Leaks and machine capability/reliability

[DNA256#]

I just got back from my DME (for help fitting with new/different mask).
I've been having lots of trouble with leaks above 35 l/min for extended periods of time,
that the reason for trying a new mask/cushion.
I am wearing a full face mask, due to mouth breathing.

So.... the respiratory therapist at the DME said that i really need to get these leaks under control because
" the Resmed Airsense 11 autoset becomes increasingly unreliable at leaks above 24 l/min"
When I asked, what problems occur at leaks between 24 and 50, he said he didn't know, just that Resmed has said that the machine's capabilities above 24 become increasingly unreliable.

But when we looked at the pressures created by the machine (during a moderately high leak 50-60 l/min, the machine was able to maintain, and sometimes actually increase pressure during the leakage episodes.

So, calling on the expertise of the forum members, What does happen that causes Resmed to say that acceptable leakage has to be less than 24 l/min? Exactly what capability or reliability is diminished?


'

[Pugsy]

The 24 L/min line in the sand is an ultra conservative line in the sand IMHO.
Gotta have a line in the sand somewhere and that's what ResMed chose to use.

From my own personal experience with leaks I found that somewhere between 30-35 L/min excess leak was when the machine started having trouble sensing, recording and/or responding to whatever was happening.

When I saw the leak hit 35 L/min I also saw "unclassified apnea" flagging...meaning the machine could tell something was going on but couldn't figure out just what to call whatever it sensed.

With leaks above 35 L/min....the data is unreliable. Meaning if it shows nothing happened was the reason it showed nothing because nothing happened or did it simply miss flagging stuff because it was clueless when faced with such a leak.
Kinda important to know don't you think?

[DNA256#]

Meeting soon with my "new" sleep doc (prior retired)

Should I suggest to him that the events observed during the >35 leak episode just be ignored? It would substantially change my AHI values as in general the majority of my events are detected during these leak periods.
I wonder how successful I'll be with that !

And, yes it's kinda important to know what the consequences are of leaks >35

[Dog Slobber]

Meeting soon with my "new" sleep doc (prior retired)

Should I suggest to him that the events observed during the >35 leak episode just be ignored? It would substantially change my AHI values as in general the majority of my events are detected during these leak periods.
I wonder how successful I'll be with that !

And, yes it's kinda important to know what the consequences are of leaks >35

No, She said the data was unreliable.

That doesn't mean that apneas reported didn't occur.

• It could still be accurate.
• It could be accurate in count but not classified correctly.
• Apneas may be occurring, but not flagged.
• Apneas may be flagged, but not occurring.
• Flagged apneas may not be classified correctly
• Other events (FLs, Snores) may be missed or incorrectly flagged.

You need to be reducing the leaks, not asking how to interpret unreliable data.

[Pugsy]

Should I suggest to him that the events observed during the >35 leak episode just be ignored?

No...that's not what I said.

What to do with that unreliable data actually depends on just how prolonged the massive leak happened to be.
Was it 10 minutes? If so we can safely ignore 10 minutes of unreliable data.
Was it half the night??? That absolutely cannot be ignored.

DogSlobber's reply.....I totally agree with.

[DNA256#]

Should I suggest to him that the events observed during the >35 leak episode just be ignored?

...that's not what I said.

HUH?
Slobber said "that's not what she said" I didn't think that I quoted you at all. So this particular part is confusing.

What I think I understood you to say was "With leaks above 35 L/min....the data is unreliable."
The "data" is what the doctor will be looking at. Data meaning AHI, A events, H events, etc.
Or were you and slobber making some other distinction about "data"?

What to do with that unreliable data actually depends on just how prolonged the massive leak happened to be.
Was it 10 minutes? If so we can safely ignore 10 minutes of unreliable data.
Was it half the night??? That absolutely cannot be ignored.

Again I'm perplexed. Are you saying that if 30% of the night is > 35 l/m, the unreliable data during those periods "absolutely cannot be ignored"? But If 10% of a night is >35 l/m, the unreliable data during that period can safely be ignored?

Does unreliable data some how become reliable as a proportion of the night?

[Pugsy]

I am sorry if my thoughts are unclear to you.
Perhaps someone else can explain it better than I can.

[DNA256#]

You need to be reducing the leaks, not asking how to interpret unreliable data.

The search for a "good" (minimal leak) mask has been going on for several months, and at some not insignificant expense. So, yes, I am/have been working on it.

[Sanjay1976]

Similar situation.
showed Oscar to doc, and he noted that 75% of my events occurred during big leak periods.
I would go to sleep for 2 hours and only have 2 events while I was at minimal leak. During the next 3 hours with leaks between 35-65, I would have 40 events. Then during the final 3 hours at leaks less than 25, I would have 2 events. Doc said, if he were grading the night he would say AHI was about 1 but he would not say 5 because he was skeptical of the large leak time.

[Dog Slobber]

Similar situation.
showed Oscar to doc, and he noted that 75% of my events occurred during big leak periods.
I would go to sleep for 2 hours and only have 2 events while I was at minimal leak. During the next 3 hours with leaks between 35-65, I would have 40 events. Then during the final 3 hours at leaks less than 25, I would have 2 events. Doc said, if he were grading the night he would say AHI was about 1 but he would not say 5 because he was skeptical of the large leak time.

And your Doctor might be wrong.

During your large leak time, Apneas are often (always?) flagged as unclassified. ResMed's don't increase pressure for UAs.

With that in mind, if during a large leak you experience an Obstructive Apnea, but it's flagged as a UA, then the device will *not* respond. Therefore, not treating the obstructive apnea can contribute to more OAs that would not have occurred, had the device been able to reliably flag them.

Missed trigger events, such as FLs, can also lead to OAs, this can then lead to the above scenario.

The fact that UAs are not responded to and missed FLs; explains that during periods of high leak more apneas are actually likely to occur.

One last point, experienced member palerider has claimed that during extended large leaks ResMed's can actually lower pressure to try to mitigate the leak. I accept this. Lowering the pressure can further contribute to OAs.

Your Doctor is interpreting unreliable data as meaning; the Apneas are likely not to have occurred. This is wrong. I've described three potential causes, where it is incorrect to be treating large leak UAs as probably not likely to have occurred.

Your Doctor doesn't understand how ResMed's behave.

[Tec5]

Similar situation.
showed Oscar to doc, and he noted that 75% of my events occurred during big leak periods.
I would go to sleep for 2 hours and only have 2 events while I was at minimal leak. During the next 3 hours with leaks between 35-65, I would have 40 events. Then during the final 3 hours at leaks less than 25, I would have 2 events. Doc said, if he were grading the night he would say AHI was about 1 but he would not say 5 because he was skeptical of the large leak time.

Your Doctor doesn't understand how ResMed's behave

Doc seems to be suggesting that the 5 hours with well-sealed mask is likely more representative of the OP's normal sleep, at least for that night.
IMO doc should be looking at many night's rather than just spotlighting one night. For all we know, that broader point of view informed the doc's skepticism of that 3 hour period in isolation.

[Max46]

Similar situation.
showed Oscar to doc, and he noted that 75% of my events occurred during big leak periods.
I would go to sleep for 2 hours and only have 2 events while I was at minimal leak. During the next 3 hours with leaks between 35-65, I would have 40 events. Then during the final 3 hours at leaks less than 25, I would have 2 events. Doc said, if he were grading the night he would say AHI was about 1 but he would not say 5 because he was skeptical of the large leak time.

Your Doctor doesn't understand how ResMed's behave

Doc seems to be suggesting that the 5 hours with well-sealed mask is likely more representative of the OP's normal sleep, at least for that night.
IMO doc should be looking at many night's rather than just spotlighting one night. For all we know, that broader point of view informed the doc's skepticism of that 3 hour period in isolation.

Yeah, Particularity if this pattern* seems to be a common occurrence.

*events being concentrated in periods of large leaks, while other 'leak minimal periods' accompanied by significantly lower event counts.