An APAP Shootout (sort of) on Academic Journal

SWS

Have a good trip & enjoy that warm climate.

Cheers

Doug

-SWS wrote:
But speaking of wonderful threads, I would like to move a copy of SamCurt's quote below into drbandage's very important thread that is currently underway:

What I can just say is, if Uncle Sam thinks it's big enough problem, NIH and other grants would prove big enough to acquire all CPAP producers in the world, let alone doing bigger investigations.

The extremely important thread:
viewtopic.php?t=16617&postdays=0&postorder=asc&start=0

As a follow up to the various posts here I want to add how I have fallen in love (once again) with Barcelona (or Barthelona when I was last there a very long time ago (Sitges actually - just nearby) ).

SAG you are a champion. Your courage & persistance is a credit to who and what you are. Please keep it up - your insights are slowly filtering through.

As to the value of researchers and their work all I can say is SAG please keep it up and please keep posting here. Gracias.


Cheers

DSM

dsm wrote: Have a good trip & enjoy that warm climate.

Thank you, Doug! I agree with all your glowing comments about SAG.

Speaking of APAP-efficacy methodologies... Someday I would enjoy hearing more details about the home-based methodology you devised to compare HI accuracy. What a wonderfully stimulating thought. I'm doubtful that I would personally be able to devise a suitable home-based methodology toward assessing HI accuracy. Indeed, leak calculation versus a fixed mask-based leak table or look-up function in firmware is but one of several HI accuracy factors. And mask leakage calculation itself is really a very easy thing to calculate with reasonable accuracy. Think of quantitatively superimposing the two factors of combined orifice-flow-rate calculations (across a 4-to-20 cm pressure spectrum) with comparisons of patient-volume-in versus patient-volume-out. Now perform a fairly easy curve match. The major advantage I see between these two viable design approaches really has to do much more with the issue of relinquishing real-time resources than mask-leakage or HI accuracy itself, in my honest opinion.

drbandage on the topic of SDB awareness wrote:As you (and others) probably can tell, I feel quite passionate about this subject.

Thank goodness, doctor. And thank you as well!!!

It only takes one well-placed stone to slay the giant of missing awareness. But me and my silly paradigm shifts... If each stone only stands a one-in-a-million chance of slaying that giant, then more than a million of us should join arms with our slingshots very carefully aimed.

Regarding the Barcelona methodology. Like SAG I have been aware of both of these bench-studies for quite some time. I personally favor the Barcelona methodology of those two. Yet there are also some machine interpretation caveats that I hope to discuss when I get back. Take care all!

SWS,

Re the HI data.

Lets start with this statement from me & see where it leads as I am sure that as the discussion unfolds, AI & HI data readouts will become part of the topic.

My experience in 2005/6 was that when I began buying & trying different cpap machines, I naively assumed there was some science to the Auto side of them in that being medical devices, they would provide consistent data such as AHI AI & HI numbers.

After trying my Remstar Auto & a Sprit Auto I was somewhat taken aback to see that the AHI AH & HI numbers bore little resemblance to each other between the 2 machines.

At this time I had no leaning toward one being more accurate that the other but I did observe both over some months & at that time concluded the data from the Spirit appeared to be closer to what was observable of my sleep (along with considerable input from my wife).

So the first point I will put forward is ...

1) That Autos appear to often provide vastly different AHI AI & HI numbers between brands.

2) If one accepts point 1) then one can ask "I wonder which machine provides the more accurate data" - I concluded at that time that the Spirit did - now 18 months later and with a lot more machine types tested (Bipaps Vpaps) plus the benefit of a PO & other general observations, I feel I got it right 1st go. My tests are not laboratory tests, just me as someone with an electronics & computer background with a penchant for solving complex problems who has done as much as I can to answer such questions for myself.

So issue one is how does one cope with the significant variations in nightly statistics produced by different brands of Auto - my approach was to tell people here to regard the nightly data as not absolute but relative to the brand they had & not to get hung up about how precisely accurate it is because to do so is to be placing ones faith in misleading data. One brand may give an AHI of 3.6 & another 0.6 for essentially the same nights sleep.
If one understands the nature of the brand & how it scores, one can make allowances & just get on with therapy. I make allowances for the higher HI scoring in the Resmeds vs the lower AHI numbers that seem to come from the Remstar.

I would like to see research data from one of the research groups on the scoring. And yes I do hear the message that it seems no two companies can agree on what a hypopnea is & how to score it (I don't fully buy this line), but all that tells me is it is then up to me to use my best diagnostic talents to figure it out for myself - which I did.

DSM

#2 Just wanted to add that if SAG is willing to add any commentary re the accuracy of the AHI data from one brand to another - he could end this discussion before it begins (for me that is) as I would accept his judgement as being as good as I could get on the issue - i say this in respect of him being a medical researcher with a lot of the best facilities at his disposal
_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): CPAP, AHI, Hypopnea, auto

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): CPAP, AHI, Hypopnea, auto

StillAnotherGuest wrote:Course, on the other hand, in order to get someone to do a large scale study on humans using the different APAPs, I mean, y'gotta eat, so who's gonna cough up the dough to take say, 6 different machines, x-hundred patients, and do multiple night PSG (a single night won't do it, you really need a mean over several nights)? The time factor alone, each "volunteer" patient needs to spend 18 nights (or maybe 21 if you want to establish a baseline) in this project. So if we got 2100 test periods per x-hundred patients, even if we run 2 patients per tech, and depending on the salary structure of the area where you do this thing, you get up to 300,000G's in acquisition costs alone real quick. Add in the professional component (Dr. ABSM-guy, what do you charge for a month of your time?), administrative costs (oh yeah, administrative costs)...

...so again, who's gonna cough up the dough? "Sponsored by a grant from the Acme CPAP Company." There ain't a lot of other motivated groups out there.

Ooh, maybe we can get Consumer Reports. Right after they're done with the infant car seat thing.

Fun Things to Know About Forced Oscillation Technique (FOT)

(FOT-- that still sounds like a potty-mouth word!)

You can end up with a lower pressure:

AutoSet vs. Somnosmart

The Autoset titration pressure (P95) was on average significantly higher than the Somnosmart titration pressure (9.9 ± 2.6 cm H2O vs 7.0 ± 2.5 cm H2O, respectively; p = 0.005). The P50 of the Somnosmart was on average 2.4 ± 1.5 cm H2O lower than the P50 of the Autoset. Moreover, the P50 of the Somnosmart (4.5 ± 0.7 cm H2O) was quite close to the lower pressure limit of the device. Figure 1 shows the Autoset recommended pressure (P95) plotted against the Somnosmart recommended pressure. It can be seen that 12 of 15 patients had higher Autoset than Somnosmart recommended pressures (p = 0.009). Inspection of a Bland and Altman plot (Fig 2 ) displays considerable lack of agreement between the Autoset and the Somnosmart P95s. The bias was calculated at 3.0 cm H2O.

So we're talking about 3.0 cmH2O less. That's a lot.

It could very well be due to FOT (snicker)(sorry, can't help myself):

Effects of High-Frequency Oscillating Pressures on Upper Airway Muscles in Humans

In summary, we have shown that when the upper airway is briefly subjected to a pressure wave of 30 Hz with an amplitude of 4 cmH,O, there is an increase in the activity of some muscles of the upper airway. This occurred in normal subjects and in patients with a range of obstructive sleep apnea. In the latter, the response of the upper airway muscles can be sufficient to open the obstructed upper airway. The occurrence of similar responses in the sternomastoid and diaphragm suggests that a number of receptors are activated by the stimulus. It is possible that these responses could be utilized in developing new treatments for sleep apnea.

And what is an extremely interesting tidbit:

Although it is clear that the stimulus we used is an artificial one, it does mimic in part the high-frequency ressure oscillations that occur in snoring. We suggest that one of the reasons that a snorer can resist complete upper airway obstruction, despite the generation of suction pressures as high as -80 to -100 cmH,O in the upper airway, is that the pressure oscillations and tissue vibration trigger reflex activation of the genioglossus and other upper airway muscles. It is possible that a reduction or loss of sensitivity of such a reflex might be a mechanism by which snoring evolves into sleep apnea.

Now THAT may certainly give some thought about having surgery to correct snoring!
SAG

-SWS wrote:Forced oscillation is a sonar-like measurement technique and not a treatment technique, Doug. Supposedly snoring-surgery can adversely change the airway impedance-mapping characteristics or results of that sonar-like measurement.

Just saw your post regarding home-based HI accuracy methodology. Thanks! Indeed, I'd like to discuss that when I get back even though I'm not SAG. .

SWS

? the HI stuff was addressed to you - SWS ?

The FOT stuff was a reply to SAG - I was sure FOT was built into the SOMMNOsmart machines & SAG was questioning its value.

I will post an extract from a research paper that mentions it

DSM

#2

FOT ...

Since the Somnosmart device works with a dynamic signal, ie, forced oscillations, it is necessary to assess the possible role of FOT on upper airway collapsibility. Henke and Sullivan9 observed that high-frequency (30 Hz), low-pressure (< 1 cm H2O) oscillations increased the electromyographic activity of genioglossus, sternomastoid, and diaphragm in sleeping humans. Furthermore, these authors have shown that this increase of electromyographic activity was associated with a partial or complete reversal of the upper airway obstruction with an increase in tidal volume. Such reflex responses of the upper airway muscles in response to forced oscillations may play a role in lowering the need of CPAP in the Somnosmart device.

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): CPAP

SWS,

Here is the info on FOT as used to aid CPAP (from K. G. Henke and C. E. Sullivan - Australia) ...

http://jap.physiology.org/cgi/content/abstract/75/2/856

The promise is that the FOT signal can trigger a clearing of a blockage in-flight - kind of a 'tickle the tonsils' - trigger a muscle reaction that opens the airway.

SAG was 'snicker'ing (sic) at the Kessler study that claimed the SOMMNOsmart appeared able to run on average 3 cms less than a Resmed Autoset because the SOMMNOsmart uses FOT to clear obstructions while the Resmed machines use a pneumotachographm (flow sensing) & we already know that the Autoset doesn't try to clear an in-flight blockage but waits it out.

DSM

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): resmed, CPAP

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): resmed, CPAP

Sorry about re-addressing this, Doug. But I find the topic absolutely fascinating. I was aware of FOT for several years. But I was not aware of the 30 Hz 4cm stimulation effect. I'm working the subject through here, simply because it's a patient message board where patients come to discuss topics. Been working these topics through in this manner and learning along the way for quite a few years. IMHO all topics brought out for discussion here are public. Doug, I hope I didn't preclude SAG from responding---I never have before! My apologies to any who are offended, but the following FOT discussion is addressed to absolutely anyone who finds the topic itself fascinating.

But, indeed, Forced Oscillation Technique was originally devised as a SONAR-like airway impedance mapping technique. Anyone can do a search and they'll find the medical community has traditionally used FOT as a measurement technique rather than treatment or response. However, this is the incidental finding, and not the original application of FOT:

In summary, we have shown that when the upper airway is briefly subjected to a pressure wave of 30 Hz with an amplitude of 4 cmH,O, there is an increase in the activity of some muscles of the upper airway. This occurred in normal subjects and in patients with a range of obstructive sleep apnea. In the latter, the response of the upper airway muscles can be sufficient to open the obstructed upper airway. The occurrence of similar responses in the sternomastoid and diaphragm suggests that a number of receptors are activated by the stimulus. It is possible that these responses could be utilized in developing new treatments for sleep apnea.

The above incidental finding is very interesting IMHO. It truly sounds as if FOT as a measurement technique may actually have side benefits by the way of stimulating upper airway muscles. The question in my own mind is whether the mentioned 3 cm decrease in required pressure can be methodologically attributed to incidental side-benefits of FOT.

And why was snoring surgery such a spoiler? Is this effect related to FOT's airway impedance mapping having been skewed? Is this effect related to post-surgical impairment of nerve endings themselves? In that latter possibility could it prove that nerve endings that were once successfully stimulated via FOT are now impaired and thus rendered less-responsive or unresponsive to FOT?

Indeed, I would absolutely love to hear SAG's and anyone's input regarding FOT. So what if I occasionally put my own FOT in my mouth? .

dsm wrote:? the HI stuff was addressed to you - SWS ?

The FOT stuff was a reply to SAG - I was sure FOT was built into the SOMMNOsmart machines & SAG was questioning its value.

Doug, I wouldn't mind working the entire HI topic even more thoroughly when I get back. Now the FOT topic... Purely a lack of self-control on my part! And that's not a lie. I just found the FOT topic too darn irresistible. So irresistible that I couldn't help explore that topic a little. Seriously, that's an incredibly cool topic IMHO.

But getting back to manufacturer differences in HI. The entire issue needs to be collaboratively worked in my opinion. Regarding differences in residual HI score itself. How many recorded HI deltas between models are due to detection error, how many are due to manufacturer differences in scoring criteria, and how many are due to intentional differences in treatment decisions? Only one of those three are cases of HI inaccuracy. The last two are cases of different residual HI treatment patterns for different manufacturers.

In that latter case only (intentional algorithmic differences in pressure-based treatment decisions) think: 1) occasional "pressure induced hypopneas", 2) "intentionally unchallenged hypopneas", and 3) "unintentional residual hypopneas" ("pressure-strategy unique") as all being possibilities to account for differences in residual HI among different APAP brands. Also consider the fact that radically different algorithmic pressure-based approaches are virtually guaranteed to yield at least some differences in residual treatment patterns and scores in general. Clearly all the machines are not equally accurate. But it's likely IMHO that manufacturer-specific "pressure strategies" can also be expected to yield differences in treatment outcome that are either subtle or sometimes not-so-subtle on a patient-by-patient basis. Here you presumably have multiple factors (including detection inaccuracies) collectively contributing to residual HI variation from one APAP model to the next.

SamCurt wrote:Hey all, I wonder if I should summarize that other test, in which I think is more sophisticated (but still iron lung ) than this, and if yes, post in this thread or a new one.

Sam,

Can you provide any more detailed info from this report ?
http://www.chestjournal.org/cgi/content/full/130/2/312

Many thanks if you can


DSM

dsm:

That's all right, but since I'd have two faculty interviews tomorrow, I can only do it on Thursday.

Sam,

Better'n I can do - many thanks

Thanks

Doug

In our last episode, we were looking at FOT as a potential means to counter the effects of OSA:

Fun Things to Know About Forced Oscillation Technique (FOT)

(FOT-- that still sounds like a potty-mouth word!)

You can end up with a lower pressure:

AutoSet vs. Somnosmart

The Autoset titration pressure (P95) was on average significantly higher than the Somnosmart titration pressure (9.9 ± 2.6 cm H2O vs 7.0 ± 2.5 cm H2O, respectively; p = 0.005). The P50 of the Somnosmart was on average 2.4 ± 1.5 cm H2O lower than the P50 of the Autoset. Moreover, the P50 of the Somnosmart (4.5 ± 0.7 cm H2O) was quite close to the lower pressure limit of the device. Figure 1 shows the Autoset recommended pressure (P95) plotted against the Somnosmart recommended pressure. It can be seen that 12 of 15 patients had higher Autoset than Somnosmart recommended pressures (p = 0.009). Inspection of a Bland and Altman plot (Fig 2 ) displays considerable lack of agreement between the Autoset and the Somnosmart P95s. The bias was calculated at 3.0 cm H2O.

So we're talking about 3.0 cmH2O less. That's a lot.

It could very well be due to FOT (snicker)(sorry, can't help myself):

Effects of High-Frequency Oscillating Pressures on Upper Airway Muscles in Humans

In summary, we have shown that when the upper airway is briefly subjected to a pressure wave of 30 Hz with an amplitude of 4 cmH,O, there is an increase in the activity of some muscles of the upper airway. This occurred in normal subjects and in patients with a range of obstructive sleep apnea. In the latter, the response of the upper airway muscles can be sufficient to open the obstructed upper airway. The occurrence of similar responses in the sternomastoid and diaphragm suggests that a number of receptors are activated by the stimulus. It is possible that these responses could be utilized in developing new treatments for sleep apnea.

And what is an extremely interesting tidbit:

Although it is clear that the stimulus we used is an artificial one, it does mimic in part the high-frequency ressure oscillations that occur in snoring. We suggest that one of the reasons that a snorer can resist complete upper airway obstruction, despite the generation of suction pressures as high as -80 to -100 cmH,O in the upper airway, is that the pressure oscillations and tissue vibration trigger reflex activation of the genioglossus and other upper airway muscles. It is possible that a reduction or loss of sensitivity of such a reflex might be a mechanism by which snoring evolves into sleep apnea.

Rather than let sleeping dogs lie, the Dempsey group (remember them?) got a bunch of sleeping dogs and did a bunch of FOT on them (eeew!!):

Here Doggie!

And they demonstrated:

The findings from this study have implications for understanding the mechanical behaviour of the UA in snorers. Snoring is characterized by high-frequency oscillations of the soft palate, pharyngeal walls, epiglottis and tongue (Liistro et al. 1991) at a similar frequency (30 Hz) to the HFPOs artificially applied to the UA in the present study (Robin, 1968; Liistro et al. 1991). Previous investigators have hypothesized that a reason snorers can resist complete UA obstruction, despite the generation of substantial negative pressures in the UA (Lugaresi et al. 1975) is that the pressure oscillations cause reflex activation of the genioglossus and other UA muscles (Plowman et al. 1990b; Henke & Sullivan, 1993; Brancatisano et al. 1996).

and

The data from the present study also have implications for ventilatory control during high-frequency ventilation (HFV), whereby small tidal volumes are delivered at high frequency.

So the question is, why aren't all the OSA patients on jet ventilators instead of CPAP machines?
SAG

SAG wrote:So the question is, why aren't all the OSA patients on jet ventilators instead of CPAP machines?

Dunno! The question itself may go toward the immense number of factors that can make or break any given medical technology in the market place. And if a new or alternative technology does happen to make it to the market place, the duration of time required for that product to go through the research cycle, followed by a necessary development cycle, and finally onto a product-release date can alter radically on a product-by-product basis. However, one of many possible circumstances affecting a new product's release date (or whether the candidate technology is even pursued) is precisely what competing products, competing technologies, and even established in-house alternatives happen to be up to in the marketplace.

At least that is my initial thought as an outsider to the industry of sleep science itself. However, what do SAG and others here think regarding why patients may not be presently using pressure-oscillation treatment techniques? Is this perhaps another case of medically or financially unfeasible technology? Or is this a potentially viable technology that simply has yet to make it through the scientific and corporate gauntlets (that can be both unique and highly circumstance driven)?

As an avid researcher of medical research itself, our SAG just may have the answer for us! But in modern sleep science there are undoubtedly many more questions than answers. Talk about a challenging field...