S9 missing 2x Hyopopneas - in detailed view [bug?]

[dave21]

Hi,

I had unfortunately a bad night last night. Setup as usual was the S9 AutoSet with the Swift FX (trying to still get used to the mask and thought I had finally got used to it).

So my AHI is usually 0.1 on average every night (almost perfect), but last night my AHI was 0.6. From an AHI perspective that doesn't sound like a bad night, however I know that I awoke a huge amount of times last night. Checking the graphs I had 16 abnormalities in the sine waves - it looks like I re-adjusted my sleep position 16 times and 5 times where I awoke through the night.


Usually I get 7.5 hours sleep but only managed 6 hours sleep last night and figured there was no way I was going to doze off for another 1.5 hours so at that point got out of bed.

Summary
Summary shows I had 1x Apnea and 5x Hypopneas making an AHI = 0.6. This is also what the S9's basic and detailed screens show (AI = 0.1 and AHI = 0.6 thus HI = 0.5)


Full night detailed graph
Detailed graph only shows 1x Apnea and 3x Hypopneas. So AHI = 0.6 - 0.1 - 0.3 = 0.2. So there's 2x Hypopneas missing somewhere that ResScan detailed graph doesn't think is a Hypopnea but the S9 and the ResScan Summary screen thinks are Hypopneas.


To make sure there's no hypopneas closer together, here's a zoom


Hypopnea #1


Hypopnea #2


Hypopnea #3


Apnea #1


So where are the missing Hypopnea's? Here's my thoughts looking through the graphs that I think I've found 1x 8 second Hypopnea and 1x 12 second Hypopnea and 1x 10 second Hypopnea.





There's no other Hypopneas showing in detailed graph that are potentially >5 seconds. Bearing in mind that a Hypopnea isn't rated as a Hypopnea until it's lasting for at least 10 seconds, I would say the last two graphs could/should be classed as a Hypopnea in the detailed graph view but the first graph technically is shown as an 8 second so I think we can discard this as it isn't really a Hypopnea for arguments sake.

Any thoughts at why ResScan detailed graphs wouldn't be detecting the last two as Hypopneas?

Thanks
Dave

[DreamDiver]

It looks like you're seeing the data skew bug I posted about earlier. Check the beginnings and endings of your nights' flow data for the last few nights and see whether they line up perfectly with other line-graph data. Mine have been skewing approximately 15 seconds forward for everything but flow every night for the last week or so. I'm now at about two minutes skewed from the correct positioning of events. See the above post for how I reset my skew data... I'm not sure if this helps, but it looks similar.

[dave21]

At first that was the first thing that crossed my mind, but the data lines up perfectly for the Apnea and three Hypopneas, it's just missing two Hypopneas. I'll try to reload from the card into a new patient record and see if that makes any difference but it looks like it's possibly another bug where although the data is aligned it's not actually picking up a couple of Hypopneas.

My guess here is that there's some other logic that the Detailed view is using to determine if there is a hypopnea vs what the Summary screen and machine are using. My gut feeling is that maybe the machine writes out the summary data and that's what the Summary tab and machine are using, but it's possible that either the S9 got it wrong when writing out the summary data or that ResScan is getting it wrong.

I hit a similar problem earlier this month where the AI and AHI figures didn't match up on the SUMMARY vs SLEEP RPEORT screen on the S9. It showed different on the S9 vs what ResScan was showing - viewtopic.php?f=1&t=50288

[DreamDiver]

At first that was the first thing that crossed my mind, but the data lines up perfectly for the Apnea and three Hypopneas, it's just missing two Hypopneas. I'll try to reload from the card into a new patient record and see if that makes any difference but it looks like it's possibly another bug where although the data is aligned it's not actually picking up a couple of Hypopneas.

My guess here is that there's some other logic that the Detailed view is using to determine if there is a hypopnea vs what the Summary screen and machine are using. My gut feeling is that maybe the machine writes out the summary data and that's what the Summary tab and machine are using, but it's possible that either the S9 got it wrong when writing out the summary data or that ResScan is getting it wrong.

I hit a similar problem earlier this month where the AI and AHI figures didn't match up on the SUMMARY vs SLEEP RPEORT screen on the S9. It showed different on the S9 vs what ResScan was showing - viewtopic.php?f=1&t=50288

This happened to me before I noticed the data skew. There was one central that had no similar flow data in it's vicinity at all. Suffice it to say, there are definitely still some bugs in ResScan 3.10 - but it's still the best software for viewing data out there. Far better than EP2.2.

[dave21]

Yeah I agree it's still great software.

I tried to do a complete new patient import from my backed up data but still see the same problem (although data isn't skewed), so it looks like it could be another bug.

Thanks
Dave

[LoQ]

Your first post is pretty long, there's way too much data, and I confess I didn't read the whole thing, just long enough to get the gist. I could easily be mistaken, therefore. And if I am, I certainly apologize for that in advance.

That said, I personally don't see anything missing and wonder if you are confused about the definition of AHI/HI/AI versus number of events.

You slept for 6 hours; you had 4 events (1 apnea, 3 hypos). Your AI is 1/6, which is 0.1 if you truncate, which is what appears to be the case. Your HI is 3/6, which is 0.5. Your AHI is 4/6, which is 0.6 when you truncate. AI + HI = AHI. The math works out just fine, since .1 + .5 truly does equal .6.

It seems to me, from reading your post, that when the data reported by the machine says AI = .1 and HI = .5, you interpreted that as 1 apnea and 5 hypopneas. If you had slept for 10 hours, that would be true. But the data from the machine shows that you slept just a few minutes over 6 hours. Therefore, you didn't have 5 hypopneas, only 3, as reported by the machine.

[Arizona-Willie]

Yes, I think the problem is in rounding.

1 apnea divided by 6 hours =.166666.

Their rounding isn't very good. They should have rounded that to .2 not chopped it to .1.

Last night I also had an AHI of .6
Obstructive apnea = .4
Central apnea = .1
Unknown apnea = .1

But I slept 7 hours and 21 minutes which is 7.35 hours.
The graph showed 2 obstructive apneas.

Ooops I just expanded my graph and one of the apneas was actually 2 so that made 3 apneas. One covering the other on an 8 hour scale. So when I change my calculation to 3 / 7.35 they got it right .408.

Might want to zoom in on those apneas and see if you got a double somewhere.
I thought my machine was wrong at first then it turned out to be me that was the dummie .

[dave21]

I agree, the rounding seems to be a rounding down incorrectly and not rounding up.

Sorry I also forgot to mention as the original post was very long that the S9 summary review screen showed 0.3 Hypopneas but the Sleep Review screen showed 0.5. So this seems very similar to what I experienced in viewtopic.php?f=1&t=50288

I think it's a couple of small glitches in ResScan possibly as opposed to the S9.