Remstar Auto Reported AHI -vs- Mask Models

Hello again hose heads, hope everyone had a great New Year!

I just received a brand new Respironics Remstar Auto w/ A-Flex (DS510SH). With all the cool data you can get out of it, I
wanted to compare the machine reported AHI for the two masks I have. I was expecting the machine reported AHI for both masks to be in the same ballpark, but it turns out they weren't even on the same planet. Naturally, this raised more questions than it answered, what else is new lol ...

For both nights, I set my machine to CPAP mode with a pressure of 4 and a C-Flex of 3. I left the humidifier and ramp off.

Night #1 with the Zzz-Mask Full Face Mask
http://www.michaeltolchard.com/sleep-ap ... 090102.pdf
AHI = 51.7

Night #2 with the Hans Rudolph 7600 Series V2 Full Face Mask
http://www.michaeltolchard.com/sleep-ap ... 090103.pdf
AHI = 3.4

Question #1: What could cause such a wide AHI range (51.7 -vs- 3.4)?

Question #2: Is my very high Vibratory Snore Index screwing up the AHI numbers?

Question #3: I noticed last night, using C-Flex, as I was drifting off, that when I snore on exhale as I tend to do it can cause the pressure to do a harmonic oscillation type thing. I assume the C-Flex function is not on a fixed curve and in fact varies based on pressure feedback. Does anybody know if this is the case? Could this be causing my high VS numbers, and in turn screwing up my AHI numbers somehow?

Any input would be greatly appreciated!

Michael

Hello Michael,

I know it's just soooo tempting to try one thing for a night and try to compare it to another on the next night. However, there seems to be a "settling in" process for new settings and it's generally recommended to change settings no less frequently than every 2 weeks.

As for the difference in your results, one of the things I noticed is that your leak rate seems high for such a low pressure. That can throw off your AHI results. IMHO, I'd try to first get leaks managed before experimenting to reduce AHI.

Mindy

Thanks Mindy. Regarding high leak rates, do they tend to distort the index numbers up or down? I would assume downward, but am not sure ...

Also, the documentation that came with the masks tell me he exhaust flow rate for the mask. In a perfect world, would the leak rate reported by the machine be equal to the exhaust flow rate for the mask?

Peaktagger wrote:Thanks Mindy. Regarding high leak rates, do they tend to distort the index numbers up or down? I would assume downward, but am not sure ...

Also, the documentation that came with the masks tell me he exhaust flow rate for the mask. In a perfect world, would the leak rate reported by the machine be equal to the exhaust flow rate for the mask?

If the leak becomes a "large leak", then the machine apparently stops recording AHI. However, if the leak rate is noticeably higher than it should be, the machine tries to compensate --- but it looks like your machine is in CPAP mode and a pressure of 4 in straight cpap mode seems awfully low - similar to no therapy at all.

I believe you are correct that the exhaust flow rate for the mask would equal the machine-reported leak rate if you had a perfectly sealed mask

If you have an APAP, you might try starting with a range - such as 4-10. If a straight cpap, I'd be tempted to try first around 7. Use ramp if you need to for ease of getting to sleep.

If you look on the Respironics web site, you can find explanations of C-Flex and A-flex showing the curves.

Mindy

Mindy,

Yes, I was trying to mimic no therapy which is why it was set to 4. I was willing to put up with 2 poor nights of sleep at low pressure (napped later at therapeutic pressures in APAP mode to get some more rest) to see what the machine thought my baseline AHI was. I was pretty blown away by the vast range lol, and as we all want a low AHI, I want to be able to trust the AHI number this machine is spitting out, and at this point, I don't ...

Michael

Peaktagger wrote:Mindy,

Yes, I was trying to mimic no therapy which is why it was set to 4. I was willing to put up with 2 poor nights of sleep at low pressure (napped later at therapeutic pressures in APAP mode to get some more rest) to see what the machine thought my baseline AHI was. I was pretty blown away by the vast range lol, and as we all want a low AHI, I want to be able to trust the AHI number this machine is spitting out, and at this point, I don't ...

Michael

Michael,

I'm not sure it's really valid to try to mimic "no therapy". The difference between the masks, besides leak rates, shouldn't make that much difference in AHI. I generally see a difference of 1-3 in my AHI from day-to-day even with a stable leak rate. I'd stick with one mask and work on the rest of it.

Mindy

Peaktagger wrote:Mindy,

Yes, I was trying to mimic no therapy which is why it was set to 4. I was willing to put up with 2 poor nights of sleep at low pressure (napped later at therapeutic pressures in APAP mode to get some more rest) to see what the machine thought my baseline AHI was. I was pretty blown away by the vast range lol, and as we all want a low AHI, I want to be able to trust the AHI number this machine is spitting out, and at this point, I don't ...

Michael

Yes 4cm would be like no therapy. Keep in mind the machine senses changes it flow & pressure to be able to know whether we are breathing or not, so leaks have a HUGE impact on the sensors. A large leak can prevent anything from being sensed at all.

mindy wrote:
Michael,

I'm not sure it's really valid to try to mimic "no therapy". The difference between the masks, besides leak rates, shouldn't make that much difference in AHI. I generally see a difference of 1-3 in my AHI from day-to-day even with a stable leak rate. I'd stick with one mask and work on the rest of it.

Mindy

Thanks for your input Mindy ... curiosity led me to try this, and I didn't get the expected result, so am really glad I tried it, and it is a great learning experience to figure out why. I was expecting there to be little difference in masks with regards to AHI over time, along the lines of what you reported, but was also expecting some reasonable difference as this was only a 2 night test. That's the problem, there is a huge difference that I'm trying to explain ... and as you said in your experience you see only a difference of 1-3, and I think I should be seeing something similar. As I said at this point I can't trust my numbers, at least the index ones, and I want to be able to have some faith in my equipment ...

Michael

GumbyCT wrote:Yes 4cm would be like no therapy. Keep in mind the machine senses changes it flow & pressure to be able to know whether we are breathing or not, so leaks have a HUGE impact on the sensors. A large leak can prevent anything from being sensed at all.

Thanks for your response Gumby ... I didn't have any 'Large Leaks' as reported by Encore Viewer Pro either night or with either mask, so I will assume at no time did the machine outright miss any events. That leads me to wonder, if during sizable leaks that are not classified as large leaks, does the ability to detect events diminish proportionality as the quantity and/or duration and/or size of leaks increase?

Also, does the classification of events occur at the machine, and the number of events are reported by the smartcard to encore viewer, or does all the raw data come across the smartcard and is interpreted by the encore viewer software? Yes, I have been a programmer for 25 years, so I couldn't resist asking lol ...

Michael

The raw data comes across and is interpreted by EncoreViewer. I have EncorePro and my data is saved in a SQL Server database (when I use my Respironics machine). Data can be exported to Excel in XML format. Note that it's quite a large amount of data.

Although you didn't have the "large leaks" that result in cessation of recording, you did have some rather hefty leaks over 100. Also, the pattern of your leaks was all over the place. The pattern for your Hans Rudolph was better with at least some straight "zipper" segments. When the leaks look like your ZZZ mask pattern, you're probably getting way less than 4 cm and perhaps that could affect your AHI. You are also showing mucho snoring although whether or how much that affected your AHI I don't know.

If I were you, I'd get the leaks under control and then experiment with pressures - pretty much the reverse of the way you are doing it.

Mindy

At 4cm I don't think it would take much of a leak to make a difference. Keep in mind as the pressure increases - it becomes more difficult to keep a seal.

mindy wrote:The raw data comes across and is interpreted by EncoreViewer. I have EncorePro and my data is saved in a SQL Server database (when I use my Respironics machine). Data can be exported to Excel in XML format. Note that it's quite a large amount of data.

Although you didn't have the "large leaks" that result in cessation of recording, you did have some rather hefty leaks over 100. Also, the pattern of your leaks was all over the place. The pattern for your Hans Rudolph was better with at least some straight "zipper" segments. When the leaks look like your ZZZ mask pattern, you're probably getting way less than 4 cm and perhaps that could affect your AHI. You are also showing mucho snoring although whether or how much that affected your AHI I don't know.

If I were you, I'd get the leaks under control and then experiment with pressures - pretty much the reverse of the way you are doing it.

Mindy

Thanks for the info on the data ... I have installed EncorePro, but can't get it to run without an SQL error, and sadly gave up on it quickly. It looks like a nice piece of software. If it can export data, I may try to get it running again as I would certainly like that capability.

Yes the Hans looks cleaner than the ZZZ, but the average leak rate for both masks was essentially the same, 29 and 30, so would have to say average leak rates did not create this large AHI difference. I would be curious to know if the cleaner leak line with the Hans made a difference ...

I think the huge snoring is having an impact, but can't explain why, and both nights were pretty excessive, but not so different to create such a huge AHI difference in my opinion. Hopefully somebody can chime in ...

Yes, leaks need further work and am still in the learning process. I think at 4cm I am having a hard time detecting them ha ha ... I normally run at 12cm, and detecting leaks is no problemo ...

GumbyCT wrote:At 4cm I don't think it would take much of a leak to make a difference. Keep in mind as the pressure increases - it becomes more difficult to keep a seal.

I'd have to agree with you on that ... the average leak rates for the two nights was 29 and 30, essentially the same, so I think something else might be at work here ... but dunno what lol ....

Michael

Yes the Hans looks cleaner than the ZZZ, but the average leak rate for both masks was essentially the same, 29 and 30, so would have to say average leak rates did not create this large AHI difference. I would be curious to know if the cleaner leak line with the Hans made a difference ...

The average leak rate is not the best way to assess leaks. The graph tells the whole story. The less up and down on the leak line the better. You want to aim for the straightest line possible. I would rather see a straight line at a slightly higher number than a jagged line with a good average. It looks to me that the ZZZ wasn't sealed hardly at all. The Hans got some seal but needs work. Leaks will ruin this for you. Gotta fix the leaks before the rest of the data will really help.

Brenda

bdp522 wrote:

Yes the Hans looks cleaner than the ZZZ, but the average leak rate for both masks was essentially the same, 29 and 30, so would have to say average leak rates did not create this large AHI difference. I would be curious to know if the cleaner leak line with the Hans made a difference ...

The average leak rate is not the best way to assess leaks. The graph tells the whole story. The less up and down on the leak line the better. You want to aim for the straightest line possible. I would rather see a straight line at a slightly higher number than a jagged line with a good average. It looks to me that the ZZZ wasn't sealed hardly at all. The Hans got some seal but needs work. Leaks will ruin this for you. Gotta fix the leaks before the rest of the data will really help.

Brenda

Makes sense ... steady as she goes is better ... Maybe I should wear the masks under pressure awake for a period of time to perfect them ...