CPAPtalk.com

I am scratching my head as to how my REMStar Auto with C-Flex is reacting to my sleep patterns. I am basically hypopnea exclusive, but the apap does not appear to be reacting to hypopneas. Let me illustrate with a couple of charts, in which I have combined Encore Pro's apap pressure chart with the events chart.

In the first case I have a night with basically no hypopnea activity until after 7 hours into the session. Yet the pressure rose and ended up with a 90 percentile pressure of 13 cm. The AHI for the night was 0.7. There appears to be little reaction (if any) to the hyponeas.



Last night we have a different story. I had hypopneas distributed throughout the night, with an AHI of 2.7. But the pressure did not seem to react strongly to the hypopneas, and it stayed much lower with a 90 percentile pressure of 8.9 cm.



So I ask myself: what is going on? Is it the snores that drive the pressure up? Certainly the first night had many more snores. ( In a recent set of snore experiments that I posted here I showed that the Auto definitely responded to snores) Are the Auto's algorithms much more sensitive to snores than more severe restrictions? Was there a difference in my breathing patterns on the first night that allowed the so-called "pro-active" algorithms to predict hypopneas ant take action to prevent them? I have many more nights in which the Auto does not seem to respond to hypopneas.

Maybe there is a simple explanation. I would truly love to see some more details on Respironics event detection and "pro-active" algorithms. It's easy for a manufacturer to make claims of sophisticated features in a system, and then hide behind "proprietary" reasons for not disclosing any actual performance details. We, the end users, are left to take on faith the claims of the manufacturer. So I guess my ultimate question is: does the Auto really do what it calims to do?

derek

My experience has been that the REMstar auto responds to snores more so than the Puritan Bennett 420E I have. I have also read posts by several people who have increased their bottom pressure in an attempt to eliminate hypopnias, which leads me to believe they felt their hypopnias where going untreated. I have wanted to take a closer look at this myself, but have been dealing with a cold for a couple of weeks and due to work (over) commitments I have had a very irratic sleep schedule. I intend to take a serious look at how therapy with the REMstar auto compares to therapy with the PB420E.

I don't have much confidence in the snore index. My opinion is that it is not very accurate.

I am not a snorer, and have noted a correlation between a recorded snore and when the pressure blows my mouth open.

Also, a couple of nights ago, it took me about two hours to fall asleep, but a snore was recorded during that time.

I know that I don't snore while I'm awake, so my theory is that I must have made a grunt or a groan as I moved in bed which was recorded as a snore.

You may want to search the discussions here as I remember someone performed tests on the Remstar's ability to detect snores.

Regards

Hugh,

That was derek, the originator of this thread, that did the snore detection experiment.

I agree with derek, it sure would be useful to have some independant analysis of the ability of these machines to do what they claim. This may not be so much of a concern for users with moderate to severe apnea, but for those of us with mild apnea and the desire to tweak our treatment till it's optimum, the finer details about the functionality of each algorithm would be useful.

Hi all,

To get a closer idea of your sleep, a pulse oximeter would be helpfull.

If you have restless leg syndrome, there is nothing an autopap can do for you. I have actually seen that happen.

If you have OSA and RSL, you will take care of the OSA, but still have arrousals from the RSL. That will wreck your sleep just as much as OSA.

I am a firm believer in the PSG. It looks at so much information.

I have had patient's with PVC's that were first detected during the PSG. It could be a live saver.

Regards,

Ted

PS. Check out my new Avatar

Wader,
Here's a scatter chart showing the nightly APAP 90 percentile pressure vs. total snores for the night. (The chart using the SI is very similar). I have plotted a trend line that shows a small positive slope, but as you can see the correlation coefficient (R^2) is very small.



Hard to draw any real conclusions here.

I have a similar scatter chart showing the relationship between 90 percentile pressure and HI for the night. Again it shows a slight trend but a very low value of the correlation coefficient (R^2). No real conclusion there either.

derek

Ted,
MY PSG showed no problem with O2 desaturation. I was diagnosed with Restless Legs, but Neurontin has zonked that completely and my wife marvels at just how tranquil I now am while sleeping.

derek
ps -I am surprised that your avatar does not show an Aura! By the way it looks as if you have a severe mouth leak problem!

Ted wrote:I am a firm believer in the PSG. It looks at so much information.

I am in no way being critical of the diagnostic value of the PSG. I just feel that the high cost combined with the insistance by CYA Docs to refuse to prescribe therapy without one, is forcing many to go untreated.

Derek,

I think correlation between pressure and snore/hypopnia has to be analysed by looking at the pressure response relative to snore/hypopnia frequency increases/decreases. I think where your analysis is missing this is that you are using snore/hypopnia and pressure indicies which are averaged over the course of each sleep period. I don't think the Encore pro provides the data to make a relevant observation other than looking at areas of increased snore/hypopnia frequency and then comparing the pressure response.

I'm glad you're thinking about these things! I'm just trying to keep my head above water right now.

I wonder if one of the manufacturers could incorporate a pulse oxymeter into a nasal mask. Make the mask out of an opaque material and shoot the light through the nose.

Wader,
I just happened to have that data in a spreadsheet - so I threw the plot together. Not pretending that it is anything more than it is

Now I thought about this as I drove in to work after posting the above stuff. Here's a conjecture as to what Respironics might be doing:
The pressure is incremented strongly only when the total rate of all events combined (FL, OA, H, S) exceeds some minimum value per hour. An HI of 2.7 means a hypopnea only once every 23 minutes or so on average. The interval between events may be just too large to significantly bump the pressure.

On this basis you can see why the first chart would react the way it did, rising strongly on the intervals of dense snores, but the second one hardly reacted at all. Just a thought...

The implication would be that those of us fortunate to have a low AHI below such a threshold need to carefully tailor the minimum pressure to treat the vast majority or our events, because the REMStar Auto will not raise the pressure for us to a satisfactory treatment level. If it is active snoring that does the work it almost sounds like we need to rethink CPAP
derek

Derek,

Not suggesting that you were "pretending that it is anything more than it is." Just suggesting my opinion as to why the data you showed didn't support the assumption of a correlation between snore/hyponia and pressure increase.

I do concur with you conjecture that the frequesncy of the events/combined events is what the REMstar auto is reacting on. I also agree that those of us with mild OSA/H and a low pressure requirement may need to raise our lower pressure to prevent our less frequent events. Unfortunately this eliminates the benefit of keeping the pressure below our titrated level in between events, but then again, those of us with low pressure requirements have not had to face the high pressures that some do. The auto will still provide the ability for the machine to prevent obstuctions that require pressure higher than our titrated pressure.

Any chance your career will branch off into the research and development of auto-apap algorithms?

Rhetorically speaking: what's represented in a snore tick mark on the Encore reports? Accoustical magnitude? Duration? Neither? I have noticed when the snore tick marks show dark red, the RemStar Auto's pressure response is greater. As it stands, to simply compare each snore tick mark as if the RemStar Auto assigns equal trigger criteria may very well be in error----since trigger criteria may very well entail more than simple frequency of snore.

I agree it would be VERY nice to get a description from Respironcs on trigger criteria for the RemStar Auto!

Derek, your previous snore experiments showed that the RemStar Auto pressure response really does vary according to a snore's accoustical characteristics introduced to the machine's sensors.

Wader,
Funny you should say that.... wink, wink, say no more... (to quote Monty Python)
derek

As soon as I figure out the table structure on the SQL server, I will give ya some numbers to crunch Derek.... Haven't had enough time to look at it...