Question about how the S9 scores hypopneas

[robysue]

I've been looking at some of my data more closely because right from the start it has puzzled me that my S9 scores relatively very few hypopneas even though the vast majority of my events in my diagnostic sleep study were hypopneas with arousal and on both titrations, hypopneas with arousals outnumbered the apneas.

At any rate, I've noticed something in at least a subset of my data:

If there's an obvious large decrease in flow AND there's a flow limitation at the same time, there seems to be a very high probablity that the S9 will label it as a hypopnea.

If there's a similarly obvious large decrease in flow AND there's NO flow limitation at the same time, there seems to be a very LOW probablity that the S9 will label it as a hypopnea.

Does the S9 only record hypopneas if there is both a flow limitation AND a 40 or 50 percent decrease in flow?

[john_dozer]

I've just started going through my data.

As for lab vs. the S9, I have some events declared as Apneas by the S9/Software, but if you look at the curve there is a trace of flow oscillation that matches my breathing just at a very small amplitude. I think it could be some hair splitting as to whether they get declared as Apneas or Hyponeas. I also suspect how you eventually react to those events may influence the determination. I usually have a sudden burst of heavy breathing (make up breathing) following these and suspect that enters the calculation somehow. A lab may have different criteria. I have hyponeas that look nearly the same with the exception of the make up breathing and that includes flow limitation changes which are similar.

I guess I'm saying evaluate what is happening and not so much what label it gets.

[robysue]

I've just started going through my data.

I have some events declared as Apneas by the S9/Software, but if you look at the curve there is a trace of flow oscillation that matches my breathing just at a very small amplitude. I think it could be some hair splitting as to whether they get declared as Apneas or Hyponeas. I also suspect how you eventually react to those events may influence the determination. I usually have a sudden burst of heavy breathing (make up breathing) following these and suspect that enters the calculation somehow. A lab may have different criteria.

Well, I know that the S9 literature says that an apnea requires and 80% reduction in airflow for at least 10 seconds. And you may be right in that whether there is some kind of reaction in terms of a sudden burst of heavy breathing being involved in the calculation.

The important thing is whether the consequences and treatment are the same? If so, then that's what I would focus on.

Well, that's the point in my own case: I am still more tired and more sleepy in the daytime now (on CPAP) than I was before being diangosed. And part of me can't help but wonder if I'm still having a number of "hypopnea with arousal" events that are waking me up (and hence causing the daytime problems), but are not showing up in the hypopnea data for some reason. On many nights I do see places where there are rather obvious decreases in flow rate that clearly last more than 10 seconds, but they're not being flagged by the S9 as hypopneas. It took me a while to discover the correlation of "not being flagged as a hypopnea" with "flow limitation curve" at "fully open". And so the question has naturally occurred to me: Even though I am now using the S9 every night, am I still having enough of the "hypopnea with arousals" that were scored in my diagnostic sleep test to disturb my sleep? And is that why I'm still so tired and exhausted even though my S9 consistently indicates that my overnight AHI is less than 1.5 and is most often less than 1.0?

I haven't yet looked that closely at how big and how long the breaths remain big after the "hypopnea-like" breaths end are yet. But that is an interesting idea, and I will keep that in mind the next time I'm starring at my data.

[john_dozer]

When I first started treatment, I felt worse. Other aspects were improved, alertness and clarity and lethargy. But I felt worse.

Someone suggested that breathing against pressure for several hours can be exhausting at first.

I also suspect that if your apnea is deep enough, you don't get that 'lousy night of sleep' feeling that non-apnea suffers get if they sleep on a rock or only get a few hours of sleep. We feel it other ways, but that particular "lousy" feeling isn't one of them.

Once you get better and start removing that sleep debt, that 'lousy' feeling become possible again. Expecially considering you have something strapped to your face and making noise just a few feet off to the side.

Basically I suspect that feeling from poor quality sleep for 'reasons other than apnea' get eclipsed or even blocked when you have poor quality of sleep from apnea

[Big Daddy RRT,RPSGT]

Were you set on 7-8cmH2O based on an in the sleep lab titration?

If so they should have used the same criteria to adjust the CPAP pressure during your titration that they used for your diagnostic study, so the mild hypopneas have probably been addressed. However if you have been set on 7-8 based on the Auto CPAP at home or some random method then you might need more pressure to resolve these more mild hypopneas even if these are not being reported by your S9. Others have noticed this.

There is also some controversy regarding the criteria used between manufacturers, some saying that certain manufacturers are "under reporting" so their numbers look better. I'm not sure about that, but it is something to be considered especially if you are not sleeping/feeling better.

[robysue]

When I first started treatment, I felt worse. Other aspects were improved, alertness and clarity and lethargy. But I felt worse.

Mental fog and inability to focus have been severe problems that started for me when I first started using CPAP. I've been working with the PA in my sleep doctor's office concerning this rather alarming development.

Someone suggested that breathing against pressure for several hours can be exhausting at first.

This has been such a major problem that I will soon be switched to a BiPAP. I'm hoping that once I can wake up no longer feeling like I'm running a marathon in my sleep every night things might start getting back to normal.

I also suspect that if your apnea is deep enough, you don't get that 'lousy night of sleep' feeling that non-apnea suffers get if they sleep on a rock or only get a few hours of sleep. We feel it other ways, but that particular "lousy" feeling isn't one of them.]

My apnea is moderate (and on the mild side of moderate). And during my diagnostic study, there were no o2 desats. My main problem was with "hypopneas with arousal" (RERAs), not apneas or hypopneas with desaturations on the diagnostic study.

Once you get better and start removing that sleep debt, that 'lousy' feeling become possible again. Expecially considering you have something strapped to your face and making noise just a few feet off to the side.

I'm 12 weeks into therapy. I'm not having any problem sleeping with the mask on. I am still feeling worse than I did last summer before I was diagnosed. And at this point, I am skeptical of explaining my difficulties away with "sleep debt". Sorry for sounding snarly, but since I did NOT feel sleepy until AFTER I started CPAP, I don't really think sleep debt is a major issue for me. Evidence from my sleep studies and from the quality of my pre-CPAP and post-CPAP sleep indicates that I never had much trouble getting into REM either before or after CPAP, but that I may be having some real problems getting into stage 3/4 sleep WITH CPAP.

Basically I suspect that feeling from poor quality sleep for 'reasons other than apnea' get eclipsed or even blocked when you have poor quality of sleep from apnea

I don't understand what you're trying to get at here. Presumably since the S9 reports my AHI as less than 1.0 on most days and almost always less than 1.5, the S9 is effectively treating my apnea. So I should no longer be having "poor quality of sleep from apnea" right?

But to re-iterate, my question is this: Given the following facts:

• my known OSA problem is lots of hypopneas with arousal based on my diagnostic sleep study,
  my S9 data says my AHI is usually well less than 1.0,
  my daytime sleep apnea symptoms have gotten WORSE (not better) since starting CPAP,
  and I am not talking about "occasional" bad nights with CPAP, I'm talking every day I feel less well with CPAP than without CPAP

is it possible that the S9 is significantly underestimating the number of hypopneas I am actually having at night and that the S9 AHI numbers make it look like the CPAP therapy is more effective than it actually is?

Or is it more likely that there is something about CPAP itself that is triggering (new) disruptions of my sleep patterns that for now are more disturbing (in the daytime) than the apnea itself was last summer? And in this case, how much longer will it take before the CPAP-induced distruptions become no more distrubing (in the sense of triggering daytime symptoms) than the moderate apnea itself?

NOTE: I do not intend to give up on CPAP. But I do want to understand why CPAP seems to be more disturbing to my quality of life---in terms of daytime symptoms---than my apnea itself was. I think this is an important issue to get at in addressing the fact that so many people have trouble sticking with CPAP therapy in the long run: No matter how well the therapy treats the apnea itself (and prevents or minimizes the associated health problems), if the treatment itself causes more problems and more daytime symptoms than the apnea itself does, it takes an incredible amount of self-discipline to not throw the towel in and abandon CPAP altogether.

[robysue]

Were you set on 7-8cmH2O based on an in the sleep lab titration?

If so they should have used the same criteria to adjust the CPAP pressure during your titration that they used for your diagnostic study, so the mild hypopneas have probably been addressed. However if you have been set on 7-8 based on the Auto CPAP at home or some random method then you might need more pressure to resolve these more mild hypopneas even if these are not being reported by your S9. Others have noticed this.

There is also some controversy regarding the criteria used between manufacturers, some saying that certain manufacturers are "under reporting" so their numbers look better. I'm not sure about that, but it is something to be considered especially if you are not sleeping/feeling better.

My first titration study led to a prescribed pressure of 9cm based on 30 minutes of sleep at the end of the study, having been at 7cm for most of the night. 8cm was simply skipped over because the last pressure increase was done so close to the end of the night.

I could not tolerate 9cm of pressure at home: Had serious side effects and the sleep doctor's PA then ordered a week of auto-titration at home. The 95% level for that week came in at around 7.8cm. So at the end of that week, my prescription was changed to APAP with a range of 7--8cm. This range is marginally more tolerable, but still difficult for me to put up with. My daytime symptoms are not as bad as they were when I was using straight 9cm, but I still have worse daytime symptoms than before starting CPAP in the first place. The S9's AHI numbers at 9cm and at 7--8cm are statistically indistinguishable: Most nights AHI is below 1.0 (often well below 1.0), almost always below 1.5, and the occasional outlier night above 2.0. [Outlier night being less than one night out of every two weeks.]

Since I've continued to have problems with dealing with the pressure and have continued to have more problems with daytime symptoms on CPAP than pre-CPAP, the sleep doc ordered a bi-level titration. That titration study came up with a titrated set of pressures of: IPAP=8cm and EPAP=6cm based on a whopping 17 minutes of sleep right at the end of the night after I had been on a IPAP/EPAP of 6/4 for most of the night. I am scheduled to pick up a PR S1 BiPAP Auto next week from my DME. So I don't know how well I will do on these titrated levels.

[john_dozer]

Basically I suspect that feeling from poor quality sleep for 'reasons other than apnea' get eclipsed or even blocked when you have poor quality of sleep from apnea

I don't understand what you're trying to get at here.

I guess what I'm saying is imagine a person without apnea having a night of sleep. Now imagine a person without apnea having a night of sleep with a mask strapped to their face and a noise generator making periodic hissing sound next to them.

Their quality of sleep, even though they don't have apnea, will go down. Likely in a discernable way. Probably for days, weeks, or even months before they become use to it.

Now force them to breath against pressure and their quality of sleep will go down again. I was trying to say there's more to sleep quality than apnea. And being able to fall asleep is not an indication you won't be affected. If anything your sleep debt from apnea will help you fall asleep under such circumstances that might otherwise not be tolerable.

I didn't want to make a big deal out of it, because its probably not the major factor here. But take steps to make sure your comfortable during sleep eventhough you fall asleep. Make sure the device is quiet enough for you, you don't get contrained by the hose, you can roll your head into comfortable positions, etc.

[billbolton]

This has been such a major problem that I will soon be switched to a BiPAP.

Why are you sweating on the details of your present treatment approach if you know you are being moved onto another treatment approach

Cheers,

Bill

[DHC]

But to re-iterate, my question is this: Given the following facts:

• my known OSA problem is lots of hypopneas with arousal based on my diagnostic sleep study,
  my S9 data says my AHI is usually well less than 1.0,
  my daytime sleep apnea symptoms have gotten WORSE (not better) since starting CPAP,
  and I am not talking about "occasional" bad nights with CPAP, I'm talking every day I feel less well with CPAP than without CPAP

is it possible that the S9 is significantly underestimating the number of hypopneas I am actually having at night and that the S9 AHI numbers make it look like the CPAP therapy is more effective than it actually is?

Or is it more likely that there is something about CPAP itself that is triggering (new) disruptions of my sleep patterns that for now are more disturbing (in the daytime) than the apnea itself was last summer? And in this case, how much longer will it take before the CPAP-induced distruptions become no more distrubing (in the sense of triggering daytime symptoms) than the moderate apnea itself?

NOTE: I do not intend to give up on CPAP. But I do want to understand why CPAP seems to be more disturbing to my quality of life---in terms of daytime symptoms---than my apnea itself was. I think this is an important issue to get at in addressing the fact that so many people have trouble sticking with CPAP therapy in the long run: No matter how well the therapy treats the apnea itself (and prevents or minimizes the associated health problems), if the treatment itself causes more problems and more daytime symptoms than the apnea itself does, it takes an incredible amount of self-discipline to not throw the towel in and abandon CPAP altogether.

robysue,

Your 'situation' sounds similar to mine - in that my sleep study reported an AHI of 39/hour without CPAP - all hypopneas. With CPAP at 7 cmH2O, the AHI count was reduced to 2/hour.

Your post motivated me to go back to the data I have been collecting since September to parse the AHI into AI (Apnea Index) and HI (Hypopnea Index) for both the S9 and the DeVilbiss IntelliPAP. Here are those summary results:

DeVilbiss IntelliPAP

AHI: 2.41
AI: 0.21
HI: 2.20

ResMed S9 Autoset:

AHI: 0.40
AI: 0.32
HI: 0.08

As you can see, there is a HUGE difference in the AHI and HI values between the devices - with the greatest disparity in reported hypopneas. DeVilbiss allows configuration of the "Event Detection" parameters as seen here:



You asked; "is it possible that the S9 is significantly underestimating the number of hypopneas I am actually having at night and that the S9 AHI numbers make it look like the CPAP therapy is more effective than it actually is?"

I believe that is EXACTLY the case. My sense is that the S9 grossly under-reports hypopneas *IF* using the fairly common definition of a "hypopnea" as being an airway constriction of greater than 50 % (but less than 90 % - when it becomes an "apnea") for a period of at least 10 seconds.

In this post -- viewtopic.php?f=1&t=56345&st=0&sk=t&sd=a#p528275, jnk provides a comparison table of several vendor's differences in detecting apnea and hypopneas. The ResMed device referenced is their S8, and the ResMed 'S9 Autoset Data Management Guide' provides the following definitions of Apnea and Hypopnea:

Apnea
An apnea is the temporary absence or
cessation of breathing. An apnea is scored
when there is reduction in breathing by
75% of the baseline breathing for at least
10 seconds.

Hypopnea
A hypopnea is an episode of shallow
breathing during sleep. A hypopnea
is scored when there is a reduction in
breathing by 50% of baseline breathing
with partial upper airway obstruction for 10
seconds or more. The event is scored after
10 seconds of the hypopnea.

DeVilbiss defines them thusly (reference their 'AutoAdjust Clinical Overview' - found here -- http://www.mediafire.com/?f1965ks6mjkjm02):

Apnea - Amplitude is reduced to 10% of full-scale signal for 10 seconds

Hypopnea - Amplitude is reduced to 50% of the most recent baseline signal for 10 seconds.
NOTE: The AutoAdjust algorithm constantly adjusts the full-scale signal to match the patient’s current
breathing pattern: therefore, the amount of signal loss needed to reach a 50% reduction in amplitude will vary.

As I read this, it suggests the S9 and the IntelliPAP should be producing AHI values that are at least somewhat consistent with one another as both begin event identification with a 50% reduction in airflow. Clearly, this is not the case, and my feeling is that the S9 is purposely under-reporting events.

To your other question - "Or is it more likely that there is something about CPAP itself that is triggering (new) disruptions of my sleep patterns that for now are more disturbing (in the daytime) than the apnea itself was last summer?"

The other factor that *may* be affecting your quality of sleep is ResMed's algorithm for making pressure changes. Illustrating this clearly are a couple of graphs from the 2 devices I posted here -- viewtopic.php?f=1&t=56345&st=0&sk=t&sd= ... 05#p545496. Whereas the IntelliPAP hugs the lower limit setting, the S9 aggressively changes pressures throughout the night. There were some other graphs posted earlier in that topic that further illustrate the point.

In the DeVilbiss Clinical Overview linked above, there is this commentary about their pressure algorithm:

DeVilbiss, in collaboration with a clinical advisory team, developed the AutoAdjust algorithm including the Event Set Measurement. The Event Set Measurement is unique to the DeVilbiss AutoAdjust and allows our devices to make pressure changes every minute and in varying increments which can improve patient comfort and reduce arousals caused by increasing pressure based on a single event, overreacting to artifacts and noise in the signal or decreasing
pressures before breathing has stabilized.

The following is a simplified version of algorithmic activity:
• Every 60 seconds, the algorithm adds a new Window-One to the Event Set, shifts all previous windows down the Event Window’s time continuum, and bumps the old Window-Six out of the Event Set.
• Every 60 seconds, as they occur during that minute, the algorithm scores all breathing events in Window-One.
• Every 60 seconds, the algorithm searches for density trending using two time periods: the most recent minute (Window-One) and the most recent 6 minutes (the full Event Set).
• Every 60 seconds, the algorithm decides whether to increase, decrease, or maintain the current therapy pressure. If the algorithm determines that a pressure change is beneficial to the patient’s therapy, the incremental amount of pressure change is also determined at this time.

Some manufacturers and many bench-top studies tout the speed with which some devices respond to events. However, this may not be a viable advantage. Sleep lab titration protocols using manual techniques commonly respond to events as they are seen by the clinician and with regard to recent event trending.

There are two reasons for responding less quickly to breathing events: fast response times wake patients or bring them to partial arousal; and fast response times have a greater percentage of false responses. The AutoAdjust mimics sleep lab manual procedures by initiating a moderate and controlled response to breathing events.

In another topic billbolton suggested that sensitivity to pressure changes should result in using the device in CPAP (rather than APAP) mode. While that may be true for some, there are plenty of reports to indicate that higher pressures are problematic for many xPAP users, and it follows that an auto-titrating algorithm prioritizing keeping pressure as low as possible while still providing efficacious therapy is desirable.

If my experience is any help, the S9's pressure changes definitely create arousals during the night, and I simply do not sleep as well. Maybe you can use Bolton's suggestion and try straight CPAP for a couple of days just to see if the constant pressure - without the adjustments - might help?

[KatieW]

This is from the Resmed S9 Data Management Guide:

Hypopnea
A hypopnea is an episode of shallow
breathing during sleep. A hypopnea
is scored when there is a reduction in
breathing by 50% of baseline breathing
with partial upper airway obstruction for 10
seconds or more. The event is scored after
10 seconds of the hypopnea.

The Data Management Guide has disappeared from Resmed's site. But I found it here:

http://photos.medwrench.com/manuals/3012-1039.pdf

[robysue]

DHC,

Thanks for much food for thought.

KatieW,

I've read that language many times, but it just doesn't seem to jive with the data that I've seen from my S9 and that's why I'm wondering.

And billbolton,

It's not that I'm really sweating the details: As a mathematician, I'm curious about the data.

But as an apnea sufferer about to switch machines, I'm also trying to make sure that I don't overreact if there's a sudden rather large increase in my AHI and HI numbers. I know the companies use different algorithms and that numbers cannot be compared between different companies. As a mathematician I sure wish it were easier to find real information about how an individual machines record these events so that it would be possible to make a more informed decision concerning what machine to purchase and to make more sense of what's really going on in the data. In my case, I'm also wondering if my adjustment problems really indicate that I need a biPAP? Or is there something specific to the S9 that has made it not the best choice for my apnea? And in that case, am I just shooting in the dark (again) hoping that a different company's machine will somehow be a better match for me?

After all, playing mask roulette is one thing. But if we're also forced to play machine roulette to find a machine that actually works for us, well, that's a lot more expensive than going through multiple masks---particularly if you have a 50% copay (like I do) or have to buy out of pocket.

[jnk]

As for ResMed's S9 event definitions, the whole concept of 'deviation from baseline' is meaningless, medically speaking, when a company fails to release how it calculates the baseline. So those definitions aren't worth the electrons used to generate them, in my opinion.

And I would like to stress again that the purpose of the home-machine numbers is ONLY to have some general indicator to use for finding the best pressures over time. So it isn't so much about what the numbers are and how they were collected; it's about how the numbers are used. And since the only pressure-change choices available at any given time are (1) leave it alone, (2) raise it, or (3) lower it (ignoring, for a moment, the complications of EPR and APAP ranges), the exact nature of how the trending information is collected and reported is not nearly as important as seeing how the numbers change over time with pressure changes.

I realize in this case that hypopneas are the beasts you wish to hunt down and kill, and for that reason it would at first seem that how a particular home machine scores hypopneas would be the important consideration. But, that is not so, in my opinion. That is because even if one machine scores a bunch of hypopneas and another machine does not score them, NEITHER machine tells you which of those changes in breathing actually disturbed your sleep the night before, since neither machine has EEG leads. And that is why looking at home-machine data, which is based solely on flow, and then trying to judge how consolidated sleep may have been the night before based on that flow data alone is, to my mind, an exercise in futility that I would file under "misuse of data." I admit it would be cool if it worked that way--if hypopneas weren't such slippery, nebulous creatures and were easily defined, eradicated, and repulsed. But it just doesn't work that way in the real world, to my knowledge.

Math is useful for many medical situations, but in the area of normalizing human breathing, math can only point in a general direction for choosing a pressure that seems most to stabilize a particular person's airway on most nights. The practice of medicine is often more about the art of making judgments based on broad principles than it is about strict measurements of mathematical science. Airways don't know math.

[ozij]

Robysue,
ResMed's don't respond to hypopneas, be they S8 hyponpnea, or by they S9 hypopneas. So the scoring of those is not really relevant to the quality of therapy you are getting. Resmed's do respond - aggressively - to flow limitations, but those have to happen in order of the machine to respond to them.

Somehow, it does not make sense to me to leave a person who has many respiratory effort related arousals at a miminal pressure that lets those happen. I would say that both your APAP experience at home and your BIPAP titration hint that your minimum pressure should be 7.8 at least.

My personal experience with the S9 makes me think its response to flow limitations (of which I have many) is different from the S8's, in that that S8 would raise pressure and stay there for 20 minutes. The S9 -- if I'm reading my charts correctly -- drops back down far sooner. You may be setting yourself up for repeated arousale be leaving your minimum too low.

If it were me, at this point, and until I had a bi-level machine to try, I would try to run the S9 at a fixed pressure of 7.8 (or 8), and an EPR of 2. That's almost like a bi-level with IPAP=7.8 (or 8) and EPAP=5.8 (or 6).

O.

Edited to disable smilies and enable 8). Thank you DHC

[DHC]

Robysue,
ResMed's don't respond to hypopneas, be they S8 hyponpnea, or by they S9 hypopneas. So the scoring of those is not really relevant to the quality of therapy you are getting. Resmed's do respond - aggressively - to flow limitations, but those have to happen in order of the machine to respond to them.

Somehow, it does not make sense to me to leave a person who has many respiratory effort related arousals at a miminal pressure that lets those happen. I would say that both your APAP experience at home and your BIPAP titration hint that your minimum pressure should be 7.8 at least.

My personal experience with the S9 makes me think its response to flow limitations (of which I have many) is different from the S8's, in that that S8 would raise pressure and stay there for 20 minutes. The S9 -- if I'm reading my charts correctly -- drops back down far sooner. You may be setting yourself up for repeated arousale be leaving your minimum too low.

If it were me, at this point, and until I had a bi-level machine to try, I would try to run the S9 at a fixed pressure of 7.8 (or , and an EPR of 2. That's almost like a bi-level with IPAP=7.8 (or and EPAP=5.8 (or 6).

O.

O.,

>>Somehow, it does not make sense to me to leave a person who has many respiratory effort related arousals at a miminal pressure that lets those happen.<<

What markers/events do you consider credible as indicative of a "respiratory effort related arousal"?

In robysue's case - where her PSG indicated that her fundamental problem is "hypopneas with arousal" - and since, as you have pointed out previously, the S9 algorithm does NOT respond to hypopneas, do you consider the possibility that the ResMed algorithm just may not be the one best-suited for robysue's particular condition?

>>My personal experience with the S9 makes me think its response to flow limitations (of which I have many) is different from the S8's, in that that S8 would raise pressure and stay there for 20 minutes. The S9 -- if I'm reading my charts correctly -- drops back down far sooner. You may be setting yourself up for repeated arousale be leaving your minimum too low.<<

Assuming, of course, that a person does not mind the higher pressures from setting a higher minimum pressure. For many people, and robysue stated similarly upthread, higher pressures are problematic. Further, rapid changes in pressure are contributors to arousals. In that respect, the S9 may be creating more arousals with either/both rapid pressure increases/decreases. It seems so in my experience with the S9.

>>If it were me, at this point, and until I had a bi-level machine to try, I would try to run the S9 at a fixed pressure of 7.8 (or , and an EPR of 2. That's almost like a bi-level with IPAP=7.8 (or and EPAP=5.8 (or 6).<<

I think your suggestion that she use a fixed pressure of 8.0 was interpreted as a smiley - FYI