

[palerider]

The box is placed upsteam of the marker which doesn't seem right. I would have thought that the box should be centered around the event marker.

Why?

Why not??

because. The marker is put where the event is flagged, not some arbitrary 'it looks better here' position, the box should show the approximate event duration in the flow.

Are the event markers placed in the correct location? I can post pictures if you would lie. Thanks again.

Compare it to encore.

Don't have encore. They are certainly close enough.[/quote]
Get it, maybe it'll answer some of your questions.

When mark and I were working on the current version of SH, I compared all the respironics data. It matched, after a fair amount of tweaking.

[TedVPAP]

palerider - Maybe the picture in the subsequent post helps.