Is the S9 event data questionable at best?

brucifer wrote:The point I am making is that my S9 was unable to distinguish between apnea events that occur while I am sleeping and false-positive events that occur while I am awake.

The point you are missing is that the Auto part of CPAP, includiing scoring, is entirely predicated on autonomous sleep breathing behaviour

So ANY DATA RECORDED AND SCORED WHILE YOU ARE AWAKE is entirely meaningless for any purpose.

Cheers,

Bill

brucifer wrote: And how do you tell the difference between the two? That is my point with this thread.

I know that was your point but I have no answer for you that would satisfy you. You weren't happy with the common "awake breathing is different from sleep breathing".

A false negative would mean that no events were scored when they should have been. We wouldn't have any way to compare anything to even try to tell the difference. Nothing gets scored period with a false negative. We would have to put the entire flow graph under the microscope and try to evaluate every little minor reduction in flow to see if we thought that an event should have been flagged.
Way more work than I would want to attempt even if I did have the skill levels to "read" every little blip on the wave form chart and I don't have those skills nor do I want them.

I personally am good with the data. If my AHI is low I don't worry about if some of the events were "real" or not. It is low and even if real event or false positive not a big deal. I don't even want to go down the road that the machine may miss some events. People obsess enough over the numbers they see. I don't even want to bring up what numbers they may not see because the machine missed some.
I see false positives flagged but that doesn't mean that the big blank spots with nothing flagged are suspect in my mind.
At some point we have to either trust the data or not. I think if anything the machine manufacturers have set things up as best they can and if there is going to be an error it is on the side of false positives and not false negatives which would be harmful.

You can of course think which ever way you want to go with your thinking. Not my place to try to change your mind.

Are the machines perfect? Of course not but they are all we have if we want something other than "how we feel" to try to evaluate.

robysue wrote:

sol wrote:The machine is fooled, cannot tell if you,re asleep or wake. Even if you,re asleep there is no way of knowing if its a true apnea or not, an apnea have to be accompanied with oxygen desat

On a PSG, an OA or CA does NOT require an O2 desat. It does require that you are asleep (according to the EEG) when the OA or CA occurs,

There's a little difference in the criteria different groups use, but my understanding is the usual definition requires an x% O2 desat or a y% EEG change. Plus a x% decrease in airflow.

billbolton wrote:

brucifer wrote:The point I am making is that my S9 was unable to distinguish between apnea events that occur while I am sleeping and false-positive events that occur while I am awake.

The point you are missing is that the Auto part of CPAP, includiing scoring, is entirely predicated on autonomous sleep breathing behaviour

So ANY DATA RECORDED AND SCORED WHILE YOU ARE AWAKE is entirely meaningless for any purpose.

Cheers,

Bill

If what you are stating here is true, then one would expect the meaningless data recorded while one is awake to not have a nearly IDENTICAL event signature to what one records while one is asleep. Therefore, for comparative purposes, the event record while one is awake is indeed useful.

Furthermore, from studying my S9 graphs, how is my pulmonologist going to discern periods when I was awake from periods when I was asleep? How is the data generated by the S9 truly going to enlighten her in addressing my CPAP therapy?

Please understand, I'm not throwing the baby out with the bath water. I believe that CPAP therapy has a proven track record of being beneficial to patients. No argument from me there. Rather, I just question the validity and reliability of the data generated by the S9.

brucifer wrote:

BlackSpinner wrote:Most people's breathing while awake is very erratic. We hold our breath when we move or when we get startled. We do deep breathing when excited. When we sleep we don't do that, breathing goes on auto pilot as we sleep.

True. However, I was on my back and very still while texting away with my iPhone. I was breathing very normally except for the Tidal Volume and the Minute Ventilation. Those two wave charts showed much greater amplitude. The point I am making is that my S9 was unable to distinguish between apnea events that occur while I am sleeping and false-positive events that occur while I am awake.

Unlike Santa, who "can tell when you are sleeping and tell when you're awake", the S9 cannot distinguish between wake and sleep because it has no EEG.

The S9 does not have any programming that attempts to use the size of the inhalations to determine whether you are awake or not. If it is on and it is detecting breathing, it assumes that you are asleep and it scores events whenever the airflow drops sufficiently far from baseline. It's that simple. (The F&P Icon does try to guess when you are awake, and it must be making that guess on the size of the inhalations since all it has to work with is the wave flow, and wake breathing tends to be both more erratic and deeper [i.e. more airflow] than our normal sleep breathing.)

Because the S9 assumes that you must be asleep if you are using it, it will score an apnea of some sort anytime the airflow into/out of the lungs drops by 90% from baseline for at least 10 seconds, and "baseline" is computed using no more than the last two or three minutes of data, if I recall. Likewise, the S9 will score a hypopnea any time the airflow into/out of the lungs drops by 50% or more from baseline for at least 10 seconds as I recall.

If the amplitude of the wake inhales is much greater than normal, then both apneas and hypopneas can be scored with more airflow than you normally see during them. An example: If your normal sleep breathing inhalations normally top out around 20 L/min (like mine do), then to score apnea requires that the inhalations top out at no more than 2 L/min and a hypopnea requires the airflow to be less than 10 L/min. If your normal wake breathing tops out 60 L/min, an apnea could be scored with inhalations with peak flow as large as 6 L/min and a hypopnea can be scored with inhalations with peak airflow as large as 30 L/min.

Consequently, if my S9 cannot distinguish between what is a real event and what is not, then what is the point of depending on the data generated? In essence, we really don't know what we are looking at when we view our respective data.

If you are awake for long periods during the night with the mask on, the number of false events can present a real problem in interpreting the data. That's why Pugsy and others will often ask folks who've posted nasty looking data if they remember being really restless during the night. But if you are asleep during 90-95% of the time you are using the machine, there's simply less time to score a bunch of false events. And if 2 or 3 false events are scored during the brief amount of WAKE over the course of a seven or eight hour hour night, the marginal increase in reported AHI is statistically not that important.

Pugsy wrote:

brucifer wrote: And how do you tell the difference between the two? That is my point with this thread.

I know that was your point but I have no answer for you that would satisfy you. You weren't happy with the common "awake breathing is different from sleep breathing".

A false negative would mean that no events were scored when they should have been. We wouldn't have any way to compare anything to even try to tell the difference. Nothing gets scored period with a false negative. We would have to put the entire flow graph under the microscope and try to evaluate every little minor reduction in flow to see if we thought that an event should have been flagged.
Way more work than I would want to attempt even if I did have the skill levels to "read" every little blip on the wave form chart and I don't have those skills nor do I want them.

I personally am good with the data. If my AHI is low I don't worry about if some of the events were "real" or not. It is low and even if real event or false positive not a big deal. I don't even want to go down the road that the machine may miss some events. People obsess enough over the numbers they see. I don't even want to bring up what numbers they may not see because the machine missed some.
I see false positives flagged but that doesn't mean that the big blank spots with nothing flagged are suspect in my mind.
At some point we have to either trust the data or not. I think if anything the machine manufacturers have set things up as best they can and if there is going to be an error it is on the side of false positives and not false negatives which would be harmful.

You can of course think which ever way you want to go with your thinking. Not my place to try to change your mind.

Are the machines perfect? Of course not but they are all we have if we want something other than "how we feel" to try to evaluate.

Pugsy, no doubt, the data generated by the S9 is fuzzy at best. I guess we have to take it all with a grain of salt and a degree of faith. I'm not from Missouri, but I do have a show-me mind set. I'm not going to kid myself and call something a diamond when I believe it is a lump of coal.

brucifer wrote: no doubt, the data generated by the S9 is fuzzy at best. I guess we have to take it all with a grain of salt and a degree of faith. I'm not from Missouri, but I do have a show-me mind set. I'm not going to kid myself and call something a diamond when I believe it is a lump of coal.

You think it is fuzzy. I don't. A difference of opinions. I choose to trust the machine within its limitations. You don't want to.
If the normal explanations don't satisfy you then there is no point in trying to change your mind. Like I said...not my place in life.
You are surely entitled to your opinion. I won't be trying to change it. If you don't trust the data then don't look at it.

Pugsy wrote:

brucifer wrote: no doubt, the data generated by the S9 is fuzzy at best. I guess we have to take it all with a grain of salt and a degree of faith. I'm not from Missouri, but I do have a show-me mind set. I'm not going to kid myself and call something a diamond when I believe it is a lump of coal.

You think it is fuzzy. I don't. A difference of opinions. I choose to trust the machine within its limitations. You don't want to.
If the normal explanations don't satisfy you then there is no point in trying to change your mind. Like I said...not my place in life.
You are surely entitled to your opinion. I won't be trying to change it. If you don't trust the data then don't look at it.

No need to get snarky. Whether you want to outright admit it or not, the data is fuzzy. You mentioned the false positives yourself. A rose by any other name is still a rose. What I am attempting to do in this thread is understand WHY the machine generated the data that it did. I'm looking for LOGICAL explanations, not faith-based reasoning. I'm a trained scientist, so that's how I think.

If you choose to trust the machine within what you perceive to be its limitations, that's your choice. However, I'm still learning about this machine, so please pardon me if I feel the need to press forward and ask tough questions. You're certainly welcome to ignore my posts if my questions and comments bother you.

brucifer wrote: Furthermore, from studying my S9 graphs, how is my pulmonologist going to discern periods when I was awake from periods when I was asleep? How is the data generated by the S9 truly going to enlighten her in addressing my CPAP therapy?

Chances are you pulmonologist will never study the S9 graphs at all. She'll use the summary data to get a working approximation of how well your therapy is working and she will combine that data with the subjective data based on your answers to questions like: How do you feel now? How would you rate the quality of your sleep these days? Have any of your symptoms returned? If things look good and you report things as going OK, she'll assume the machine is doing it job.

If the summary data looks iffy (the AHI is too high or the leak rate is too high), she may briefly examine a few overviews of whole night waveform data to see if the problem appears to be nasty clusters surrounded by periods of calm or whether the whole night seems to be "not great". If there are clusters, she'll probably assume they are REM or supine sleep related. In either case, if you tell her you think you are sleeping at night and the number of CAs is small, she'll likely just prescribe a small increase in pressure and ask you to let her know if that doesn't help.

If the number of CAs is alarmingly high night after night and you report sleeping soundly with the machine, she won't spend any time analyzing the wave form from your machine; she'll use that data to justify bringing you back to the lab for another sleep study, most likely with an eye to doing a BiPAP ST or ASV titration.

Please understand, I'm not throwing the baby out with the bath water. I believe that CPAP therapy has a proven track record of being beneficial to patients. No argument from me there. Rather, I just question the validity and reliability of the data generated by the S9.

The machine AHI data is best understood as an approximation of the true AHI data, and as approximate data, the data is accurate enough for trending purposes. In other words, I'd never believe that Kaa gets my exact AHI right night after night. But if Kaa is consistently scoring so few events (real or false) that the computed AHI is less than 2.0 AND I'm feeling like I was asleep for 85-90% of the night most nights, I can safely assume that Kaa really is doing his job of preventing the vast majority of my events from taking place. And if Kaa suddenly started scoring enough events night after night that my AHI figures jumped to the 4.0 - 5.0 range and stayed there for a week or two AND I still thought I was asleep for 85-90 percent of the night, that would be enough data to warrant some further consultation with the sleep doc about what might need to be done.

robysue wrote:

brucifer wrote: Furthermore, from studying my S9 graphs, how is my pulmonologist going to discern periods when I was awake from periods when I was asleep? How is the data generated by the S9 truly going to enlighten her in addressing my CPAP therapy?

Chances are you pulmonologist will never study the S9 graphs at all. She'll use the summary data to get a working approximation of how well your therapy is working and she will combine that data with the subjective data based on your answers to questions like: How do you feel now? How would you rate the quality of your sleep these days? Have any of your symptoms returned? If things look good and you report things as going OK, she'll assume the machine is doing it job.

If the summary data looks iffy (the AHI is too high or the leak rate is too high), she may briefly examine a few overviews of whole night waveform data to see if the problem appears to be nasty clusters surrounded by periods of calm or whether the whole night seems to be "not great". If there are clusters, she'll probably assume they are REM or supine sleep related. In either case, if you tell her you think you are sleeping at night and the number of CAs is small, she'll likely just prescribe a small increase in pressure and ask you to let her know if that doesn't help.

If the number of CAs is alarmingly high night after night and you report sleeping soundly with the machine, she won't spend any time analyzing the wave form from your machine; she'll use that data to justify bringing you back to the lab for another sleep study, most likely with an eye to doing a BiPAP ST or ASV titration.

Please understand, I'm not throwing the baby out with the bath water. I believe that CPAP therapy has a proven track record of being beneficial to patients. No argument from me there. Rather, I just question the validity and reliability of the data generated by the S9.

The machine AHI data is best understood as an approximation of the true AHI data, and as approximate data, the data is accurate enough for trending purposes. In other words, I'd never believe that Kaa gets my exact AHI right night after night. But if Kaa is consistently scoring so few events (real or false) that the computed AHI is less than 2.0 AND I'm feeling like I was asleep for 85-90% of the night most nights, I can safely assume that Kaa really is doing his job of preventing the vast majority of my events from taking place. And if Kaa suddenly started scoring enough events night after night that my AHI figures jumped to the 4.0 - 5.0 range and stayed there for a week or two AND I still thought I was asleep for 85-90 percent of the night, that would be enough data to warrant some further consultation with the sleep doc about what might need to be done.

Excellent explanation, robysue. Thank you.

Sorry if you think I was snarky. Didn't particularly mean it that way. I can get snarky and down right nasty mean but this time I really didn't mean it that way.
Meant just a simple statement... if you don't trust the data then don't bother with it. I wouldn't bother with something I didn't believe in. Wouldn't waste my time.

False positive is probably not the best term to use but I can't think of anything any better that is easy to explain to people.

Pugsy wrote:Sorry if you think I was snarky. Didn't particularly mean it that way. I can get snarky and down right nasty mean but this time I really didn't mean it that way.
Meant just a simple statement... if you don't trust the data then don't bother with it. I wouldn't bother with something I didn't believe in. Wouldn't waste my time.

False positive is probably not the best term to use but I can't think of anything any better that is easy to explain to people.

I believe you and I appreciate that you didn't mean to be snarky.

Just to be clear, I'm trying to understand how to use the data and to what degree I can rely on it. I don't want to cast it aside if it can be useful. robysue gave a very good explanation just now it how it can be useful. She explained how the data, as nebulous as it may be at times, can be used a co-diagnostic tool by a sleep doc and by the patient. I'm very okay with that.

Yesterday I had a follow-up appointment with the sleep doc, when I open the laptop to show her the download - she didn't want to look at the download. She just said - your compliance is good, sleeping solid 7 hours every night and what matter most you,re feeling good and not sleepy during the day so go home do what you,re doing and see in a year or sooner if anything changes.

I don't understand why you would call the flagged events "false positives"? I stop breathing all the time when I'm awake. Before cpap I rarely noticed because it is usually while I'm doing something else, like typing, pouring, slicing, dialing, texting....
I trust my S9 data.

Lizistired wrote:I don't understand why you would call the flagged events "false positives"? I stop breathing all the time when I'm awake. Before cpap I rarely noticed because it is usually while I'm doing something else, like typing, pouring, slicing, dialing, texting....
I trust my S9 data.

In theory, very few people will stop breathing long enough while awake to cause any adverse health effects.

While sleeping, apneas may cause O2/CO2 problems, and problems with sleep disturbance that adversely affect your health.

If you stop breathing long enough to harm yourself while awake or even partly awake, you have an unusual condition that probably requires a different kind of treatment than normal sleep apnea.