Flow Runs? (with 420e)

[rested gal]

Is there in fact reasonable empirical evidence for the guideline: Set your minimum to a value such that most of the time is spent at that pressure?

No empirical evidence that I know of. A good range just depends on each person, I suppose. I know it would be nicer to have more guidance to go by, but I don't think it's out there...not where autopap ranges are concerned. Wide open 4 - 20, or any min/max between... I'd think it's just going to be what suits each person. And even then, a particular range might not suit the same person on some nights as well as it usually does.

I'm sorry...I didn't make it clear that the a, b, c's of what I look for on the Silverlining "Detailed" graph don't mean that's what I think others should look at and base their pressure setting decisions on. Those are simply the areas of the graph I look at and are what I expect/want to see based on the minimum pressure I chose by a method that had nothing to do with what I'd been seeing in my data. That method being something I discovered purely by accident:

Over a year ago, one night when I layed down to sleep with mask on, machine going, and began to drift off to sleep, I realized that when the inside of my throat relaxed, my throat slammed shut. When I noticed that, I deliberately kept it as relaxed as I could and tried to draw in a breath...no way. Not a thread of air could come through. That epiphany happened while I was still awake enough to notice it. At that time, I had my autopap set at 6 - 20, so obviously 6 wasn't good enough...for me...not even awake... if there was a lot of relaxation.

I didn't change anything right then. Just repeated the relaxation in other sleep positions and felt the same thing happen consistently. Made a mental note to myself..."I'm gonna raise that lower pressure tomorrow!"

Next day, I tried the same experiment while not using the machine. Tried it while sitting up, wide awake at the computer. Deliberately let the back of my throat relax and BAM... soft palate, back of tongue, back of throat... all slammed shut together. Tried to inhale...nada.

That afternoon, I put on the mask, fired up the machine and tried various straight pressures while laying down. 7 kinda' helped. 8 almost prevented it. 9 was BINGO! So, that's how I arrived at what I wanted MY minimum pressure to be.... 9. In effect, probably almost the pressure I'd likely be prescribed if I were using straight cpap.

Actually, I knew that even as relaxed as I tried to deliberately make it, the total relaxation of sleep would probably cause closure even at 9, and that really I'd likely have been prescribed at least 10 or 11 if I had had a study. But I settled on 9 as I knew from trying it that 9 would keep my throat fully open at least while drifting off to sleep, and maybe even through light sleep on my side, which was my preferred sleeping position anyway.

So, that's simply what I found out about how my own OSA seems to happen. Someone else might not need any cpap pressure at all...until they are in certain positions and/or certain stages of sleep. Others might need some, but not much, until certain conditions were ripe for events for them. Those people would probably do fine with the minimum pressure set considerably lower than "prescribed". For me, though, since I already knew total blockage could happen while I was wide awake AND sitting straight up if my throat relaxed, it seemed pointless for me to try to use a pressure lower than the 9 that I found would prevent it.

That's not as elegant a way to arrive at a minimum pressure as teasing it out of data or having an established rule to go by. It just happened to be a "noticed this happening" simple way for me to decide how to set my minimum.

One could ask, "If you're going to set the minimum up pretty close to or at your prescribed pressure, what's the point of having a more expensive autopap? Why not just use straight cpap at your prescribed pressure?" For me, I want a margin up top...a maximum pressure for those nights, or times during a night, that more might be needed. Without having to set a "worst case scenario" single high pressure going all night.

11 seemed to be what worked for me during "worst case scenario"...in REM, on my back, during a PSG sleep study I had a couple of years after I had begun autopap treatment. And that does seem to be what my data from different brands of autopaps shows most of the time, most of the nights... rocking along for the most part at 9, sometimes 10, occasionally 11, and a few rare brief excursions up to 12 or 13 -- not often and not for long.

Ah, silly me. I was assuming that all events were being reported on the graph. But if this is true, events are occurring, being responded to, but not being reported. That does make it difficult to infer much from the graph.

Yes...if we're talking about precursor indicators of an impending "event". The word "event" usually is taken to mean a flow limitation bad enough to get marked as such, and hypopneas, and apneas. Doesn't mean that "events" which should earn a tick mark are happening and being dealt with behind the scenes but not being marked because they were dealt with.

Gee, I hope the code doesn't look like this. If it does, that would certainly explain some of the slowness of response.

Heheh, I sure hope it doesn't either!

If a tree falls, and no one is in the forest....

Ah, but the question isn't whether anyone will hear (have heard) the sound. The question is whether the hidden Markov model will have predicted the sound .

I knew that wasn't quite to the point when I wrote it...but, heck, I liked the "sound" of it. I'm doing well to just spell "algorithm"!!

But more details of the model would be nice to know in any event.

Yes, indeed. Very interesting stuff, and I do enjoy reading about it even though I don't grasp much of it. That's wayyyy out of my league, ghmerrill. -SWS is the man for that!

[Snoredog]

ghmerril wrote:

I have now looked at the "Performance of Auto-adjust Nasal CPAP Devices ..." document by Eiken and McCoy. I guess I don't want to start a big debate about this work here, and consequently won't offer a scientific opinion of it. But I should say that I do not find it to be "helpful", if you take my meaning. I suppose that I could start with their references list and unravel things from there, though that material is now at least five years old. I suppose I should just contact our information management people and tell them to send recent articles on APAP studies Smile.

I would take that study (not really a study anyway) with a grain of salt. That comparison was done on a "simulator" with no peer review, very hard for simulators to duplicate a human response to anything. If you read the original article out of the sleep rag where that report came from (see below), you will see all the problems with it as pointed out by Puritan Bennett and missing from the study. Notice how the Resmed Spirit always came out on top? In the article you find where they were the only financial contributor and technical advisor. You could fart in that study and that machine would still come out on top Click below and read PB's comments.

http://www.sleepreviewmag.com/article.p ... 006/05&p=9

Your current situation: If you are having ANY CA's show up on the Silverlining reports you need to change the way you look at setting up the 420e machine. Even though that machine is supposed to be the best at avoiding central and mixed apnea, you need to help it along. That means you need to DROP your Max. pressure down to your 90% pressure or even 1cm below it, that means down from 15cm to 11cm or even 10cm. Next, you need to drop the Min. pressure from 9cm down to like 6.5cm or 7cm, go as low as you can while still allowing for comfortable breathing air. Don't starve yourself with the Min. setting. Then pull a report, I bet you will see a marked difference the next day.

Biggest mistake people make is increasing the floor pressure in an attempt to lower HI's, sometimes it works but for many it doesn't. My experience is when you increase that Min. pressure to 9cm, it seems to "mask" events that the machine would otherwise would need to "see" as a pattern trigger and either ignore or respond to later on. Then when these events manifest themselves later the pattern is not complex enough for the machine to respond to. It's like it took 3 skips, one hop and a jump before you did "this". With the bottom pressure being up, it misses seeing the 3 skips and one hop and only sees the "jump", if you catch my drift.

I would set your 420e to:

Min. Pressure: 6.0 or 6.5cm
Max. Pressure: 10cm or 11cm.

Lowering the Max. pressure should reduce the number of CA's seen. Lowering the Min. should reduce the flow limitation runs (because it cannot see them now with current pressure settings masking them). I would turn 1FL back On. With it Off, it won't respond to flow limitations, BUT it WILL record them even if masked by pressure.

then again, I could be all wrong

[ghmerrill]

rested gal & snoredog:

Thanks much for these comments. I will use these to guide some additional experimentation. I don't have quite the same problem as rested gal, though I can certainly empathize with her. About a decade ago I had farily extensive sinus surgery, which was definitely worth doing and solved a problem with recurring low-grade sinus infections. At that time, the surgeon also carved a bit off the soft palet in an attempt to reduce my snoring. It helped a little bit, but didn't solve the problem.

I note that the recommendations of rested gal and snore dog actually go in different directions: one to increase the minimum and one to decrease it. I've pretty much been following the "increase minimum" approach so far. Just for grins, I'll try snoredog's approach and see what happens. After that, I may well start to treat this as a real experiment, record data, do analyses, etc. .

Regarding the Eiken/McCoy study. I'm afraid I can't take this with a grain of salt (I know that Snoredog is actually trying to be kind here). The problem doesn't lie (as several list contributers have suggested) in using a simulation. It lies in the more fundamental methodologies. I read both articles, and my reaction to PB's comments was that I felt they were incredibly restrained. I can well imagine the sequence of circumstances (involving first markenting, then R&D, then legal) that resulted in that carefully crafted and very temperate response. It really is well done. It sends one quite intelligible message to the community of APAP users, and it sends another quite direct message to the scientific community.

I am more than a bit surprised that -- having funded the "study" -- Resmed actually allowed it to be published. That, I would speculate, was a decision of their marketing people rather than their scientific people. I would speculate that their R&D folks are almost certainly embarrassed by it. But of course, it wasn't really published, it was "published". These are, of course, only speculations from a neutral onlooker. I have no dog in that fight, and I do not work for a company that develops or markets durable medical equipment.

[rested gal]

I note that the recommendations of rested gal and snore dog actually go in different directions: one to increase the minimum and one to decrease it. I've pretty much been following the "increase minimum" approach so far. Just for grins, I'll try snoredog's approach and see what happens. After that, I may well start to treat this as a real experiment, record data, do analyses, etc. .

Good idea, gh. One of the many benefits of having an autopap and the software is to be able to experiment to see what works best for you.

[ghmerrill]

Well, last night I set it to a min of 7 and a high of 10, and with IFL1 turned on. I feel like I didn't sleep at all, although I know I did for some periods.

The graph shows the pressure starting at 7 and then almost immediately going up to 10 and sitting there all night. 92% of the time was spent at the max. This is consistent with my previous experience with IFL1 turned on.

There were 3 apneas, 4 CA, 9 hypopneas, 7 hypopneas(fl), and 105 acoustical vibrations (!). The graph shows little tick marks along the (virtually straight) pressure line that seem to correspond roughly to the hypopneas and central apneas.

I suppose I shouild try increasing the max, but I suspect this will result in pretty much the same behaviour at a higher pressure.

[-SWS]

Regarding the Eiken/McCoy study. I'm afraid I can't take this with a grain of salt (I know that Snoredog is actually trying to be kind here). The problem doesn't lie (as several list contributers have suggested) in using a simulation. It lies in the more fundamental methodologies. I read both articles, and my reaction to PB's comments was that I felt they were incredibly restrained. I can well imagine the sequence of circumstances (involving first markenting, then R&D, then legal) that resulted in at carefully crafted and very temperate response. It really is well done. It sends one quite intelligible message to the community of APAP users, and it sends another quite direct message to the scientific community.

I am more than a bit surprised that -- having funded the "study" -- Resmed actually allowed it to be published. That, I would speculate, was a decision of their marketing people rather than their scientific people. I would speculate that their R&D folks are almost certainly embarrassed by it. But of course, it wasn't really published, it was "published". These are, of course, only speculations from a neutral onlooker. I have no dog in that fight, and I do not work for a company that develops or markets durable medical equipment.

Ghmerrill, if you're insinuating the original purpose and methodology of the test seem "trumped up" toward serving the funding company's vested market interest, several posters on the message boards have taken that stance in the past. Your point that competing companies were likely constrained by "tempered" legal and corporate considerations, in issuing their own responses, is unquestionably valid.

Regarding the methodology itself: toward serving the purpose of accurately comparing APAP algorithms and responses, a crucial patient feedback loop is open, thus clearly rendering the entire study null-and-void. Toward serving the sole purpose of feeding a simulated open-loop pattern into all APAP machines... indeed, such a methodology seems conducive to yielding artificial outcomes that would clearly favor the funding company's APAP machine. If your presumption of underlying motives is correct, then the study would be exceptionally "well done" (as you said) toward achieving misleading data objectives. The fact that the scientific community would instantly perceive this methodolgy as flawed (regardless of intent), and that the APAP user community would not, is correct in my opinion as well.

I also have no vested interest in the PAP market. However, I prefer to simply point to the study's methodology and point out that the patient-response feedback loop that is so crucial has been opened and thereby renders the entire study null-and-void. I have always believed in giving people the benefit of the doubt... perhaps to a fault.

BTW, the PB 418P patent number is 6532959. Therein lies much algorithmic commonality with the 420e. You had mentioned that you have a research staff who has access to a common medical-study search repository. On an altogether different topic, it would be interesting to eventually open one or more threads to compare the few pharmaceutical products that purportedly target obstructive apneas in some patients. We haven't had any in-depth discussions of how valid or invalid pharmaceutical approaches might be for certain obstructive apnea etiologies. And I think that topic would fall under your areas of expertise. Undoubtedly your first objective should be to continue focusing on your own PAP therapy and results. However, I just couldn't help piquing your interest toward helping us with that second topic as well... in the distant future perhaps.

[ghmerrill]

-SWS:

Again, thanks for the detailed comments.

I was not intending to insinuate anything about that study's being skewed by their funding source. In fact, I hadn't noticed that connection prior to (rather quickly) forming my view of the methodologies (and I use the term loosely) employed in the study. Even independent of the absurdity of the absent paitient-response feedback loop, I cannot imagine that study being accepted by a a referee.

You had mentioned that you have a research staff who have search access to a common medical study database. On an altogether different topic, it would be interesting to eventually open one or more threads to compare the few pharmaceutical products that supposedly diminish obstructive apneas in some patients. We haven't had any in-depth discussions of how valid or invalid pharmaceutical approaches might be for certain obstructive apnea etiologies.

Yes, that would be interesting. Unfortunately, some of that is a bit outside my particular area of expertise, though that in itself wouldn't deter me. That's what colleagues are for, eh? If one were, for example, to start to run heavy-duty data mining algorithms on very large observational data bases containing hundreds of millions of prescription and treatment records, "interesting" things might be discovered. But one could not do this even in one's spare time because of certain legal and moral constraints, not to mention regulatory issues. It would be an excellent exercise for academics, and a wonderful opportunity for a research grant. Alas, the academics (outside of collaborative arrangements) don't have access to the data.

As I'm sure you really are aware, I can't begin to go there in a venue such as this. I could at best point to published reports and studies, and it would take years to get those out. Still, it is something to think about.

In my work (whatever that is), I tend to make use of examples based on my own experience with (and understanding) of medical conditions. So far this has involved making use of examples based on such conditions as Tetralogy of Fallot, Hodgkin's Disease, prostate cancer, etc. (you can see what a fun time we've had in our family ). As I learn more about sleep apnea, I can start to make use of that as well. At the moment, I don't see any action around here pertaining to treatments for apnea -- though of course it is of interest as a comorbidity or effect. Still, something to keep in mind.

[Snoredog]

Well, last night I set it to a min of 7 and a high of 10, and with IFL1 turned on. I feel like I didn't sleep at all, although I know I did for some periods.

The graph shows the pressure starting at 7 and then almost immediately going up to 10 and sitting there all night. 92% of the time was spent at the max. This is consistent with my previous experience with IFL1 turned on.

There were 3 apneas, 4 CA, 9 hypopneas, 7 hypopneas(fl), and 105 acoustical vibrations (!). The graph shows little tick marks along the (virtually straight) pressure line that seem to correspond roughly to the hypopneas and central apneas.

I suppose I shouild try increasing the max, but I suspect this will result in pretty much the same behaviour at a higher pressure.

understand that higher pressure can trigger more centrals. so increasing the Max. pressure is NOT what you want to do.

And it explains WHY you were titrated at 7cm. Anything over that and it appears you have more centrals. Your centrals went up or were not seen before, now they are. I suspect your acoustical vibratory snores is the cause of the problem. Machine increases to eliminate those snores which trigger more centrals, a stupid viscous cycle it is. You are going to have to lower the Min to 6.0cm and lower the Max to 7cm or set them both to 7cm. Note: that machine is only able to detect 62% of central apnea seen, so you could be having 38% more. The sensitivity values are shown below. How is 2FL set?

PB says:

However, we feel a hypopnea without a flow limited breath does not warrant a pressure increase due to the probability that such an event may not be due to airway obstruction.

I take the above to mean they think hypopnea seen alone are central hypopnea. So it is very possible those 9 Hypopneas seen above are actually central hypopnea (the other 7 have flow limitations associated with them).

NOTICE below that Snore detection is more sensitive than Hypopnea and even Central detection? This means the machine will see those snores and forget about your hypopnea and centrals. The only way you can stop that is by limiting the Max. pressure so it doesn't trigger the centrals. I'm also afraid they needed one more parameter on the 420E, that is 1SL to turn snore detection On/Off.

Hey don't feel bad, join the club, that is the exact reason I cannot use a lot of these autopaps on the market either. The ONLY autopap on the market that may help you is the Remstar Auto w/cflex and it is only because it limits snore detection response. Even though it does have a high priority snore detection "circuit" it specifically limits said response to that trigger. Some of these other machines say ah I hear a snore and that's all I'm going to listen for until they are gone. Sometimes they just don't go away. Resmed machines are the worst for this condition.

Now what you need to determine is WHAT pressure triggers those centrals and limit your machine to avoid that pressure (look at any PSG titration table). These apap mfg's need to rethink this 10cm limit, I'm finding it is much lower.

Apap mfg's should also incorporate the following:

snore detection enable/disable
snore detection sensitivity: allow you to increase/decrease sensitivity.
HI+flow limitation enable/disable
CA limit pressure: allow you to input pressure instead of hard coding 10cm.

FYI:
The GoodKnight 420 Evolution uses the same event detection technology as does the GoodKnight 418P.
For your information, provided below are the sensitivity and specificity figures for the GoodKnight 418P.


Event Sensitivity Specificity
Apnea 100% 99.8%
Central apnea 62% 100%
Hypopnea 52.3% 99.8%
Snoring 90.8% 99.6%
Runs of flow limitation 86.1% 96.5%
GoodKnight 418 P

then again, I could be all wrong to. It would be interesting to see what SWS thinks.

[/quote]

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): 420E, resmed, Titration, Hypopnea, auto, APAP

[Snoredog]

Now that I read this all again, I would make 1 change and stay the course with the current pressure settings 1 more night if you can (still would rather see you down to 6.5cm on the Min. side though).

Turn 1FL back off or to 0 (zero). Let's make sure those pressure runaways seen last night were from the FL detection and not a snore response. If with 1FL off and it still pegs at 10cm then you'll know it was snores as the trigger. Any hypopnea will be taken care of with 2FL, so leave that on.

Leave 2FL on as it is by default (hypopnea + FL detection), this is the 2 skips and the hop.

[ghmerrill]

Snoredog:

This business about "missed" CAs doesn't make a lot of sense to me, given the data I'm looking at over the past month.

You seem concerned that if I set a pressure range of, say 8-15 it will mask CAs. But the data doesn't seem to indicate that. Looking across my various experimental settings (with each typically used for at least several days), I see that CAs range from 1 to 15 per night. This with a setting around the 8-15 or 8-14 range and with IFL1 set off.

The 15 is an anomaly, but most often I see CAs of 1 to maybe 3 or 4. This is consistent with what I saw last night with IFL1 on and the lower range of 7-10. So in terms of CAs being displayed by the software, there is virtually no difference between my prior settings (8-15, IFL1 off) and the new setting (7-10, IFL1 on). And in addition, with IFL1 off and the 8-15 range, the pressure rarely if ever goes above 12, centers around 9, reduces the snores, and I feel much more rested. It just doesn't seem to be the case (at least on the basis of what last night showed) that I have more CAs. My CAs with the broader range are typically between 1 and 5, and it looks like the average is 3.

One could, I suppose, argue that the higher (e.g., 8-12) is triggering more CAs and that these are "not seen", but there's absolutely no empirical evidence for that. Empirically speaking, if the system is undertecting CAs, it may as easily be undertecting them as much (or more) at the lower pressure as at the higher.

I understand that higher pressure can trigger more centrals. I just don't feel I have any evidence that in my case it is triggering more centrals.

The reason I got the APAP was that:

• Over time, it seemed that my Resmed CPAP set at 7 had become less effective (both in terms of snores and in terms of responding to apneas -- I was being awakened at times with apneas).
  Changing the CPAP setting to 8 seemed to improve this (in both eliminating snores and the awakenings).

Using the 420e, with IFL1 off and a range of, say, 8-14 or so, I seemed to be getting some pretty good results and the actual pressure used seems to center around 9 as the mode. What started this thread was my concern about "runs".

Your suggestion appears to be that in fact a more serious condition (centrals) is being masked and should be addressed. I'd certainly like to determine if this is the case, but it's not obvious that I have the tools to do it, and it doesn't look as though the data supports it -- unless I've misunderstood the issues.

[-SWS]

The central-apnea rate ghmerrill described sounds extremely normal and extremely incidental to me. The flow runs index of 25 in all likelihood pose absolutely no problem as well. Flow limitations do not have a standard scoring criteria in the sleep industry as far as I am aware. My own observations are that some patients will score more flow limitations on the 420e than other APAP models by competing manufacturers.

Even though flow limitation runs in those quantities are usually not a problem, I personally think it's not a bad idea to methodically experiment toward possible improvement in therapy. I think ghmerrill is doing a great job of analyzing his data toward possible improvements.

Snoredog, PB announced a couple/few years ago that the 420e sensitivity & specificity specs were identical to the 418P. That was when the 420e was first released, however. I am unaware of any changes to the 420e specifications since then. Please note the specifications of the less-severe obstructive events, such as flow-limitation-runs and hypopneas, are only relative to the respective manufacturer-defined scoring definitions for each of those event types. To the best of my knowledge there still aren't universally or internationally accepted scoring definitions for hypopneas or flow limitations. Sleep science is a very young branch of medicine with much standardization and very many discoveries yet to come in my opinion.

[Snoredog]

Snoredog:

This business about "missed" CAs doesn't make a lot of sense to me, given the data I'm looking at over the past month.

You seem concerned that if I set a pressure range of, say 8-15 it will mask CAs. But the data doesn't seem to indicate that. Looking across my various experimental settings (with each typically used for at least several days), I see that CAs range from 1 to 15 per night. This with a setting around the 8-15 or 8-14 range and with IFL1 set off.

The reason is I am NOT ignoring your original titration of 7cm. You have by setting your minimum pressure well above it at 9cm.

I am not concerned at all about the flow limitations. FL's don't cause desaturations, but CA's can and should be avoided at all costs.

When you set your range to 8-15cm, everything under that 8cm pressure is masked. That means if you previously had 4 CA's appear at 7cm pressure, those will not be counted. So you increased pressure and the number of CA's went down, but did they really?

You feel like crap so you increase pressure even more.

You were originally titrated in the lab where 7cm pressure was found. If that PSG was done correctly, that found pressure resulted in the best sleep with fewest events and it didn't trigger central apnea.

That titrated pressure is fairly LOW. WHY? Why did they stop there and not continue on to 12cm like the autopap found?

Is it because they may have seen CA's along the way? If they did, it should be on your titration report and explains why they stopped at 7cm. Maybe the PSG tech seen your snores and didn't care about them.

Next, you use a autopap with a wider pressure window and your pressure jumps up to 12cm. Why is that?

Is the machine missing something seen during the manual PSG titration that the autopap cannot detect? The autopap doesn't have EEG, a chest strap and snore microphone or a human being to help make its decision. The lab tech could have ignored the snore mic and stopped your titration when they seen centrals.

The next question is, what is causing the pressure runaways to the Max. pressure setting on the 420E?
the 1FL setting?
or the acoustical vibrations?

You changed 2 parameters at once so we don't know. You are at risk of pressure induced centrals, that number moves around as you change pressures. It moves either because they are now unmasked or "seen" or the change resulted in more events. In the case of CA's they go up with pressure if they are pressure induced.

For example; If you have centrals seen at 8cm pressure and you set the Min. pressure to 9cm, the machine will NOT see and detect those events below the 9cm pressure value. So the scoring for them drops, but have you really eliminated those CA's seen at 8cm?

Always use how you feel as the indicator.

[Guest]

For example; If you have centrals seen at 8cm pressure and you set the Min. pressure to 9cm, the machine will NOT see and detect those events below the 9cm pressure value. So the scoring for them drops, but have you really eliminated those CA's seen at 8cm?

If

• I have centrals seen at 8cm pressure, and
  I set the min pressure to 9cm, and
  the centrals seen at 8cm are then masked, then

wouldn't I expect to see fewer centrals at the 9cm setting (since the ones being masked would not be seen)?

If the average number of centrals is iinvariant with pressure, doesn't this imply that higher pressure settings are not masking centrals? The only alternative to this conclusion is the view that (on average) exactly as many centrals are induced by the increased pressure as are masked by it -- since this is the only condition that would result in the invariance. I would have to regard this as a possible -- but not plausible -- hypothesis.

How could I test this hypothesis? Well, if I had an EEG I suppose I could . Short of that, it seems to me the best I can do is to do a sequence of tests on a number of nights , each at a constant pressure setting (7, 8, 9, 10, etc.). If the CAs are invariant across these tests, then it would strain credulity to suppose that CAs were being masked. I shall perhaps do this.

[ghmerrill]

That last was from me, of course. Somehow didn't get logged in.

[Snoredog]

[quote="-SWS"]The central-apnea rate ghmerrill described sounds extremely normal and extremely incidental to me. The flow runs index of 25 in all likelihood pose absolutely no problem as well. Flow limitations do not have a standard scoring criteria in the sleep industry as far as I am aware. My own observations are that some patients will score more flow limitations on the 420e than other APAP models by competing manufacturers.

Even though flow limitation runs in those quantities are usually not a problem, I personally think it's not a bad idea to methodically experiment toward possible improvement in therapy. I think ghmerrill is doing a great job of analyzing his data toward possible improvements.

Snoredog, PB announced a couple/few years ago that the 420e sensitivity & specificity specs were identical to the 418P. That was when the 420e was first released, however. I am unaware of any changes to the 420e specifications since then. Please note the specifications of the less-severe obstructive events, such as flow-limitation-runs and hypopneas, are only relative to the respective manufacturer-defined scoring definitions for each of those event types. To the best of my knowledge there still aren't universally or internationally accepted scoring definitions for hypopneas or flow limitations. Sleep science is a very young branch of medicine with much standardization and very many discoveries yet to come in my opinion.