Squishing the belly increases AHI

[NotMuffy]

Ok, I have done a very quick calculation of sleep efficiency within each cuff condition. Correct my calculations if they are wrong. I calculated sleep efficiency as: sleep (in epochs)/total epochs under each cuff condition. This is what I got:

Deflated: 69.1+/-5.1%
Mid: 65.2+/-4.4%
Max: 59.3+/-4.6%

There was no significant difference in sleep efficiency (p=0.132)

Not wishing to be argumentative (just kidding!) I am saying that unstable sleep creates unstable breathing. Sleep Efficiency <75% is abysmal, so this entire patient study is affected (you got bad, worse, worst).

I would also offer that attempting to create a standard for this (admittedly, we're inventing it as we go along), one shouldn't a 0-100% scale, one should use like a 60-100% scale, with the decay of the curve against what you're measuring (let's say EDS, for example) sigmoid in nature (you can go from 100% SE down to about 93% or so and not feel too bad, and then less than that, everything rapidly hits the fan. At 60% SE, we're at the point of "why even bother going to bed".

[split_city]

Ok, I have done a very quick calculation of sleep efficiency within each cuff condition. Correct my calculations if they are wrong. I calculated sleep efficiency as: sleep (in epochs)/total epochs under each cuff condition. This is what I got:

Deflated: 69.1+/-5.1%
Mid: 65.2+/-4.4%
Max: 59.3+/-4.6%

There was no significant difference in sleep efficiency (p=0.132)

Not wishing to be argumentative (just kidding!) I am saying that unstable sleep creates unstable breathing.

I can turn it around and say unstable breathing creates unstable sleep.

Sleep Efficiency <75% is abysmal, so this entire patient study is affected (you got bad, worse, worst).

I would also offer that attempting to create a standard for this (admittedly, we're inventing it as we go along), one shouldn't a 0-100% scale, one should use like a 60-100% scale, with the decay of the curve against what you're measuring (let's say EDS, for example) sigmoid in nature (you can go from 100% SE down to about 93% or so and not feel too bad, and then less than that, everything rapidly hits the fan. At 60% SE, we're at the point of "why even bother going to bed".

What's you lab's average patient sleep efficiency? By definition, patients come in because they have crappy sleep. I thought I would look over some of our lab's data. I looked at about 8000 diagnostic studies in which each patient had at least 4 hrs time in bed and the average sleep efficiency for the entire population was 70.2%. When separating into no OSA (AHI<15), mild OSA (15-<30), moderate OSA (30-<45) and severe OSA (45+), the sleep efficiency data were 71.9, 71.6, 69.1 and 66.7% respectively.

I just found this study to compare:

To date, the clinical use of unattended home-based polysomnography (PSG) is not recommended. To assess whether sleep efficiency is better at home, we have performed a prospective, crossover, single-blind study comparing unattended home- versus attended in-hospital PSG in a population referred for high clinical suspicion of obstructive sleep apnoea syndrome (OSA). Within 2 weeks, all the patients underwent both PSG performed by the same sleep technician, which were analysed by another blinded technician. Payments for each procedure were also calculated. Sixty-six patients (mean age: 49 ± 13 years; mean body mass index: 30 ± 7; mean Epworth Sleepiness Scale: 10 ± 5) were included. The quality of recordings was poor in 1.5% of the attended PSG versus 4.7% for unattended PSG (P = 0.36). Sleep efficiency at home was better (82% versus 75%, P < 0.001), and sleep duration longer (412 min versus 365 min, P < 0.001). Sleep latency was also shorter at home (28 min versus 45 min, P = 0.004), and patients spent more time in rapid eye movement sleep (19% versus 16%, P = 0.006). Apnoea–hypopnoea index (23 versus 26, P = 0.08) was similar at home and in the sleep lab. Sixty-seven per cent of patients preferred home-based PSG. PSG payment was also lower at home (268 Euros versus 1057 Euros). We conclude that home-based PSG is associated with a better sleep efficiency. It also appears as feasible and reliable in patients with high preclinical suspicion for OSA. It is also more comfortable for the patients whose sleep efficiency is better and allows cost saving related to the absence of hospitalization.

So the sleep efficiency in their lab setting was 75% compared to our 70%.

Coming back to a comment by you:

I would back all the way out and do full NPSG, baseline and then each pressure, so now you have like 4 nights total. Sleep efficiency <85% excludes the study.

Geez, patients in our lab and in others aren't getting anywhere near 85% during a standard PSG. What chance do they have with all the additional equipment used in my study? Basically means we, nor any other research group for that matter, should run any studies like this.........

On a side note, does that mean we should call all our patients up and say "sorry, your sleep efficiency was terrible meaning your OSA is simply due to you having bad sleep in the lab. Please stop using your CPAP machine because your results are invalid."

[NotMuffy]

I can turn it around and say unstable breathing creates unstable sleep.

Sure.

What's you lab's average patient sleep efficiency?

Low to mid 90s.

By definition, patients come in because they have crappy sleep.

Perhaps, but SE is only one component of that, and also by definition, EVERYBODY says anything under 85% is abnormal.

I thought I would look over some of our lab's data. I looked at about 8000 diagnostic studies in which each patient had at least 4 hrs time in bed and the average sleep efficiency for the entire population was 70.2%. When separating into no OSA (AHI<15), mild OSA (15-<30), moderate OSA (30-<45) and severe OSA (45+), the sleep efficiency data were 71.9, 71.6, 69.1 and 66.7% respectively.

Maybe you need to add more channels. Like NPT.

So does that mean we should call all these patients up and say "sorry, your sleep efficiency was terrible meaning your OSA is simply due to you having bad sleep in the lab. Please stop using your CPAP machine because your results are invalid."

No, tell 'em you'll send them to a place where they can actually "sleep". I'll send you some brochures.

[split_city]

What's you lab's average patient sleep efficiency?

Low to mid 90s.

Nice. Is that common across all sleep centres though?

Do you guys have a research arm?

So does that mean we should call all these patients up and say "sorry, your sleep efficiency was terrible meaning your OSA is simply due to you having bad sleep in the lab. Please stop using your CPAP machine because your results are invalid."

No, tell 'em you'll send them to a place where they can actually "sleep". I'll send you some brochures.

Hahaha, well our lab does have the oldest beds in the hospital. That's a long story for another day. Political nightmare

[NotMuffy]

OK, seriously...

On an 8-hour record, you're looking at about 2.5 hours of wake time. That's HORRIBLE! How can you possibly assess sleep if the (or an) issue is sleep maintenance? Are you matching the patients' sleep times with lab Lights Out/On (i.e., don't start the test too early)? Is there a pattern to the wake?

[split_city]

OK, seriously...

On an 8-hour record, you're looking at about 2.5 hours of wake time. That's HORRIBLE! How can you possibly assess sleep if the (or an) issue is sleep maintenance? Are you matching the patients' sleep times with lab Lights Out/On (i.e., don't start the test too early)? Is there a pattern to the wake?

We based the lights off time on what each patient's normal sleep time was. While we did get this pretty close to the designated time, delays in setup pushed that out by up to 30mins. The latest we allowed them to sleep till was about 6am because that's when the clinical patients were woken up.

[NotMuffy]

What's you lab's average patient sleep efficiency?

Low to mid 90s.

Nice. Is that common across all sleep centres though?

I don't know, but since most references list 85% as the point at which insomnia begins (insert debate about definition of insomnia here), I think a lab needs to target ~>93% SE to be able to diagnose sleep architectural issues.

In order to assist interpretation, SE is rated out as (and if there was a reference for this, I have since forgotten):

>95% - Excellent
85 - 94% - Good
75 - 84% - Fair
<75% - Poor

Do you guys have a research arm?

Negative.

We based the lights off time on what each patient's normal sleep time was. While we did get this pretty close to the designated time, delays in setup pushed that out by up to 30mins. The latest we allowed them to sleep till was about 6am because that's when the clinical patients were woken up.

In that case, you got me. That schedule would certainly seem to increase SE (perhaps artificially so).

[split_city]

What's you lab's average patient sleep efficiency?

Low to mid 90s.

Nice. Is that common across all sleep centres though?

I don't know, but since most references list 85% as the point at which insomnia begins (insert debate about definition of insomnia here), I think a lab needs to target ~>93% SE to be able to diagnose sleep architectural issues.

In order to assist interpretation, SE is rated out as (and if there was a reference for this, I have since forgotten):

>95% - Excellent
85 - 94% - Good
75 - 84% - Fair
<75% - Poor

I actually tried to find any reports quoting sleep lab SE numbers. The only one I could find was the one I mentioned earlier:

To date, the clinical use of unattended home-based polysomnography (PSG) is not recommended. To assess whether sleep efficiency is better at home, we have performed a prospective, crossover, single-blind study comparing unattended home- versus attended in-hospital PSG in a population referred for high clinical suspicion of obstructive sleep apnoea syndrome (OSA). Within 2 weeks, all the patients underwent both PSG performed by the same sleep technician, which were analysed by another blinded technician. Payments for each procedure were also calculated. Sixty-six patients (mean age: 49 ± 13 years; mean body mass index: 30 ± 7; mean Epworth Sleepiness Scale: 10 ± 5) were included. The quality of recordings was poor in 1.5% of the attended PSG versus 4.7% for unattended PSG (P = 0.36). Sleep efficiency at home was better (82% versus 75%, P < 0.001), and sleep duration longer (412 min versus 365 min, P < 0.001). Sleep latency was also shorter at home (28 min versus 45 min, P = 0.004), and patients spent more time in rapid eye movement sleep (19% versus 16%, P = 0.006). Apnoea–hypopnoea index (23 versus 26, P = 0.08) was similar at home and in the sleep lab. Sixty-seven per cent of patients preferred home-based PSG. PSG payment was also lower at home (268 Euros versus 1057 Euros). We conclude that home-based PSG is associated with a better sleep efficiency. It also appears as feasible and reliable in patients with high preclinical suspicion for OSA. It is also more comfortable for the patients whose sleep efficiency is better and allows cost saving related to the absence of hospitalization.

I did speak to my lab manager and he said that sleep efficiency numbers for lab studies reported in the literature lie between low 70s and high 80s. He left before I could get hold of any of these reports.

[NotMuffy]

I did speak to my lab manager and he said that sleep efficiency numbers for lab studies reported in the literature lie between low 70s and high 80s. He left before I could get hold of any of these reports.

Those that I have seen are from quite a while ago, and few and far in between as well.

However, while it may sound reasonable and scientific to attribute poor SE to "First Night Effect", if I was the Manager of a Sleep Lab that was consistently putting up 70% SEs (no offense intended), there'd be a whole new Sleep Lab staff there by "Second Night".

[split_city]

Well, the next time you visit Australia to take on the challenge of being able to get >85% SE in a study of mine, you can also come and show the techs how the sleep studies should be run.

[NotMuffy]

Well, the next time you visit Australia... you can also come and show the techs how the sleep studies should be run.

I'll need 4 more tickets. I actually have little (if anything) to do with why we have 90%+ SE.

Can we go to a good burger place when we get there?

And don't be sneakin' in one of them roo burgers.

[split_city]

But roo meat is quite nice........would you settle for a Hungry Jacks meal (Burger King for you guys)?

If you don't like that, I guess I can always throw 'nother shrimp on the barbie