What is 90% Pressure all about?

Hi,

Can anyone tell me what exactly is the 90% pressure? and why is it important?

Is it that the system found this pressure most effective at treating apneas?

I'm trying to work out how to calculate it.

With Resmeds it is 95% pressure, the pressure AT or BELOW which you spent 95% (or 90% if a PR) of the night. And, yes, my understanding is that that pressure is the pressure needed to reduce or eliminate most "events" (the apneas and hypopneas). Usually, once the pressure level needed to eliminate most all apneas has been determined, then another 1cm to 2 cm is used to reduce or eliminate the hypopneas.

Heck, I don't know what I'm talking about. Let jnk, RG, Muffy, whoever, answer this. Not me.

Actually the Respironics definition is that the 90% is the pressure that you spent 90% of the time at and below. That is all. They don't mention this would be the "ideal" pressure. It might be a number close to where you optimal pressure might be or it also can be an inflated number if there are leaks and the machine is trying to fix the leaks or even if there are some rogue events that might on occasion need a higher pressure. It is a good starting point though.

For me after nearly 2 years on APAP with Min of 10 and max near 20 my 90% will be around 11 to 13. Sometimes less than 11 and on occasion more than 13. I saw one night earlier this week where it was 16. Rogue events, and a little leak seemed to push it up. Big "debate" on this very 90% discussion just yesterday in another thread. It got rather ugly. I can look over my reports and see that I don't often go above 12 (90%) and since I prefer to stay at 10 minimum for comfort and my AHI is often less than 1. I don't see the need to raise the minimum to match the 90% number. Even if I were to go to straight cpap I would probably choose 10 because it seems to be adequate for my needs for the majority of the time. The overall pressure average I might see a night would be more around 10.5 to 11.

So I never worried so much about the 90% number except in sort of passing. It is easily seen if pressure is driven up for some reason and easily seen if there is a pattern to it. If it is not falsely inflated because of leaks, it won't be all that different from the minimum.

Look at your daily details and you will see where it tends to want to stay when it isn't being driven by other forces. If there are times where it gets higher look at the reason it went higher and see if it was leak or events that drove it up.
Remember also that the APAP at whatever pressure minimum you choose will probe a bit and the 90% will likely never exactly match the minimum just because of the little probes if nothing else.

So while the 90% pressure might be close to your optimal pressure it can also be misleading if it is falsely inflated due to odd factors like a leak or rogue stubborn events. Sure I could use a pressure of 16 as my minimum or on cpap and I would likely zap the events, but who wants to do that when 10 min gives me great numbers too? This is why I use APAP, so I can use the lowest minimum that deals with my events effectively and let the machine come up to deal with the stubborn ones.

Edited to correct my typo. I thought the "and below" but fingers didn't work right. I didn't catch that till it was brought it to my attention. It was an error on my part. Mind and fingers working independently, again....

Thank you, Pugsy. Pressures are not my forte. You explained it very nicely.

Except I do want to clarify that Respironics' 90% and Resmed's 95% is the pressure the pressure AT or BELOW which you spent 90% (95%) of the night.

Here are the 2 nights I mentioned and used as an example in the other thread.
It is obvious what drove up the pressure on the night I hit a 16 cm 90% number. Bad event drove pressure up and then leak drove it up even more. Most of my nights are like the one where things went smoothly or I have just minor bumps. Even if I saw 16 every night but the majority of that night was lower, I would still use 10 minimum and just let the machine deal with the oddball stuff.
We of course have to look at the overall night. AHI good? Don't change. If AHI not so good night after night, then obviously some change in minimum might be warranted if we can't blame a higher 90% on leaks...

In other words, I look at the overall picture and I don't zero in on one specific detail because sometimes one detail might be misleading and that includes AHI (overall could be good but maybe have lots of clusters over the night) or leak (could be short time with bad leak and rest of night quite acceptable) or in this case the 90 % number. No one number is the holy grail of numbers to denote perfection. It doesn't exist. Just too many variables available to skew with numbers Best we can to is optimize things base on what we see on our own reports. Not what someone else might do or say.

Slinky wrote:I do want to clarify that Respironics' 90% and Resmed's 95% is the pressure the pressure AT or BELOW which you spent 90% (95%) of the night.

Correct. The 90% figure (Respironics) or 95% figure (ResMed) means the machine used THAT pressure AND pressures BELOW that pressure, 90 (or 95) percent of the time.

It does not mean the machine spent 90 or 95% of the time AT that one pressure.

Example: A person is using a ResMed S8 or S9 autopap.

He's using his autopap in auto mode, set for a range of 9 - 20. Minimum pressure 9, maximum pressure 20.

He checks his numbers the next morning in the little LCD window on the machine, and sees that last night's 95% pressure number was 16. The machine used THAT pressure AND pressures BELOW that pressure 95% of the night.

If he had the software to see details, he might see that the machine used pressures of 9, 10, and 11 MOST of the night... hours and hours spent down at those pressures.

He might see that the machine went up to a pressure of 17 for a short time once during the night.

He sees that the machine used these pressures... 12, 13, 14, 15 and 16 ... for just a few minutes each.

MOST of the time, pressures of 9, 10 and 11 were all that his machine had to use to keep his airway open.

BRIEFLY, the machine had to use pressures of 12, 13, 14, 15, 16, and 17.

So, the 95% pressure is going to be reported at 16. The person spent 95% of his time AT and BELOW 16. And he spent only 5% of the time above 16 (when the machine went up to 17.)

Why, then, is the 90th or 95th centile pressure reported at all, if a person might have spent just a few minutes at that pressure and spent hours down at lower pressures? Why don't they just say, "17 was the highest pressure used, but 10 is the pressure that effectively kept the airway open MOST of the time."

I think it's because when autopaps were first developed, the manufacturers expected them to be used mostly for temporary use... to titrate someone at home if the person could not go to the sleep lab for an attended PSG titration.

So, if the autopap is used for a one or two week trial at home, and the overall 90% or 95% number is 16, the doctor is going to probably order the machine set at that pressure or even perhaps a cm or two more. That's the SINGLE pressure that can take care of "worst case scenario" even if "worst case scenario" happens only for very short times each night. And even if 10 or 11 would take care of things fine MOST of the night.

REM (rapid eye movement) is pretty important "sleep." REM is when most apneas hit most people the hardest. They also hit most people hard when sleeping "supine" (on your back.) REM+Supine together = "worst case scenario" for most people, so if you're gonna use CPAP at a single pressure, it needs to be high enough to keep the airway well and truly open when sleeping on your back AND in REM. Thus, the 90 or 95% number from an autopap. Gives the doctor a useful single number to prescribe when the doc says, "Ok, put them on CPAP at ___ cm H2O."

EXCELLENT explanation! RG does it again... If it's not already there, that belongs in the Wiki...

Thanks.. That makes a lot more sense..

I should be able to figure out how to calculate it now..

Well IYAM, I think the 90th Percentile concept can be very misleading. If you're having a lot of events at Pmax, or if you have some position- or stage-dependent events, locked therapy at P90 can leave a lot of SDB on the table.

For instance, in this example



there appears to be a lot of significant stuff occuring at >P90.

I think the "significant stuff" was mostly occurring during periods of high leakage. NO conclusions should be drawn with a leakline like that. (Except maybe mask-related conclusions!

P90 is meaningless under those conditions.

LinkC wrote:I think the "significant stuff" was mostly occurring during periods of high leakage. NO conclusions should be drawn with a leakline like that.

You're getting colder...

NotMuffy wrote:You're getting colder...

Is this a test?
That posted report would look an awful like mine would if I didn't have the minimum pressure high enough. While the leak isn't perfect it it still within the parameters that the machine should be able to deal with for the most part and not completely invalidate the results. I suppose the groups could be REM related or even positional but that would be why I wanted APAP "just in case".

Thanks for posting that chart, NM.
To me, that chart shows what can happen with a fairly wide-open pressure range with a person who spends a good portion of the night on their back. It also shows how ridiculous it is to rely on a "90%" (or 95%) number to set one's pressure. AND, why using APAP in a range of pressures is NOT necessarily a good idea. This person had fewer events when the pressure was at the lower numbers or was dropping back down from the peaks. My guess is that the pressure and events dropped after the person rolled over onto his/her sides. The leakage is a little problematic but the average numbers are still within "tolerances". It also leads me to believe that those clusters and the higher leakage at those times are due to supine sleeping.


Den

Pugsy wrote:Is this a test?

Naturally.

Pugsy wrote:That posted report would look an awful like mine would if I didn't have the minimum pressure high enough.

And the biggest reason why P90 unto itself may offer incomplete information. 90% of stupid is still stupid.

Wulfman wrote:It also leads me to believe that those clusters and the higher leakage at those times are due to supine sleeping.

Also a superior reason why P90 may miss the boat.

However, in this case, I don't think that's the reason.