Data, at last!

[jules]

No, I turned off the C-Flex completely on my old machine (it had been set on 3) about a month ago, and I haven't used either C-Flex or A-Flex on the new machine at all.

I knew you had turned it off on the plus - wasn't sure on this new machine.

[riverdreamer]

I'm running off to see my pulmonologist, so don't have a lot of time. Just wanted to say I am thinking about you, and hope something can be found to help. I have a couple of ideas that I will come back to later. In the meantime, it is very common in our office to see people react strongly to very tiny interventions or changes. Some have to take tiny portions of a prescription pill in order to tolerate it. Some even react badly to tiny fragments of certain vitamins. So it is entirely possible the BB is having a hard time with tiny leaks or tiny pressure discrepancies, or even other tiny changes in her life or environment.

Hang in there BB! Everybody is pulling for you.

[BleepingBeauty]

I'm running off to see my pulmonologist, so don't have a lot of time. Just wanted to say I am thinking about you, and hope something can be found to help. I have a couple of ideas that I will come back to later. In the meantime, it is very common in our office to see people react strongly to very tiny interventions or changes. Some have to take tiny portions of a prescription pill in order to tolerate it. Some even react badly to tiny fragments of certain vitamins. So it is entirely possible the BB is having a hard time with tiny leaks or tiny pressure discrepancies, or even other tiny changes in her life or environment.

Hang in there BB! Everybody is pulling for you.

Thanks again, riverdreamer. Interesting points to consider, and I hope you do expand on them when you have time. My AHI has been lower the past two nights (8+ each night), a definite improvement over the three previous nights (24, and then two nights at 14+). I'm still tired, but I don't feel quite as lethargic as I did earlier. (I'll take any improvement I can get!)

BTW, I don't think I've ever mentioned this to you, but I absolutely LOVE the pic in your avatar!

[Muffy]

"Episodic CompSAS flareups?" Most mainstream practitioners were shocked as hell to discover that 15% to 20% of the sleep patients they had been titrating on CPAP presented CompSAS. And right now most of them are utterly clueless that "post-adaptation" CompSAS can so episodically and infrequently crop its ugly head in patients who were supposedly adapted to CPAP.

I have a question, -sws. What means that phrase, "Episodic CompSAS flareups"?

As I read this post (and pardon me if I'm misinterpreting this), the implication is that 15% to 20% of "adapted" patients have CompSAS, and I don't think that's what that statistic means.

Muffy

[Muffy]

I like o.'s observation of the leaks. This may be an actual case of "CO2 washout".

We need a real PFT.

Muffy

[-SWS]

"Episodic CompSAS flareups?" Most mainstream practitioners were shocked as hell to discover that 15% to 20% of the sleep patients they had been titrating on CPAP presented CompSAS. And right now most of them are utterly clueless that "post-adaptation" CompSAS can so episodically and infrequently crop its ugly head in patients who were supposedly adapted to CPAP.

I have a question, -sws. What means that phrase, "Episodic CompSAS flareups"?

As I read this post (and pardon me if I'm misinterpreting this), the implication is that 15% to 20% of "adapted" patients have CompSAS, and I don't think that's what that statistic means.

Muffy

I agree the statistic was never meant to imply that 15% to 20% of CPAP-adapted patients have CompSAS. Now get ready for a humorous confession...

When I wrote the phrase "episodic CompSAS flareups" the sheer redundancy of those terms "episodic" and "flareup" made me cringe. True story. But I wanted to informally emphasize that I think CompSAS can be extremely episodic in some patients. I wanted to make that proposed concept really stand out. So I "double popped" it!

But back to your correct statement about that 15% to 20% statistic. That original quantitative assessment was about as unrefined as any epidemiological statistic can be IMHO. Initially, it didn't factor biological adaptation---or any longitudinal presentation characteristics, for that matter. While some clinicians, such as yourself, believe that figure to be inflated, yet others suspect the figure may actually be low. My main point is that medical understanding of CompSAS---epidemiology, pathology, etiology/pathophysiology---is largely a mystery. And science is only at the very beginning of peeling the layers of the CompSAS mystery---as if it were a complex onion of sorts.

It was only after those initial 15% to 20% estimates published in 2005/2006 when multiple research teams realized---based on analysis of historical data---that a good percentage of those CompSAS patients would go on to adapt to CPAP. That work represents a second layer of the CompSAS mystery onion having been nicely peeled and analyzed. However, I think there are plenty of remaining layers of the CompSAS mystery onion waiting to be methodically peeled toward understanding the essence of this problem in physiology.

And I think one of those epidemiological layers that needs to be peeled and further analyzed is that some CompSAS patients are presenting their complex SDB symptoms in episodic fits and spurts---typical of many disorders in which neurological dyscontrol plays a central role. I believe additional longitudinal analysis will reveal that some of those supposedly "biologically adapted" patients are, in fact, presenting occasional CompSAS flareups or episodes, based on an array of presently undetermined pathophysiological triggers---perhaps even somewhat individualized subsets/constellations of subtle triggers, similar to what happens with the VCD patient population. If that latter conjecture bears true, then combinational management techniques (versus any single proposed mitigating technique) just may hold promise for some of those CompSAS patients.

I like o.'s observation of the leaks. This may be an actual case of "CO2 washout".

We need a real PFT.

I think it's a good idea to analyze sustained or baseline CO2 associated with eupneic breathing. However, my very unqualified hunch is that this whale of a physiology problem is hiding in the transient domain of stimulus/response---where survival-based neural adaptive plasticity just may "dysfunctionally supersede" central homeostasis---on a short-term albeit "runaway" basis. If the CompSAS problem in etiology were simply based in the inability to sustain correct CO2 levels during eupneic-breathing, then I think EERS or yet other therapeutic means of CO2 elevation would be far more effective across the CompSAS population. Unfortunately that does not seem to be the case.

Well, the ice has melted, the trusty fishing shanty and hot stove have long since sunk to the bottom of the lake, and we're standing chest-high in hot August water---still engaged in fascinating chat.

[Jason S.]

Me too. AHI 1.4 the second night and 3.6 the first night, but that number was skewed due to recorded events that happened before I fell asleep. So could be between 1-2 for both nights.

[Muffy]

I think it's a good idea to analyze sustained or baseline CO2 associated with eupneic breathing.

I think it would be more telling to monitor CO2 during apnea rather than eupnea.

However, my very unqualified hunch is that this whale of a physiology problem is hiding in the transient domain of stimulus/response---where survival-based neural adaptive plasticity just may "dysfunctionally supersede" central homeostasis---on a short-term albeit "runaway" basis.

I don't think it's hiding at all, it's just that nobody wants to spend the 16 grand for the pair of glasses that would allow them to see it.

If the CompSAS problem in etiology were simply based in the inability to sustain correct CO2 levels during eupneic-breathing, then I think EERS or yet other therapeutic means of CO2 elevation would be far more effective across the CompSAS population. Unfortunately that does not seem to be the case.

I believe the biggest obstacle to making EERS work is leak control. Leaks at the mask rapidly negate the effectiveness of this modality. The concept of using increased FICO2 was around long before CSDB was invented, it's simply that nobody can make it cheaply, reliably and safely work outside of the laboratory.

While the Respironics software looks at leaks in a very coarse fashion (compliments of the "25/32 Rule") I think the pattern of leaks in BB's case shows that her event salvos are always associated with an unstable leak pattern, and that event-free breathing is always associated when Observed Leak approximates Designed Leak.

Finally, I believe what makes BB so susceptible to this phenomenon is her low CO2 reserve. This consideration will be better seen in looking at FRC in complete PFT. Her physical stature should make her at risk for a low value here.

Me too. AHI 1.4 the second night and 3.6 the first night, but that number was skewed due to recorded events that happened before I fell asleep. So could be between 1-2 for both nights.

I think you're in the wrong movie there, Brad.

we're standing chest-high in hot August water---

Water temperature is just a little cool here in the Atlantic surf, a welcome refuge from the otherwise oppressive southern heat.

Muffy

[-SWS]

I think it's a good idea to analyze sustained or baseline CO2 associated with eupneic breathing.

I think it would be more telling to monitor CO2 during apnea rather than eupnea.

Makes sense relative to a transient CO2 issue...

Finally, I believe what makes BB so susceptible to this phenomenon is her low CO2 reserve. This consideration will be better seen in looking at FRC in complete PFT. Her physical stature should make her at risk for a low value here.

And that makes sense since that transient issue relates to a CO2 window or threshold.

While the Respironics software looks at leaks in a very coarse fashion (compliments of the "25/32 Rule") I think the pattern of leaks in BB's case shows that her event salvos are always associated with an unstable leak pattern, and that event-free breathing is always associated when Observed Leak approximates Designed Leak.

I'm going to take a closer look at the patent description regarding leak detection. I agree that BB's event salvos always accompany a much more jagged albeit "low-level" leak line.

The question is whether that jagged leak line is legitimate or just measurement/calculation artifact, with respect to periodic breathing or hyperventilation that BB likely presents.

IIRC, part of the leak algorithm heavily factors detecting disproportionate rate-of-flow changes relative to either recent average or typical flow slopes. And yet another part of the algorithm factors differences between inspired and expired flow volumes---since they should be equal pairs. You're right about the leak algorithm being coarse. Anyway, that jagged leak line may be legitimately jagged because of slight leak variations. But accurate measurement subtleties in instantaneous leak might not be the case for this algorithm...

Rather, I suspect this may be a case of a basic leak algorithm having difficulties with those highly erratic flow-rate changes introduced by periodic breathing or hyperventilation. Associated breath stacking can additionally skew that other part of the leak algorithm looking for deltas between inspired and expired volumes.

However, my very unqualified hunch is that this whale of a physiology problem is hiding in the transient domain of stimulus/response---where survival-based neural adaptive plasticity just may "dysfunctionally supersede" central homeostasis---on a short-term albeit "runaway" basis.

I don't think it's hiding at all, it's just that nobody wants to spend the 16 grand for the pair of glasses that would allow them to see it.

Well, apparently they never met Swimdaddy, who even has a blue "haggle" button in his auctions.

j/k

But measurement and attribution of challenging phenomena are quite the tricky pair...

[Muffy]

I'm going to take a closer look at the patent description regarding leak detection. I agree that BB's event salvos always accompany a much more jagged albeit "low-level" leak line.

The question is whether that jagged leak line is legitimate or just measurement/calculation artifact, with respect to periodic breathing or hyperventilation that BB likely presents.

IIRC, part of the leak algorithm heavily factors detecting disproportionate rate-of-flow changes relative to either recent average or typical flow slopes. And yet another part of the algorithm factors differences between inspired and expired flow volumes---since they should be equal pairs. You're right about the leak algorithm being coarse. Anyway, that jagged leak line may be legitimately jagged because of slight leak variations. But accurate measurement subtleties in instantaneous leak might not be the case for this algorithm...

Rather, I suspect this may be a case of a basic leak algorithm having difficulties with those highly erratic flow-rate changes introduced by periodic breathing or hyperventilation. Associated breath stacking can additionally skew that other part of the leak algorithm looking for deltas between inspired and expired volumes.

I don't believe BB's cycle-generating rogue leaks are "low level". Indeed, they are typically 50 - 100% greater than Designed Leak:



Further, I don't believe that BB's central response to these insults is so violent as to generate artifact (as might be envisioned in a response to purely obstructive events). Even if she did have a substantial hyperventilatory response to events, I have confidence in the reliability of the Respironics Leak Calculation algorithm to accurately represent data. Here is an example of a guy with apparent obstructive events (unfortunately, I am not privy to the actual raw data in this case, so I'm taking the scorer's word for it.)(Not because I want to, I believe that there is also a central flavor underfoot here.)(But I think there's enough similarity to offer this case as evidence of the leak detection accuracy.)

BTW, this is with Quattro at 17 cmH2O so designed leak is about 50 LPM:



If we were to offer Leak Bounce as an explanation for artifact, then the associated Frequency Response with this case of higher pressures would seemingly generate even more spurious data. Yet, this does not appear to be the case here. Baseline leak is generally within a narrow window (subject to 25/32 limitations of measurement).

Muffy

[-SWS]

Pardon a quick detour to mull over the graph posted immediately above...

Here is an example of a guy with apparent obstructive events (unfortunately, I am not privy to the actual raw data in this case, so I'm taking the scorer's word for it.)(Not because I want to, I believe that there is also a central flavor underfoot here.)(But I think there's enough similarity to offer this case as evidence of the leak detection accuracy.)

No preliminary snores... No precursor flow limitations... And almost all of the SDB events are supposedly obstructive apneas, to the tune of 76 per hour. What an atypical presentation.

I would sure think that person is a candidate for some of the less common obstructive-apnea etiologies:

Anatomically narrow and excessively collapsible upper airways are seen in most patients with obstructive disease. Associations include native cephalometric abnormalities, acquired soft tissue abnormalities such as enlarged tonsils, obesity, effects of androgenic hormones, dysfunctional protective upper airway reflexes, upper airway neuropathy/myopathy, and increased airway length.

Dysfunctional upper airway reflexes in particular seem like a possibility for such a dense and precursor-free obstructive event distribution. If counterproductive vagal afferent pressure stimulation is involved, they may want to experimentally turn C-Flex off, hoping for at least marginal improvement. Good luck to that patient.

__________________


But there are clearly different sets of dynamics that can go into both genuine leak variations and artifact-based measurement variations. An example of genuine leak dynamics would be those somewhat uncommon conditions presented by an obstructive patient with fully-occlusive and atypically sudden apneas. Not all obstructive apneas are sudden and/or complete airway closures. But the apneic closures that are both sudden and completely occlusive are specifically the events that will present a very sudden step or spike in flow-impedance. And those sudden flow-impedance spikes in turn will create an instantaneous pressure spike that can more easily breech interface seals.

I distinctly recall Berthon-Jones commenting (in one of his patent descriptions) that instantaneous pressure spikes induced by sudden apneic closures are prone to breeches at the interface seal. That type of leak-prone patient, presenting sudden and very sharp impedance spikes, may very well be what we see on the unusual graph above. Spikes are the operative trait. Notice how very "spike oriented" that patient's leak graph happens to be throughout the entire night. Those comparatively sharp leak dynamics are sudden and frequent versus gradual or sustained. I would guess those particular signature spikes represent genuine leak variations versus measurement or calculation artifact.

Back to my earlier comment about genuine leak variations being different than artifact/measurement variations.

I have confidence in the reliability of the Respironics Leak Calculation algorithm to accurately represent data.

Well, as a rule, citing any presumed accurate signal or data processing instance never proves measurement sensitivity or specificity across the board. That obstructive presentation above probably does not present the algorithm with the same narrow-window or instantaneous flow baseline comparison challenges that various periodic breathing patterns can present. And with a single sensor inside the machine, transient proximal leaks cannot be directly measured. Respironics instead elects to estimate leaks based on variations in instantaneous flow patterns combined with averaging methods.

Unfortunately inferential leak estimates do not make for highly accurate leak measurements. If any single-sensor leak estimate method were submitted for peer review as if it were accurate leak measurement methodology instead, I honestly think the method would be quickly rejected regarding accuracy claims. But inferential leak estimates perform adequately for treatment algorithms IMHO.

[Muffy]

Here is an example of a guy with apparent obstructive events (unfortunately, I am not privy to the actual raw data in this case, so I'm taking the scorer's word for it.)(Not because I want to, I believe that there is also a central flavor underfoot here.)(But I think there's enough similarity to offer this case as evidence of the leak detection accuracy.)

No preliminary snores... No precursor flow limitations... And almost all of the SDB events are supposedly obstructive apneas, to the tune of 76 per hour. What an atypical presentation.

Interestingly, this patient was also placed on supplemental oxygen, so from this thread's perspective, I think this is more like "Oh look. There goes another one."

I would sure think that person is a candidate for some of the less common obstructive-apnea etiologies:

My money is this is Garden Variety CompSAS that came up and bit 'em in the butt, but they didn't turn around to look and see what bit 'em.

I distinctly recall Berthon-Jones commenting (in one of his patent descriptions) that instantaneous pressure spikes induced by sudden apneic closures are prone to breeches at the interface seal.

Keeping in mind that BB is on 12 cmH2O and the other patient is on 17 cmH2O, isn't Mikey Berthon-Jones the guy who invented an algorithm that says obstructive apneas don't happen after 10 cmH2O?

I have confidence in the reliability of the Respironics Leak Calculation algorithm to accurately represent data.

Well, as a rule, citing any presumed accurate signal or data processing instance never proves measurement sensitivity or specificity across the board. That obstructive presentation above probably does not present the algorithm with the same narrow-window or instantaneous flow baseline comparison challenges that various periodic breathing patterns can present.

I believe "That obstructive presentation above" is heavily-flavored with a periodic central component, so it's not "narrow-window", it's "the same window".

And with a single sensor inside the machine, transient proximal leaks cannot be directly measured. Respironics instead elects to estimate leaks based on variations in instantaneous flow patterns combined with averaging methods.

To be clear, with a single sensor inside the machine, transient proximal leaks CAN be directly measured, it's simply Observed Leak - Designed Leak. You just can't do it with a Smart Card.

On the other hand, speaking of MB-J, his software package can.

Muffy

[Muffy]

Further, I believe BB has historical evidence of real OAs, demonstrated at sub-therapeutic pressure, complete with near-diagnostic snoring events and non-lethal Leak Bounce:



Muffy

[-SWS]

Keeping in mind that BB is on 12 cmH2O and the other patient is on 17 cmH2O, isn't Mikey Berthon-Jones the guy who invented an algorithm that says obstructive apneas don't happen after 10 cmH2O?

Well, there are statistical likelihoods and there are statistical twistyhoods.

Here's what Berthon-Jones REALLY said:

The vast majority of obstructive apneas are already well controlled by 10 cmH2O. At higher pressures, events are more likely to be "central" in origin...

Looks like we'll have to disagree on an entire variety of points in this thread, Muffy. But at least we agree on some key points: I believe that BB has an obstructive component, which is typical of CompSAS. As usual, thanks for sharing your views.

[Muffy]

As usual, thanks for sharing your views.

And thanks for having me! I think these were very interesting cases!

Muffy