Doing my own sleep study - surprising results

[ozij]

When you read drastically different AI and HI numbers from two different AutoPAPs, you are less likely reading an AI/HI measurement error in any one machine. Rather, you are more likely reading that one machine's algorithm happens to be treating you better.

I am not clear how it can be good that there is such a discrepancy between reports from 2 main brands of AUTO. Saying it is simply a difference in how 2 different manufacturers interpret results seems a very weak argument.

DSM,
I don't think anyone was saying the difference is good. "Saying it is simply a difference in how 2 Saying it is simply a difference in how 2 different manufacturers interpret results" was not a an argument for one thing or the other, it was an explanation of a fact of life. Hypopneas have different definitions -
http://www.aasmnet.org/PDF/hypopneaposition.pdf

Apnea, characterized by a cessation of airflow for 10 seconds
or more, is easily recognized and agreed upon. In contrast, the
definition of hypopnea, characterized by a reduction without cessation
in airflow or effort, is not consistent.2 A duration criteria
of at least 10 seconds is generally agreed upon in adults. More
variable definition features include the degree of airflow or respiratory
effort reduction, inclusion and degree of oxygen desaturation,
and inclusion of arousal from sleep.2 To compound the
problem further, each of these is dependent on the method of
detection.

- and it's the definition that guides the software. Software is what helps the APAP decide what it is identifying, and how to react to it.

That is the reason that for some of us, one APAP gives much better treatment than another. Your Remstar may be no good for you, and perfect for someone else, and your Sprit may be wonderful for you, and dreadful for someone else.

O.

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): Arousal, Hypopnea, auto, APAP

[neversleeps]

Saying it is simply a difference in how 2 different manufacturers interpret results seems a very weak argument.

I don't see it as an argument, but rather a statement of the facts.

The seemingly different data reported by both machines is likely accurate in both cases. Both machines have fairly good sensitivity and specificity where apnea/hypopnea detection is concerned. Recall that you are reading "left over" apneas and hypopneas in both cases. And, also recall that both machines do not react merely in responsive mode. While both machines necessarily respond to obstructive events, both machines also leverage proactive treatment. The Remstar Auto, for instance, will intentionally attempt to induce an inconsequential flow limitation toward measuring airflow deltas, and then, in turn, extrapolating the next target pressure.

When you read drastically different AI and HI numbers from two different AutoPAPs, you are less likely reading an AI/HI measurement error in any one machine. Rather, you are more likely reading that one machine's algorithm happens to be treating you better. After all, in both cases you are reading "left over" apneas/hypopneas after the machine's responsive algothithmic techniques coupled with that same machine's proactive techniques takes it's best shot at treating your particular breathing patterns.

Different AI and HI does not necessarily mean either machine is measuring either obstructive sleep event incorrectly.

One other question deals with either manufacturer's definition or cut-off point between scoring an obstructive sleep event as a hypopnea versus scoring that same obstructive sleep event as an apnea. While the Remstar Auto uses the more sensitive flow sensor of those two machines, either sensor type is more than capable of measuring flow limitation well enough to distinguish apnea versus hypopnea. However, the apnea/hypopnea definitions or cut-off point may actually be different between these two models by design.

AutoPAP manufacturers do not necessarily agree on how to score sleep events, and there are subtle differences in scoring criteria. Anyone whose sleep events tend to consistently border between hypopnea and apnea may return drastically different AI-to-HI ratios between two different manufacturers for this reason.

So let me see if I've got this right.

The different results from machine to machine are not errors. The results are correct for the algorithms and scoring criteria used by the particular manufacturer.

An individual's particular breathing patterns could conceivably be better interpreted by one manufacturer's algorithms over another.

[FL andy]

So let me see if I've got this right.

The different results from machine to machine are not errors. The results are correct for the algorithms and scoring criteria used by the particular manufacturer.

An individual's particular breathing patterns could conceivably be better interpreted by one manufacturer's algorithms over another.

I hope others, more knowledgable than I am, respond to this summary. The way neversleeps states this above - I agree that is the way I read this thread. IF this is true, it sure seems like hit and miss whenever one orders any Auto-Titration xpap machine. ESPECIALLY when one has some medical (is it limited to "medical"?) condition that tends to alter their own breathing pattern.

Seems to me the new Respironics Auto BiPap eliminates this problem of hit or miss on breathing patterns - based on my limited knowledge. Good for Bi Level needs, but regular Auto seems to be somewhat risky unless one knows their own breathing pattern agrees with their machine's algorithms.

Andy

[Guest]

AutoPAPs are hit or miss. But absolutely nothing in this thread should lead us to believe that they miss with intolerable frequency across the SDB patient population. DSM's experience cannot be generalized. Aspirin is hit or miss and so is acetaminophen. When one person yields an adverse reaction to aspirin and nothing but benefits from acetaminophen, we are not so tempted to generalize that aspirin does not work well for most----or that acetaminophen
is the superior solution for everyone.

AutoPAP is, indeed, hit or miss---but with a more than acceptable rate in my opinion. AutoPAP actually "hits" correctly for quite a few patients where CPAP or BiLevel happens to "miss". Any prescription or therapy known to mankind is truly a "hit or miss" proposition when you think about it.

Neversleeps, I think that statement nicely summarizes the likelihood of the situation. However, even therapeutic machines and diagnostic tests that have reasonable sensitivity and specificity ratings still manage to "botch" that which they aim to detect. My own take is that across a statistically robust patient population: 1) some patients are likely to be treated better by any given APAP's algorithm (versus yet some other algorithm---with satisfactory detection yet a less-than-optimal algorithmic pressure-response strategy occurring in that latter case), 2) far fewer patients are likely to experience apnea any hypopnea misdetection, while 3) far more patients than the previous two exceptions are likely to fare just fine on any arbitrarily selected AutoPAP algorithm.

As a general rule apnea and hypopnea detection are not inclined to be algorithmic shortcomings so much as are: 1) precursor detection (typical snore and flow limitations), 2) uncharacteristic/atypical events that can mask or otherwise skew characteristic sleep events (asthma, hypoventilation, hyperventilation, non-snore accoustic vibrations, etc), and 3) overall pressure response strategy. As an example of this latter statement, a relatively "fast-changing" pressure response strategy is great for some while a slower pressure rate of change is ideal for others. One APAP's algorithm is ideal for one person yet dreadful for yet another person as Ozi pointed out. Again, this non-panacea type statement is true of any medication or therapy, however.

Perry, thank you for coming back to the boards!

[neversleeps]

So let me see if I've got this right. (I may be slow, but I'm persistent!)

While the different APAP manufacturers use different algorithms and scoring criteria, the majority of people will do well on any manufacturer's APAP.

I guess, as in most things, we are made aware of the exception more often than the rule. We don't hear from the thousands of well-treated satisfied individuals (and why would we; they're off happily being well-treated and satisfied). We hear for those who find some anomaly and are seeking answers. This exception is brought to the forefront and we all take a look at it; trying to figure it out. I must say, from the perspective of a neophyte (that would be me), this information is absolutely invaluable and completely engrossing. What a fantastic education this thread has been!!

[Andy]

AutoPAPs are hit or miss. But absolutely nothing in this thread should lead us to believe that they miss with intolerable frequency across the SDB patient population. DSM's experience cannot be generalized. Aspirin is hit or miss and so is acetaminophen. When one person yields an adverse reaction to aspirin and nothing but benefits from acetaminophen, we are not so tempted to generalize that aspirin does not work well for most----or that acetaminophen
is the superior solution for everyone.

AutoPAP is, indeed, hit or miss---but with a more than acceptable rate in my opinion. AutoPAP actually "hits" correctly for quite a few patients where CPAP or BiLevel happens to "miss". Any prescription or therapy known to mankind is truly a "hit or miss" proposition when you think about it. ...

One APAP's algorithm is ideal for one person yet dreadful for yet another person as Ozi pointed out. Again, this non-panacea type statement is true of any medication or therapy, however.

1. Would "Guest " please state his name and his qualifications. If Guest is a registered, previous poster, then we can judge whether he or she is qualified to make those definitive, authoritative statements. Are you real or not? If you do not want to reveal your name and qualifications, then some of us would like to know *why* you post.

2. Comparing APAP equipment to aspirin or Rx medication as being comparable examples, and implying that hit or miss treatment with APAP is not perfect but OK and acceptable - is rubbish!

Once a patient gets an APAP, (a) his/her insurance is not likely to pay for a second machine for a long, long time. It is nothing like getting a second doctor's opinion paid for. And (b) to get another APAP "to see if *that* one is better for him/her, is expensive and it comes out of his own pocket - a bit more money than buying a different pain reliever or getting another Rx medication.

3. You said: "DSM's experience cannot be generalized". You want to know the most important reason why no one else compared their experience with another mfg's machine? APAPs cost so much money most people CANNOT AFFORD TO BUY TWO OR THREE DIFFERENT MACHINES TO FIND THE *ONE* THAT WORKS BEST FOR THEM, AS INDIVIDUALS.


Guest, either state the reasons why you want annyonmity or give us better, credible reasons to support your authoritative statement. Please.

Neversleeps and DMS, my good friends, thank you for your posts.

Andy

[wading thru the muck!]

Andy,

My guess is that this particular "guest" is a knowledgable and respected former regular to this forum and others. There is at least one that I can think of that may choose to chime in occationaly without the fanfair of an official return.

[Guest]

1. Would "Guest " please state his name and his qualifications. If Guest is a registered, previous poster, then we can judge whether he or she is qualified to make those definitive, authoritative statements. Are you real or not? If you do not want to reveal your name and qualifications, then some of us would like to know *why* you post.

The administrators of this board have intentionally allowed for guest posting. True any guest post on any forum comes with neither certification nor validation. However, no registered post comes with any real credence. In my opinion any any message on any message board is but an idea or opinion and comes with neither credential nor credence until verified as fact---from registered or unregistered posters. The only information I ever see posted on this or any message board are: 1) opinions, 2) ideas, and 3) unverified fact. Again, registration is not required here. You are welcome to ignore my posts or any guest posts. [/quote]

2. Comparing APAP equipment to aspirin or Rx medication as being comparable examples, and implying that hit or miss treatment with APAP is not perfect but OK and acceptable - is rubbish!

My point was that APAP does not have a 100% efficacy rate. Neither does aspirin. Nor does any RX. APAP is an Rx with less-than-perfect efficacy and it is, indeed, "acceptable" according to the FDA. To my knowledge any other convening government agency throughout the world that has ever reviewed AutoPAP has also deemed it "acceptable".

Once a patient gets an APAP, (a) his/her insurance is not likely to pay for a second machine for a long, long time. It is nothing like getting a second doctor's opinion paid for. And (b) to get another APAP "to see if *that* one is better for him/her, is expensive and it comes out of his own pocket - a bit more money than buying a different pain reliever or getting another Rx medication.

Yes, APAP is much more expensive than aspirin. That still doesn't change the fact that APAP, CPAP, BiPAP, anyPAP may or may not work for any given patient---just like aspirin. And it still does not change the fact that any RX prescription--- APAP/CPAP/BiPAP included---is deemed "acceptable" by the FDA and similar agencies despite less-than-perfect efficacy.

3. You said: "DSM's experience cannot be generalized". You want to know the most important reason why no one else compared their experience with another mfg's machine? APAPs cost so much money most people CANNOT AFFORD TO BUY TWO OR THREE DIFFERENT MACHINES TO FIND THE *ONE* THAT WORKS BEST FOR THEM, AS INDIVIDUALS.

I'm not sure why this particular rant is aimed at me. I happen to think that most people who can tolerate PAP therapy (fifty-some percent?) are likely to do well with any APAP algorithm. However, I also happen to think that as with and RX, some patients will not take to this particular RX well.

Guest, either state the reasons why you want annyonmity or give us better, credible reasons to support your authoritative statement. Please.

Most mildly reclusive eccentrics prefer anonymity don't they? Sorry if it annoys you. Please kindly disregard any posts that don't meet your personal criteria. Please.

[FL andy]

The administrators of this board have intentionally allowed for guest posting.

Yes, you are correct. I did indeed forget that.

I apologize to you and to all others.

Andy

[dsm]

Backgrounder:
Regarding my own self-therapy, I am someone doomed to experimenting & evaluating and also doomed in that I can at my age (all kids long gone from home) afford to spend some money on such experimenting. On the other hand, I regard any expenditure that will prolong my good health when there are obvious problems, as being money well spent. Also I have been extraordinarily lucky in obtaining some great gear at low prices (I certainly could not have done the various experiments using gear sold at retail prices).

My most recent aquisition has been a N-200 Nellcor pulse oximeter. I am really keen to measure what that tells me against what the various xPAPs I have are telling me.

I would like to try to come up with a clear relationship between blood-oxygen sat at night vs feelings of energy during the day. The night time blood oxygen sat is going to depend on the xPAP devices being used *and* as good an understanding as possible of what my own breathing problems are.

I did a sleep study 15 years ago and was advised I did not have OSA
Before the current sleep study (mid 2005) my respiratory physician said
after an examination, that he was expecting me to agin show up negative.
The results of the 2005 study however gave me an AHI of 40.

Some conclusions I am coming to include:
1) I do have bouts of OSA and the sleep study picked that up accurately
2) I do slow breathing & also will stop, unrelated to any OSA event.

Of 2 above I am not really sure what this is about. Is it 'centrals' ? - I don't know. I do know that in my younger years I spent hours practising slowing my breathing as a form of meditation (25 years of practice). MY wife is pretty well convinced that it has become a habit that I enter into at start of sleep & just before waking. The spirit results tended to echo this pattern in a way that led me to believe it was accurately recognizing what was happening.

My wife says that if I go on my back it is almost a certainty that as soon as I start to 'sleep' I slow & stop breathing & she then nudges me - she usually asks me to sleep on my side as this slowing of breathing doesn't seem to occur. She has also said that she gets woken by the CPAP or APAP appearing to strain (I have ceased breathing it the machine is not moving much air other than through the masks co2 leak holes). When I practised the process of reducing breathing I always did it on my back. I was also able to quickly lower my blood pressure using this meditation technique.

I have been seriously wondering for some time if this is what I have been doing and if this has been lowering my BO sat at night & combined with getting older & being around 15 stone has combined along with the other Apneas, to create my own set of symptoms.

But, my sleep clinic folk assure me that breathing while asleep is not under conscious control & they doubt that any meditation exercises I may have once done could be the cause of my symptoms.

The thing that really caught my attention was the fact that for about 2.5 months CPAP seemed to have transformed my energy & alertness levels, then the effects began to wane & again my wife would say, you are slowing & stopping your breathing again!. It seemed to me I was just adapting to the xPAP cms and reverting to a prior sleeping pattern.

Many folk here have commented on the experience we usually call 'cpapers high', I believe I know exactly what that phenomenon is. But when the sleepiness & lack of alertness began to return despite having the best APAPs I began to wonder if I had only been partially correctly diagnosed & that I might have some other effect related to the slowing & stopping breathing.

That was when, after deciding the Spirit data was telling me the best match to my wife's observations, that I should try a BiLevel esp one with Timed or Assisted control. That is where I am now & all appears to be good again ...
for now

I do believe that different brands of APAPs have an obligation to report some form of consistency in the data they provide. I tend to feel any statement that implies that these different brands can produce quite different HI & AI values and that this is quite ok, is 'obfusticating' a problem. The statements re the discrepancies read more like expert political opinion than meaningful measurable statements of fact
Sorry to seem to be so harsh on the comments made re differing readouts but that was how it came across to me.

I will be going back to my sleep clinic and to the doctor managing my case (a respiratory specialist doctor). I expect to get another prescription from him for a BiLevel (even though I already have it).


I am also going to see if an undergrad or post grad student from our local uni & who works with these people, is willing to use me as a guinea pig to either prove or disprove my belief that I was only partially diagnosed correctly for what ever condition it is that I actually have.

Summary:
One point I really feel needs hammering is that almost everyone going on xPAP (esp at higher than average cms settings), wants to get some relief from the exhalation effort, this is especially so for those of us who have erratic congestion in our nasal passages. This relief can be by way of
CPAP with Cflex, APAP, and best of all IMHO, BiLevel. Some are able to live with straight CPAP at high cms but I suspect they have good nasal passages. On top of that there is the issue of maintaining a good flow of clean air (min co2) into the lungs during sleep. Timed & Assisted BiLevels appear to help there.

Cheers DSM

[dsm]

<snip>
That is the reason that for some of us, one APAP gives much better treatment than another. Your Remstar may be no good for you, and perfect for someone else, and your Sprit may be wonderful for you, and dreadful for someone else.

Ozij,

I am not saying that the RemStar wasn't giving as good 'therapy' results to me as the Spirit. I am saying that the stats produced differed in a way that left me losing confidence in the RemStar printouts.

The RemStar has been my fav xPAP machine until the PB330. In terms of its sophistication, its robustness, its quietness, its build, its quality, it is tops!. But, the stats were at odds with personal observation over a long period. For me this became the problem.

Also I agree that in terms of therapy, one AUTO can well provide better *therapy* to a user than another.

These are 2 very different issues.

Cheers

DSM

[dsm]

<snip>

The different results from machine to machine are not errors. The results are correct for the algorithms and scoring criteria used by the particular manufacturer.

I used the word discrepancies - never said 'errors' ?

But if someone says that 1 machine can report say 4x HI & 1x AI & another can
report this as 1x AI & 4x HI and than this is perfectly ok, just doesn't cut ice for me.

I believe the machines need to be better aligned as to their meanings and measurements of these datum.

Cheers

DSM

[rested gal]

My guess is that this particular "guest" is a knowledgable and respected former regular to this forum and others. There is at least one that I can think of that may choose to chime in occationaly without the fanfair of an official return.

I think you're absolutely right, Wader. I'm always glad to see thoughtful posts like this Guest's have been, whatever "name" a person chooses to use....or not use.

[neversleeps]

I do believe that different brands of APAPs have an obligation to report some form of consistency in the data they provide. I tend to feel any statement that implies that these different brands can produce quite different HI & AI values and that this is quite ok, is 'obfusticating' a problem.

I believe I understand what you're saying, and therein lies the crux of the issue being debated. The fact of the matter is, as I understand it, the different manufacturers do not have any such obligation. We may wish they did, but they do not. If they did, would they not then ultimately be reduced to using identical algorithms and identical scoring criteria?

I personally don't view the acceptance of different results from manufacturer to manufacturer as obfuscating a problem, but rather demonstrating the point that they are, in fact, using different methods in obtaining those results.

[Guest]

DSM, a couple thoughts regarding some of your comments and observations:

My most recent aquisition has been a N-200 Nellcor pulse oximeter. I am really keen to measure what that tells me against what the various xPAPs I have are telling me.

I would like to try to come up with a clear relationship between blood-oxygen sat at night vs feelings of energy during the day. The night time blood oxygen sat is going to depend on the xPAP devices being used *and* as good an understanding as possible of what my own breathing problems are.

Either desats or cortical arousals may result in poor sleep. You may discover that all your PAP machines result in no significant SpO2 drops. Yet if any one machine happens to fail at keeping cortical arousals at bay, then your subjective assessment, or better yet a PSG, is the better tool. It's great that you're measuring your SpO2 since you may discover a shortcoming in one or more of your xPAP machines relative to your own breathing pattern.

Some conclusions I am coming to include:
1) I do have bouts of OSA and the sleep study picked that up accurately
2) I do slow breathing & also will stop, unrelated to any OSA event.

Of 2 above I am not really sure what this is about. Is it 'centrals' ? - ...

If you had a bout of central apneas during your PSG, the sleep study would have picked that up accurately via a respiratory effort belt. If you had bouts of central hypopneas during your PSG the standard sleep study would not have differentiated those central hypopneas from obstructive hypopneas for a typical lack of technical means---a central hypopnea registers some respiratory effort during the PSG making hypopnea differentiation extremely difficult (but not impossible).

Of 2 above I am not really sure what this is about. Is it 'centrals' ? - I don't know. I do know that in my younger years I spent hours practising slowing my breathing as a form of meditation (25 years of practice). MY wife is pretty well convinced that it has become a habit that I enter into at start of sleep & just before waking...
...But, my sleep clinic folk assure me that breathing while asleep is not under conscious control & they doubt that any meditation exercises I may have once done could be the cause of my symptoms.

The bold words in your sentence I have emphasized are key in my own mind. Truth be told conscious breathing directives from the brain do influence how we breathe during: 1) sleep onset, 2) intervening points of wakefulness throughout the night, and 3) during those final emergence-related sleep stage shifts as we finally wake at the end of a night's slumber. In the following reference by Peter J. Hauri, Ph. D (of the Mayo Clinic Sleep Disorders Center) anxiety-induced hyperventilation during these three non-slow-wave sleep stages are documented:
http://www.talkaboutsleep.com/sleep-dis ... manual.htm

The anxiety-based hyperventilation documented in the link above by Dr. Hauri actually induce central apneas during those three "light" levels of sleep
mentioned above. Not that you experience anxiety-induced hyperventilation
during sleep onset or points of wakefulness. Merely to demonstrate that that breathing during those "light" stages of sleep may not always be purely autonomic. If the conscious mind's anxiety can influence breathing during these sleep stages, so might the conscious mind's well-engrained tendency to breathe very relaxed influence those same stages of near wakefulness.

tend to feel any statement that implies that these different brands can produce quite different HI & AI values and that this is quite ok, is 'obfusticating' a problem. The statements re the discrepancies read more like expert political opinion than meaningful measurable statements of fact..

I think you just may be missing the point of what some here and on other message boards have been saying: different APAP algorithms necessarily entail different results across the entire patient population---just as different pain-relieving pharmaceuticals also yield different results across the patient population. Substitute the words "APAP-brand A" for the word "acetaminophen" and "APAP-brand B" for "aspirin" and a statement similar to yours reads: "The fact that aspirin worked as advertised for me but acetaminophen did not, implies there's a serious problem here---because these two pain relievers should have yielded the same exact pain indices for me". For you there may be a shortcoming with both APAP-brand-A and acetaminophen, but that cannot be generalized to the patient population in either case. The RemStar Auto algorithm and the Resmed Spirit algorithms are as different in algorithms as aspirin and acetaminophen are in chemical ingredients----yet each pairs strives for the same objective.

Your statement about discrepancies in "readouts" alludes to APAPs as if they were test or measurement equipment rather than being primarily therapeutic machines. APAPs aim to treat, then measure those treatment results. You should, indeed, expect different brand test meters to yield the same "readouts". But you should never expect different treatment chemicals or algorithms to yield identical results or datum across
a patient population. When is this ever expected of equivalent alternatives in any other therapy or medicine? Would you expect Prilosec to yield the same exact residual acid-reflux index as Protonix for every GERD patient that takes both?

I believe the machines need to be better aligned as to their meanings and measurements of these datum.

Again, you seem to attribute this discrepancy to a measurement error---as if you expected to compare apples with apples. Set both machines at the same fixed pressure, then compare AI and HI. Only then will you be comparing apples with apples regarding measurement datum. Until then you are measuring the results of two different algorithms, and datum-wise ("readout"-wise) that is tantamount to comparing apples with oranges. Again, expecting to measure the same AI and HI indices out of these two different algorithms is like expecting to measure the same "pain indices" after taking aspirin, then acetaminophen.

I don't see any of your words as harsh and I hope you don't see any harshness in my own words. Interesting topic.