Apnea Board Forum - CPAP | Sleep Apnea
S9 AHI can be erroneous - Printable Version

+- Apnea Board Forum - CPAP | Sleep Apnea (https://www.apneaboard.com/forums)
+-- Forum: Public Area (https://www.apneaboard.com/forums/Forum-Public-Area)
+--- Forum: Main Apnea Board Forum (https://www.apneaboard.com/forums/Forum-Main-Apnea-Board-Forum)
+--- Thread: S9 AHI can be erroneous (/Thread-S9-AHI-can-be-erroneous)

Pages: 1 2


S9 AHI can be erroneous - zbaba4 - 11-03-2013

S9 AutoSet AHI display has little use. Most of us know if you have the mask on while awake the numbers don't count. For example last night I fell asleep quickly and awoke after 5 hours to use the bathroom. My AHI at that point was 1.5. I tried to go back to sleep but was very restless. One hour later I gave up. My AHI was 6.8! I looked at my data using ResScan and it showed a cluster of maybe 25 events in my awake hour. This is typical for me. Sometimes I do fall back to sleep. Even if I lie awake for only several minutes, I see clusters of events that mess up the AHI average. I also see a cluster for the time it takes to fall asleep (have to guess of course), but with a burst almost every night, I assume I am correct.

Anyway, sure I can use rescan to pull out my "real" data, but I wonder what use the S9 is for anyone that doesn't read their detailed data.

It matters to me a lot because I want to monitor my AHI as I reduce the pressure (very slowly) to see the results of a very large weight loss. So far, I have lowered the max from 13 to 10.6 and using resScan my "real" AHI has actually dropped (always below 5) and my hypopneas have dropped considerably. I have lowered the min pressure proportionately and now it is set to 8. The actual pressure spend more time near the max but through this whole process my "real" AHI has been very low (<3). I know I could just get another sleep study but I have another 40 lbs to go,then I may do it.


RE: S9 AHI can be erroneous - PaulaO2 - 11-03-2013

Congrats on the weight loss! Way to go!

The machines are doing what they are supposed to do. But they can't tell if you are awake or asleep. So, yes, in that regard, the on-screen data can be considered biased if the user is someone like you who does not go directly to sleep. But for those who do, it can be almost as good as looking at the data via the software.

I assume you are making the pressure changes over time, using data from ten or more days worth, correct?


RE: S9 AHI can be erroneous - Peter_C - 11-03-2013

I find this interesting, as it brings up a very good point for me. My machine is such that I currently can only see/use the onscreen data. I never thought about the nights that I spend lying in bed, not sleeping vs the nights when I fall asleep quickly, and the relation to my AHI numbers (which are much higher on nights when it takes time to go to sleep).

I do understand the machine can't tell if we are sleeping or not, but I did not think that my normal (awake) breathing could be counted as an AHI or AI event.

I never put my mask on until I am ready to *try* to go to sleep, no reading, or anything with my mask on. Mask on means go to sleep! But, due to my many different issues, there are times that I lie awake for an hour or two, give up and get up for 2-3hrs then try again.

I think I will pay attention and see if my numbers are indeed skewed due to this, or not.


RE: S9 AHI can be erroneous - PaulaO2 - 11-03-2013

An event is not noted unless it lasts for 10 seconds or more. When we try to do it on purpose, it seems like a long time. But when we are rolling over, or coughing, or talking, or whatever, it's not really that long at all. As we are laying in bed, unable to sleep, getting agitated about not being able to sleep, our breathing may or may not be a normal rhythm. The machine is trying to guess what is going on. It will flag events it thinks are long enough to be events.

I tend to fall asleep rather quickly once I go to bed. I have trained myself well due to my insomnia. My problem though, is when I am waking up. I wake up kinda, go back to sleep sorta, wake up again, go back to sleep again, and my breathing is all out of whack during this. I'm rolling over, grunting (pain), arguing with pillows, etc. But I can look at the pressure with SleepyHead and tell when I woke up. The pressure drops a LOT as soon as I start that "sorta kinda waking up" game. Any events that happen during then I ignore.


RE: S9 AHI can be erroneous - vsheline - 11-03-2013

(11-03-2013, 11:39 AM)zbaba4 Wrote: It matters to me a lot because I want to monitor my AHI as I reduce the pressure (very slowly) to see the results of a very large weight loss. So far, I have lowered the max from 13 to 10.6 and using resScan my "real" AHI has actually dropped (always below 5) and my hypopneas have dropped considerably. I have lowered the min pressure proportionately and now it is set to 8. The actual pressure spend more time near the max but through this whole process my "real" AHI has been very low (<3).

Hi zbaba4, welcome to the forum!

I think I would not lower the Max Pressure setting quite the way you are doing it.

I suggest it would be safe to keep on lowering the Minimum Pressure setting (as you monitor your data in ResScan or SleepyHead to verify your sleep quality is benefiting or at least not negatively affected), but I suggest we should not lower the setting for Max Pressure if the AutoPAP machine is raising the pressure near the Max Pressure setting while we are asleep.

The AutoSet raises the pressure only when needed to prevent or minimise obstructive events.

I suggest the Max Pressure setting should be at least a couple cm H2O higher than our 95 percentile pressure, unless we are still adapting to therapy and need Max pressure artificially lowered temporarily.

Regarding your observations that lowering the Min and Max settings together has improved your AHI, it is very easy to be misled by short term variation in our AHI numbers (from changes in sleep position, mask fit, stress, illness, diet and medication, etc), especially if waiting only a week or two between making adjustments to the settings.

My sleep doc likes to see 4 to 8 weeks worth of data between adjustments, unless the need for adjustments is fairly obvious, such as if our 95 percentile pressure is close to the Max Pressure setting and the machine is spending time max'ed out.

Take care,
--- Vaughn


RE: S9 AHI can be erroneous - Peter_C - 11-03-2013

(11-03-2013, 02:02 PM)PaulaO2 Wrote: An event is not noted unless it lasts for 10 seconds or more. When we try to do it on purpose, it seems like a long time. But when we are rolling over, or coughing, or talking, or whatever, it's not really that long at all. As we are laying in bed, unable to sleep, getting agitated about not being able to sleep, our breathing may or may not be a normal rhythm. The machine is trying to guess what is going on. It will flag events it thinks are long enough to be events.

I tend to fall asleep rather quickly once I go to bed. I have trained myself well due to my insomnia. My problem though, is when I am waking up. I wake up kinda, go back to sleep sorta, wake up again, go back to sleep again, and my breathing is all out of whack during this. I'm rolling over, grunting (pain), arguing with pillows, etc. But I can look at the pressure with SleepyHead and tell when I woke up. The pressure drops a LOT as soon as I start that "sorta kinda waking up" game. Any events that happen during then I ignore.

My concern is the fact that I am not able (til I get a newer machine) to read any data, other than what I can see on the OLED screen. So, the above tells me that I might be stressing over bad numbers, that may actually be related to what has happened while I am awake.

However, after getting my Oximeter, that data is giving me valuable insights! I showed an AHI of 19 last night, yet my O2 was only under 88 for 1.2 minutes *total* the entire sleep period.


RE: S9 AHI can be erroneous - zbaba4 - 11-04-2013

Thanks for all the replies. My major conclusion is if you can't run ResScan or sleepyhead, just tape over the AHI display.

Back to my lowering pressure... I'm doing it extremely slowly, like .2 per week if I lose weight.

I have posted here before, a confusing result. I know it would seem just setting a wide pressure range would suffice to see how you are doing. Not in my case. No matter what the upper limit is (<13), mine spends more than half the time near the top. But as I have lowered the pressure, my events have not increased at all, in fact decreased. My "filtered" data, from resScan, ignoring awake times has been less than 2 events per night, sometimes zero! I am tapering the pressure down to train my brain to get off it, I am afraid to go cold turkey because without the machine I can't tell what is happening. My sleep doctor told me after about 6, it will not matter.

The S9 adjusts pressure depending on flow limitation, not events. I am thinking about getting an oxymeter as final proof (short of a sleep study) but they seem so expensive. Anyone now of a low cost oxymeter.


RE: S9 AHI can be erroneous - Peter_C - 11-04-2013

(11-04-2013, 12:59 AM)zbaba4 Wrote: Anyone now of a low cost oxymeter.

Depending on what you consider 'low cost' to be?

You can buy a CMS50F (without bluetooth) on EBay for around $99 bucks, or from a great vendor with fast shipping for $139.00

This unit does everything most want, record data for up to 30hrs, rechargeable battery, software comes with it, has a loud alarm you can set for too low, too high, etc...

I like this unit as it is a wrist unit with a finger probe that does not come off/loose during the night, like a finger-only unit could.


RE: S9 AHI can be erroneous - PaulaO2 - 11-04-2013

So you are looking at the data with ResScan or SleepyHead? Good.

The idea is to look at the data and see what your norm is then adjust the pressure accordingly.

You have the Autoset but you say the pressure is 13 so are you using it as straight CPAP then? That's fine if that's what works for you.

If you are looking at the data and the events are less than 5, then lowering in increments of .2 is more than fine. You can drop in increments of .5. But using the criteria of "if I lost weight that week" is not a good criteria to use. You need to use your CPAP data, not the scale. Weight is but one factor of many in terms of cause of obstructive sleep apnea. And it depends on where the weight is. If you have a relatively normal size neck, meaning your neck is not overly large and your weight is more around your middle rather than all over, then losing weight may not help your sleep apnea much. It will help the rest of you, of course! There's a lot of scrawny people who have sleep apnea.

So if you have it at 13 and your AHI is 2, you feel good, and it has been 7 to 10 days (or more) since the last change, drop the pressure to 12.5. Wait a week to 10 days. AHI is still 2 (or lower), you still feel good, drop to 12. Wait another 7 to 10 days (or more). The longer you wait between changes the more data you get to gather, the more you can see patterns. No two nights are the same so the longer you wait, the more you can see the trends and patterns. I know it is hard to wait between changes. I've tried experimenting and it is so hard to not just go ahead but, really, the more data you have the better. Especially for you since you know the data has to be "altered" to account for you being awake and the data being skewed.


RE: S9 AHI can be erroneous - zbaba4 - 11-05-2013

I'm stating the high setting (13). Low has been set about 4 less, AutoSet mode. So I have gone from 13 to 10.6 and the AHI has gotten better as I do it. Can't get much better because now there are only one or two events per night and they are hypopneas or Central Apneas. In fact, there is another interesting observation. Now I see most nights with 3 or 4 hours at the outset that have absolutely no events. This wasn't true at 13. That has continued to get better also. So when I get get down to 6 or so and nothing changes, off it goes. Of course I can't measure anything then.

Info: My original, and only, sleep study put me at 13. I went for the study because of snoring and all sorts of other noises my wife noticed even while napping in a chair . Now, there is nothing.

I don't know, even if one is "cured", if you have to slowly wean of CPAP so your body gets used to it. I know several times when I couldn't use it (broken, no power, etc) I had great trouble falling asleep.

So hopefully all points to success.


(11-04-2013, 02:08 PM)PaulaO2 Wrote: So you are looking at the data with ResScan or SleepyHead? Good.

The idea is to look at the data and see what your norm is then adjust the pressure accordingly.

You have the Autoset but you say the pressure is 13 so are you using it as straight CPAP then? That's fine if that's what works for you.

If you are looking at the data and the events are less than 5, then lowering in increments of .2 is more than fine. You can drop in increments of .5. But using the criteria of "if I lost weight that week" is not a good criteria to use. You need to use your CPAP data, not the scale. Weight is but one factor of many in terms of cause of obstructive sleep apnea. And it depends on where the weight is. If you have a relatively normal size neck, meaning your neck is not overly large and your weight is more around your middle rather than all over, then losing weight may not help your sleep apnea much. It will help the rest of you, of course! There's a lot of scrawny people who have sleep apnea.

So if you have it at 13 and your AHI is 2, you feel good, and it has been 7 to 10 days (or more) since the last change, drop the pressure to 12.5. Wait a week to 10 days. AHI is still 2 (or lower), you still feel good, drop to 12. Wait another 7 to 10 days (or more). The longer you wait between changes the more data you get to gather, the more you can see patterns. No two nights are the same so the longer you wait, the more you can see the trends and patterns. I know it is hard to wait between changes. I've tried experimenting and it is so hard to not just go ahead but, really, the more data you have the better. Especially for you since you know the data has to be "altered" to account for you being awake and the data being skewed.