Tracking RERAs with AirSense 10 AutoSet and online vendors

Thanks, everyone, for your insights!

It seems like we don't really know which machine is more accurate at correctly identifying RERAs. If you don't mind, I have a related question. The machine I'm trialing right now (provided by DME for 1.5 weeks), Philips Respironics 550P, generates somewhat confusing data. I initially started my treatment to address RERAs: I had a lot of arousals during the diagnostic study bug my AHI was normal. However, here is what's interesting:

1. First, AHI.
The study was very short in duration (only six or so hours) and I need a ton of sleep to be able to function properly, ideally 9-10 hours of net sleep on most nights. This translates into 10-12 hours in bed, assuming 1-2 awake time. So, obviously, the sleep lab only covered a part of my typical sleep. In fact, I had to catch up on sleep after getting back home.
Now, I'm looking at the SleepyHead data and seeing an interesting pattern: my AHI was super low for the first part of the night but was very high towards the end.
https://ibb.co/dCx3sv

I don't remember being completely awake at that time. This is also confirmed by my tracker (Oura ring which seems to be the most accurate one available on the market) which confirms that I was asleep.
https://ibb.co/fuo9Xv

However, when I zoom in on that segment of high AHI, I don't see too many events. Do you think it's just a fluke in the data? Or can there be some insights that we missed during the sleep lab study due to its short duration?
https://ibb.co/i7TXkF

2. Second, RERA.
As I said, I got the CPAP to treat RERAs. However, this is machine showing that I hardly get any. I see a few possible explanations. First, this machine is very bad at detecting them (compared to EEG). Second, all machines are very bad at detecting them (compared to EEG) – as mine have some unique nature that makes them easy to detect with EEG but hard to detect with CPAP equipment. Third, the machine is treating them so effectively that I don't see them on the charts.

https://ibb.co/kzHskF

For reference:
My RERA index, the diagnostic study in the lab: 9.5
My RERA index, CPAP in the lab: 1.7
My arousal index, the diagnostic study in the lab: 17.6
My arousal index, CPAP in the lab: 4.7
(I'm actually not sure about the difference between RERA and just "arousal" but these indices are reported separately)

I would appreciate any insights! Or perhaps if you have suggestions regarding intelligent questions I might ask my doctor to get to the root of it, I would appreciate that too.

Just for the context, here is what initially brought to the sleep lab: episodes of extreme fatigue and brain fog + a need for a lot of sleep. I apologize if my questions are too basic or if I'm missing something obvious, I'm new to the whole CPAP thing...

Thank you all!

reinvigorated wrote:Now, I'm looking at the SleepyHead data and seeing an interesting pattern: my AHI was super low for the first part of the night but was very high towards the end.
https://ibb.co/dCx3sv

I think you are looking at the AHI graph. That graph is an hourly graph...starts a new computation every 60 minutes.
Essentially a useless graph except real good at causing panic and worry in people who don't understand it.

If you are going to share detailed reports you need to include the stuff on the left side of the detailed page as well as those graphs on the right.
See this thread for examples.
viewtopic/t103468/Need-help-with-screen-shots.html

Your machine doesn't really want to do much until those wee hours of the morning and then it increases the pressure a bit.
2 main culprits for when we see this happen...either REM stage sleep (we usually have a lot more REM in those wee hours) or sleeping on your back..or maybe a little of both. Either one can cause OSA to be worse and/or need more pressure.
I see no reason why UARS would be any different.

reinvigorated wrote: However, when I zoom in on that segment of high AHI, I don't see too many events. Do you think it's just a fluke in the data? Or can there be some insights that we missed during the sleep lab study due to its short duration?
https://ibb.co/i7TXkF

It looks like it goes to 14 that one hour...
The overall AHI is an average of the entire night and that number gets reduced by the hours at the beginning of the night when nothing happened.

If this is a typical pattern for you my first suspect would be REM stage sleep. If this is something you see every night in the wee hours then if it were me I might consider increasing the minimum a bit to better prevent whatever is going on during that hour where the events cluster so much. It can't get to where it needs to go fast enough to prevent them from happening so it needs a little better head start. Some of it is ClearAirway/central and more pressure can't fix those but considering that your diagnosis is mainly for RERA or UARS then the obstructive events flagged could may be causing arousals and the centrals could very well be post arousal centrals (you were maybe sort of awake but don't remember it). If that is the case then better preventing the obstructive stuff should also reduce any arousals that might happen and thus any arousal centrals should also reduce. Lot of maybes and ifs there but not hugely impossible ifs. I have seen it often when pressures are sub optimal...fix the obstructive stuff better and post arousal centrals go away.

Pugsy,

Thank you for the insights! A lot of open questions, indeed!

According to my sleep tracker (the second link in my post: https://ibb.co/fuo9Xv) it was indeed REM sleep. And I do recall sleeping on my back.

As for SleepyHead's left sidebar info, it's my last link in that post: https://ibb.co/kzHskF Let me know if you were referring to something else.

Thank you!

Also, here is more info on hypopneas and obstructive events (left sidebar): https://ibb.co/hCqqpa They seem to be mostly clustered in the morning, indeed.

This is the chart with the overall statistics (five nights so far). https://ibb.co/eP5Uhv

reinvigorated wrote:As for SleepyHead's left sidebar info, it's my last link in that post: https://ibb.co/kzHskF Let me know if you were referring to something else.

Yes..I saw that.
I am old and set in my ways and I have certain ways I need to see things for my mind to click.
I prefer one image that has everything I want to see in one screen shot over 2 images that I have to bounce back and forth to and try to think and correlate what I see.
Years of habit and it's just what I prefer.

It would help us help you better if you did the screen shots like we recommend. No need to redo this one though.
I just mention it in case you want to share other detailed reports.

REM worse OSA...I have first hand experience. My OSA in non REM is about 12 per hour...in REM nearly 60 per hour.
I have significant differences in pressure needs during REM compared to non REM....like 6 to 8 cm more pressure.
Sleeping on my back doesn't seem to make anything worse but REM sure does...it was documented in my first sleep study.

robysue wrote:

As to the mask pressure graph, yes, when zoomed out, it's useless, when examining breaths, it can tell you more about what's actually going on, while the 'pressure' graph is the target pressure, the mask pressure graph shows measured pressure.

Again, when someone is trouble shooting, this graph doesn't really tell us anything that's not picked up in the other graphs.

In other words, in my opinion, that graph is bells and whistles. It's nice to see the "EasyBreathe" (that was NOT easy for me to breath with) in action so that you can explain to someone how EPR vs. Flex works. But it's not a critical graph for tweaking the therapy settings.

I have found it useful, and recordings of the actual pressure isn't presented anywhere else in other graphs, on any other machine.

Thank you, Pugsy! Now I see what you meant – I'll screenshot the whole screen next time, no problem.

Thank you so much for helping. Honestly, I'm blown away by how helpful people are on this forum and how they are willing to spend their time looking at someone else's charts!

reinvigorated wrote: It seems like we don't really know which machine is more accurate at correctly identifying RERAs.

That's a reasonable conclusion: We really don't know enough about the proprietary algorithms used by Resmed and PR to determine RERAs to be able to definitively tell which machine is more accurate at detecting RERAs.

If you don't mind, I have a related question. The machine I'm trialing right now (provided by DME for 1.5 weeks), Philips Respironics 550P, generates somewhat confusing data. I initially started my treatment to address RERAs: I had a lot of arousals during the diagnostic study bug my AHI was normal. However, here is what's interesting:

1. First, AHI.
The study was very short in duration (only six or so hours) and I need a ton of sleep to be able to function properly, ideally 9-10 hours of net sleep on most nights. This translates into 10-12 hours in bed, assuming 1-2 awake time. So, obviously, the sleep lab only covered a part of my typical sleep. In fact, I had to catch up on sleep after getting back home.

Several comments:

1) Since you sleep much longer than six hours on an typical night AND since OSA is often worse in REM AND since the length of REM cycles increases towards the end of the night, it is quite possible that the official untreated AHI on your sleep study is lower than your true untreated AHI.

2) You said you need 10-12 hours in bed to get 9-10 hours of sleep "assuming 1-2 (hours of) awake time". How often do you think you are spending 2 hours of wake time during your time in bed? That's an awful lot of wake time, and if you could decrease the wake time, you might actually feel better even if you were not getting much more actual sleep.

Now, I'm looking at the SleepyHead data and seeing an interesting pattern: my AHI was super low for the first part of the night but was very high towards the end.
https://ibb.co/dCx3sv

I don't remember being completely awake at that time. This is also confirmed by my tracker (Oura ring which seems to be the most accurate one available on the market) which confirms that I was asleep.
https://ibb.co/fuo9Xv

Looking at the two images together, what I notice is that the Oura ring data indicates you had a REM cycle sometime around 9:30 to 9:45, which is exactly when there's a cluster of events in the SH events table. (Although the number of CAs in this cluster is a bit strange.)

There's also a REM cycle in the Oura data around 7:30ish, and there are several sleep disordered breathing events scored during that time frame as well.

There are two smaller REM cycles in the Oura data after 11:00, and the last events that show up in the SH data may be related to those REM cycles.

Question: How much REM sleep was scored on your diagnostic PSG?

However, when I zoom in on that segment of high AHI, I don't see too many events. Do you think it's just a fluke in the data? Or can there be some insights that we missed during the sleep lab study due to its short duration?

Like Pugsy, I think you are talking about the AHI graph, which is confusing to a lot of newbies (as well as some not so newbies.). I wrote a brief explanation of how that graph works back in this thread. Somewhere I wrote a longer, more detailed explanation, but I can't seem to find it tonight.

2. Second, RERA.
As I said, I got the CPAP to treat RERAs. However, this is machine showing that I hardly get any. I see a few possible explanations. First, this machine is very bad at detecting them (compared to EEG). Second, all machines are very bad at detecting them (compared to EEG) – as mine have some unique nature that makes them easy to detect with EEG but hard to detect with CPAP equipment. Third, the machine is treating them so effectively that I don't see them on the charts.

Without going through the zoomed-in Flow Rate data, there's no way to really tell which of your hypotheses is most likely.

I will add this, however: The Oura ring data shows a large amount of WAKE time during the period between 22:00 and 2:30. How much of that WAKE time do you remember being awake? The thing about RERAs is that they are so short (usually the arousal is 30 seconds or less) that you don't remember being awake. Is there any way to get better time stamps on the Oura ring data? And what's the shortest "wake" that the Oura is capable of showing?

And I think it's also significant that the Oura ring data shows much less WAKE time after 2:30. If you could cut out the restlessness during the first half of the night, you might feel much better even if the AHI didn't decrease.

For reference:
My RERA index, the diagnostic study in the lab: 9.5
My RERA index, CPAP in the lab: 1.7
My arousal index, the diagnostic study in the lab: 17.6
My arousal index, CPAP in the lab: 4.7

We need more than this data to make sense of what's going on with your CPAP data at home.

What did the sleep staging data look like on the diagnostic study? How much WASO (Wake after sleep onset)? How much REM? How much Stage 3/4?

And what did the sleep staging data look like on the CPAP study? How much WASO? How much REM? How much Stage 3/4?

It would be even better if you could post the summary graphs from your sleep study reports. The hypnograph and the events table from both studies would be very, very useful in trying to figure out what's going on in your sleep right now.

(I'm actually not sure about the difference between RERA and just "arousal" but these indices are reported separately)

A RERA is a respiratory effort related arousal---in order for a RERA to be scored, there has to be evidence of increasing respiratory effort (usually a flow limited sequence of breaths) that ends with an EEG arousal and a return to normal sleep breathing.

An arousal that is NOT labeled as a RERA, respiratory related, or related to periodic limb movements (PLM) is called a spontaneous arousal---it's NOT caused by respiratory problems and its NOT caused by PLMs. You just aroused for some (unknown) reason. A certain number of spontaneous arousals are considered "normal" since even people with totally normal sleep will have some of them. There's a scientific paper from 2007 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2564772/) that I'm aware of that studied EEG arousals in people with normal sleep. The major conclusion of the study is that the number of EEG arousals increases in a statistically significant way with age in people with normal sleep.

I would appreciate any insights! Or perhaps if you have suggestions regarding intelligent questions I might ask my doctor to get to the root of it, I would appreciate that too.

If I were you, I would bring up the following questions with the sleep doctor:

1) Could the sleep test underestimate the real extent of your sleep disordered breathing given that you usually get 9-10 hours of sleep and your Oura ring data shows that almost all of your REM sleep occurs during the last 3 or 4 hours of your night. In other words, your Oura ring data shows very little REM during the part of your night that would correlate to the time you spent in the lab.

2) Are the number of spontaneous arousals on your CPAP sleep test in line with the number of spontaneous arousals for a person of your age?

3) Are there ways to improve your sleep efficiency by reducing the amount of time you spend in WAKE during the first half of each night?

For what its worth. I have a AirSense Auto 10 For Her model. I have run it in both "His and Her" modes and get RERA data on both using SleepyHead.

joelrk35 wrote:For what its worth. I have a AirSense Auto 10 For Her model. I have run it in both "His and Her" modes and get RERA data on both using SleepyHead.

Good to know.

Can you tell us when your machine was manufactured?

Wake Time

robysue wrote: 2) You said you need 10-12 hours in bed to get 9-10 hours of sleep "assuming 1-2 (hours of) awake time". How often do you think you are spending 2 hours of wake time during your time in bed? That's an awful lot of wake time, and if you could decrease the wake time, you might actually feel better even if you were not getting much more actual sleep.

The Oura ring data shows a large amount of WAKE time during the period between 22:00 and 2:30. How much of that WAKE time do you remember being awake? The thing about RERAs is that they are so short (usually the arousal is 30 seconds or less) that you don't remember being awake. Is there any way to get better time stamps on the Oura ring data? And what's the shortest "wake" that the Oura is capable of showing?

And I think it's also significant that the Oura ring data shows much less WAKE time after 2:30. If you could cut out the restlessness during the first half of the night, you might feel much better even if the AHI didn't decrease.

That's exactly how I feel about this wasted time as well!

What's interesting though, is that had it not been for the Oura ring data, I would have guessed that I'm probably awake for about 20-30 minutes each night (let's say about 15 to fall asleep + 5-15 more to fall asleep again after getting up to use the restroom). But Oura data shows that I'm awake for much longer, between 1 and 2hr each night, about 90 minutes on average. However, there are caveats:
- I'm not sure how accurate Oura ring really is (though it seems quite accurate at least directionally /in terms of trends)
- If longer awake time might be due to the fact of my sleeping longer than usual (I need a lot of sleep but wasn't always able to get as much; but I have more time this July, so I have been making a conscious effort of getting as much as sleep as I can in order to feel better)

REM and Deep Sleep

robysue wrote: Question: How much REM sleep was scored on your diagnostic PSG?

Interestingly, both % of REM sleep and % of deep sleep were low for my age. The numbers were a bit higher during the sleep study when I tried CPAP and during the sleep study when I tried an oral device, but still below the norm. Here is the summary of key metrics:



This is in line with my Oura ring data. According to Oura (average for July):
- Average time asleep: 9.3hr
- Average deep: 1.1hr (11.8%)
- Average REM: 0.9hr (9.7%)

Sleep Study Data: Summary

robysue wrote: What did the sleep staging data look like on the diagnostic study? How much WASO (Wake after sleep onset)? How much REM? How much Stage 3/4?
And what did the sleep staging data look like on the CPAP study? How much WASO? How much REM? How much Stage 3/4?

It would be even better if you could post the summary graphs from your sleep study reports. The hypnograph and the events table from both studies would be very, very useful in trying to figure out what's going on in your sleep right now.

Here is more data from the sleep studies. Please let me know if the more graphs/charts can be helpful. I can share tehm too.




The caveat: I felt really weird with all these sensors on me and cautious about accidentally moving some of them, so I didn't sleep very well. Also, I couldn't sleep in my favorite position on the stomach because of the sensors, so I think that also negatively affected the quality of my sleep.

Sleep Study Data: Diagnostic
All screenshots are here in one album: https://ibb.co/album/gEObyv

Sleep Study Data: CPAP
All screenshots are here in one album: https://ibb.co/album/bsvOdv

Today's Data
Today was interesting. The trend with high AHI (in particular driven by hypopneas) continues, it's still much higher than the sleep lab one. Also, I didn't sleep as well today as I usually do.
Here is SleepHead and Oura summaries:



SleepyHead, zoomed in on the problematic clusters:
All screenshots are here in one album: https://ibb.co/album/ix5dBF

Thank you so much for your help, I appreciate it so much! Also, thanks for the explanation of AHI and the difference between arousals and RERAs!

reinvigorated,

You've given me a lot of data to think about. So it may be a day or two before I get back to you.

There are a few more questions that I have for you, as well as an idea for you to think about.

1) Was your diagnostic sleep test a home test or an in lab test?

2) What's your take on the difference between the Oura ring data and the sleep stage scoring on each of the tests?

Now for the idea for you to chew on: It's just possible that many of your RERAs on the sleep studies were borderline "hypopneas with arousal". What I mean by that is: On your CPAP test, the standard for scoring hypopneas was at least a 30% drop in airflow and EITHER an EEG arousal OR a 3% O2 desat. If many of your RERA events almost met the lab's standard of a hyponea with an EEG arousal, it's just possible that your CPAP might be scoring your residual RERAs as H's instead of RERAs because they may just barely meet the PR definition of H in terms of the airflow drop and the duration needed to score an H.

Certainly your machine scored HI looks a lot like the CPAP study's RERA Index than the HI.

robysue wrote: 1) Was your diagnostic sleep test a home test or an in lab test?

All the three tests referenced above (diagnostic, oral device, and CPAP) were done in a sleep lab. I had done a home test as well prior to doing these three. I can find the report for that one too, but basically, it pointed in the direction of UARS and served as a reason for me to do these three sleep lab tests.

robysue wrote: 2) What's your take on the difference between the Oura ring data and the sleep stage scoring on each of the tests?

To be honest, I don't know. I'm more inclined to conclude that Oura tracking is imperfect, especially on a small scale (e.g. just one night). Overall, though I saw reports that positively evaluated its accuracy, for example here one blogger compared Oura data with EEG and it looked good: https://www.alexfergus.com/blog/the-one ... ing-review

That's why I wanted to test it myself and wore it during the studies. However, directionally it still seems to show the right pattern, i.e. REM/deep % seems to be lower in my case for some reason.

Also, it seems to be mostly about the timing of REM sleep. It looks like even though the sleep studies were extremely (well, by my standard...) limited in their duration, I still got between 10% and 15% of REM sleep each night. However, according to Oura, most often my REM sleep takes place towards the end of the night, sometimes after 8 hours of sleep. Oura rarely shows any REM sleep early or in the middle of the night (with few exceptions, like today). So, it's possible that Oura doesn't classify the REM sleep I get early in the night and in the middle of the night as REM for some reason and instead classifies it as light sleep. The algorithm is a little bit of a block box (machine learning, I'm pretty sure), so it's hard to say what could cause that. But maybe temperature or heart rate change patterns I have significantly differ from the average and that, in turn, skews the data.

robysue wrote: Now for the idea for you to chew on: It's just possible that many of your RERAs on the sleep studies were borderline "hypopneas with arousal". What I mean by that is: On your CPAP test, the standard for scoring hypopneas was at least a 30% drop in airflow and EITHER an EEG arousal OR a 3% O2 desat. If many of your RERA events almost met the lab's standard of a hypopnea with an EEG arousal, it's just possible that your CPAP might be scoring your residual RERAs as H's instead of RERAs because they may just barely meet the PR definition of H in terms of the airflow drop and the duration needed to score an H

Certainly your machine scored HI looks a lot like the CPAP study's RERA Index than the HI.

I guess that's possible. I still a few days left, so I'll keep tracking and will then ask my doctor about all these things. Really hope to find some practical implications.

Thank you so much again! I don't want to take too much of your time, but if you could point me in the right direction or suggest some areas to investigate with my doctor, that would be tremendously helpful. Thanks again!

This night seems to confirm the pattern: very few RERAs but more hypopneas than detected by the sleep studies. Also, still need a lot of sleep and still not getting enough deep sleep and especially not enough REM. The overall quality of sleep was much higher this time though.

I've also realized that it looks like CPAP clock is off, it's one hour ahead. E.g. what SleepyHead is showing as 11am is actually 10am.

reinvigorated wrote:I've also realized that it looks like CPAP clock is off, it's one hour ahead. E.g. what SleepyHead is showing as 11am is actually 10am.

Edit your user profile and adjust the time zone.
Respironics machines don't have a clock that you can set.
Now if you end up with the ResMed machine it has a clock that is easily set by the patient.