Why is pressure changing by 1.5 cm H2O? - Asking because my spinal joints hurt

[robysue]

Once again, I feel the need to come to the defense of the PR Auto's Search Algorithm.

Things are not quite as simple or as backwards as Palerider is suggesting.

The Search algorithm does kick in when the breathing appears to be normal:

I googled part of Dave's comment on the subject, and found this:

http://www.google.ch/patents/US20120152252

in case you're interested.

Haven't had time to fully digest it, but the stuff in the link does look very interesting.

[palerider]

Once again, I feel the need to come to the defense of the PR Auto's Search Algorithm.

Things are not quite as simple or as backwards as Palerider is suggesting.

The Search algorithm does kick in when the breathing appears to be normal:

I googled part of Dave's comment on the subject, and found this:

http://www.google.ch/patents/US20120152252

in case you're interested.

Haven't had time to fully digest it, but the stuff in the link does look very interesting.

it's on my 'to be comprehended' list... and following some of the links to other referenced patents yields even more potential gold.

[green wings]

Ha ha! Hey, you have very sophisticated dreams.

On my third night of PAPing I dreamed I was an intelligent goose that realized that I was being "fattened up" for fois gras by being force fed more air than I could ever possibly manage to breath in on my own. And I knew (in the dream) that there was nothing I could do to prevent my ultimate fate of exploding into little fois gras hors d'oeuvres at some rich person's cocktail party.

That's one of the most disturbing dreams I've ever had in my life. And I woke up feeling like I'd swallowed a basketball.

[eseedhouse]

If your high setting is 10, that is only about 1% of the normal air pressure at sea level.

[palerider]

10cm/h2o = 0.142233 psi

[vsheline]

Thanks for this info, RobySue. Is information about the PR Auto's algorithm available, either in text form or in equation form, online?

There used to be some pretty useful information about the PR Search algorithm on the PR pages that were accessible to patients. They seem to be long gone.

If I get a chance I'll try to find some links that are still live. Some of the links, however, may point to sites that are no allowed by the forum rules.

Here is an informative paper (which I think robysue posted in an earlier thread) which summarizes the ResMed, Philips Respironics and DeVilbiss APAP algorithms in Table 2. It also summarizes other algorithms like standard bi-level and ASV.

http://www.ncbi.nlm.nih.gov/pmc/articles...rt=classic

[green wings]

Thanks, Vaughn.

I have seen recommendations in this forum to set minimum pressure at about 2 cm below either 1) median pressure or 2) 90% or 95% pressure. This article seems to lean towards the second option.

If a patient enters REM sleep or changes position, the degree of obstruction may suddenly increase and by the time the device is able to adjust to the needed pressure the patient may have had desaturations or arousals. This is why most studies reporting the equivalence of AutoCPAP to in-lab titration recommend changing the EPAP minimum to the pressure the device is at or below 90–95% of the time.2,15 In our experience, many patients left on AutoCPAP 4–20 are undertreated and may present with awakenings a couple hours into sleep, residual symptoms, or difficulty tolerating PAP. Some patients are sensitive to the pressure changes, so if patients are not doing well with AutoCPAP, fixed CPAP should be tried.

Not sure if I'm interpreting the above paragraph correctly. To me, it's saying that if a patient has greatly different pressure needs in REM sleep than in NREM sleep, then the auto-adjusting algorithm can't respond adequately to this.

I draw that conclusion because if a person has much higher pressure needs during REM sleep than during NREM sleep, their 90-95% pressure value is going to be high enough to prevent apneas during REM sleep.

P.S. In the two months or so since this thread was begun, I have grown used to the pressure changes from my machine. I'm now sleeping through the night with a pressure range of 11.0-15.5 cm. (I discovered that the auto-trial function on my machine can be reset 5 times, so I have 180 days instead of 30 days of auto-trial.)

If I take the advice from this article and set my minimum pressure at my 90% pressure, I'd be using settings of 14.0-?

[green wings]

I am replying to myself here, but I just realized that the recommendation in the article I quoted in my previous post to set minimum pressure at the 90-95% pressure refers to a set of data collected with a 4-20 range on an APAP machine.

[ger365]

I am wondering if the 1.5 cm sawtooth-shaped changes in pressure that I see on this graph are normal behavior for a PR System One CPAP machine in auto trial mode.

the respironics machines seem a bit retarded at times... this is one of them.

if you look at enough breathing/pressure traces, what you'll find is that when everything is calm, quiet, breathing is very peaceful and regular, the machine starts acting like a bored kid... and starts poking at you. the explanation is that it's seeing if it can get better breathing at a higher pressure.... but it only does this when breathing is VERY regular and even.

if you look at enough breath traces, you'll see that if you're having irregular breathing, not even events, just ... not nice and calm and even... then the machine DOESN'T try to make things better, it just sits there, sucking on it's thumb.... but as soon as things get calm, it gets antsy and starts diddling with the pressure.

it's like there's a test in the routine, somewhere, that's backwards.

Once again, I feel the need to come to the defense of the PR Auto's Search Algorithm.

Things are not quite as simple or as backwards as Palerider is suggesting.

The Search algorithm does kick in when the breathing appears to be normal: The machine is looking for very, very subtle improvements in the flow rate data that fall well below the threshold of being scored as a "flow limitation" and the Search algorithm allows the PR machine to increase the pressure before significant flow limitations, snoring, or obstructive events can occur. The Resmed machines, on the other hand only react after the fact: A flow limitation or snoring or a cluster of events must be scored before the Resmed will increase the pressure.

For most people (but not necessarily the OP), the 2cm increases in pressure are small enough that the do not wind up triggering arousals or discomfort, provided the user is fully and soundly asleep. (And very regular, noneventful breathing usually indicates the user is indeed fully and soundly asleep.) The payoff for the Search algorithm is that the machine can proactively increase the baseline pressure enough to prevent or minimize clusters of events or serious flow limitations from occurring in the first place.

Next, Palerider is correct that when the breathing is inherently unstable, the PR Search algorithm is overridden and the machine no longer goes through the test searches. However, unlike Palerider implies, there is some real science behind PR's choice of programming the machine to not go into the Search algorithm when ragged breathing is being detected.

Here are two important reasons why it makes sense to override the Search algorithm when the breathing is already ragged:

1) Ragged breathing can indicate the person is awake or is in danger of waking up. In either case, the 2cm increase in pressure may prove to be disturbing and make it more likely that the person will arouse to a full wake or prolong a wake that has already occurred. This can happen even in patients who have no trouble sleeping through the test increases if they are already soundly asleep.

2) Ragged breathing can be aggravated by increasing the pressure. In other words, increasing the pressure can counterintuitively increase the amount of ragged breathing under many circumstances. The thing to keep in mind here is that unstable breathing is not always caused by a collapsing airway. Sometimes unstable breathing is indicative of a potential CO2 overshoot/undershoot cycle. Sometimes it's indicative of restless sleep caused by other non-respiratory events. Sometimes it's caused by other respiratory causes that have nothing directly to do with OSA. PSGs have shown that unstable breathing that is NOT due to OSA can become more unstable if the CPAP pressure is increased. And the more unstable the breathing becomes, the more likely the person is to wind up having an arousal of some sort that may not be due to OSA-related events.

In other words, the PR programmers understand that the protocols for manual titration of OSA patients include provisions for not increasing the pressure in the presence of unstable breathing that does appear to be directly caused by the OSA.

It's also important to understand that the rest of the PR Auto algorithm is still in effect when the Search algorithm has been suspended due to the presence of ragged breathing: If flow limitations or snoring are detected during the ragged breathing, the machine will increase the base pressure by 1cm and wait for a minute or two to see if the snoring and/or flow limitations cease; if not, the machine will again increase the pressure by another 1cm and then wait. Likewise, if two or more Hs or OAs are detected within a minute or two of each other, the PR machine will increase the pressure by 1cm and then wait for a minute or two to see if any additional events occur.

Finally the PR Search algorithm creates another big difference between the two algorithms:

A Resmed machine just starts lowering the pressure as soon as whatever triggered the pressure increase has been resolved, and it continues to lower the pressure until more snoring, flow limitations, or obstructive events occur, at which point, the Resmed machine once again aggressively starts raising the pressure.

The PR machine, on the other hand, does NOT start lowering the pressure right away. It leaves the pressure at the new level for 10 minutes or so and then it runs the Search algorithm in reverse: It tests a reduction in pressure and if any slight (sub "flow limitation") deterioration in the shape of the flow rate data is detected while the pressure is being reduced, the machine immediately raises the pressure back up to the level that had the best flow rate data. This reverse Search algorithm prevents the machine from lowering the pressure back down too soon and/or too far. And that too can prevent future events from happening.

+1