(12-13-2013 11:49 AM)jgjones1972 Wrote:
(12-13-2013 09:34 AM)Sleepster Wrote: APAP's will indeed wait until you've had an event before raising the pressure in response to them. This is a fact that the manufacturers don't advertise.
I don't think this is true of many of the new APAPs if you define an "event" as an apnea or hypopnea (as per accepted definition of apneas and hypopneas). I sometimes have an AHI of 0.0 even though my pressure raises and lowers all through the night. If I look at the flow chart I can surmise that the pressure increases are usually due to plateaus forming in the flow waves (increased FL) and/or a slight increase in vibratory snoring.
jgjones is correct here. Modern APAP Auto algorithms are designed to analyze the wave flow data pretty carefully to detect both flow limitations and snoring. The PR System One machines also have a proprietary algorithm for detecting wave flow patterns strongly associated with RERAs even though a real RERA requires an arousal in the EEG data and the System One has no way of definitively telling whether a flagged RERA is "real" or not.
Unlike the Resmed S9 machines, the PR Auto machines also have a "search" algorithm for attempting to determine the optimal pressure setting even when the sleep breathing pattern appears to be stable. The "search" algorithm has two main parts: A "test pressure increase" part that is used to proactively increase the pressure even before a FL can be scored and a "test pressure decrease" part that prevents the machine from lowering the pressure below a therapeutic level. It's this "search" algorithm that gives the pressure curve of PR Autos their characteristic sawtoothed shape.
The "test pressure increase" part of the PR "search" algorithm
When the sleep breathing appears to be stable, the PR Autos will proactively increase the pressure by 2cm over about a two minute period on a periodic basis. If no improvements in the shape of the wave flow are found after the test pressure increase, the pressure is lowered back down to the current baseline level over about 60 seconds or so. If improvements are found during the test pressure increase, then the PR System one continues to increase the pressure at a rate of 1 cm per minute as long as improvements continue to be found with the increased pressure OR the maximum pressure level is reached. At the end of the pressure increase, a new baseline pressure level is established at a new higher pressure level. The new baseline is typically 0.25-0.5cm less than the maximum tested pressure during the test cycle OR the max pressure setting if the pressure was increased that far.
You can read a bit about the PR's search algorithm at: http://www.healthcare.philips.com/us_en/...orithm.wpd
The "test pressure decrease" part of the PR "search" algorithm
The PR Auto machines use a similar "search" algorithm when they are attempting to decide whether it is safe to lower the pressure back down and to determine how far to lower the pressure. Once the breathing is stable at a given pressure for a long enough time, the machine will tentatively lower the pressure about 1 cm over the course of a minute. If any deterioration in the shape of the inhalations is detected, the pressure is then increased back up to the previous setting; if there was no deterioration in the shape of the wave flow detected during the test pressure drop, the machine will continue to decrease the pressure by 1cm per minute as long as no deterioration in the shape of the wave flow is detected. The pressure decrease will end whenever the min pressure setting is reached OR when the machine detects a slight deterioration in the wave flow, at which point it will increase the pressure back up by 1cm.
Other notes on the PR "search" algorithm
I believe that the PR "search" algorithm also has an additional part to it that determines which way to make the test changes in pressure when the new baseline pressure is something that is between the minimum and maximum pressure settings. In other words, it appears that there are times when the "search" algorithm decides to do a test increase in pressure starting with the new (higher) baseline pressure and there are other times when the "search" algorithm decides to do a test decrease in pressure starting with the new (higher) baseline pressure.
I have not been able to sort out exactly how the PR Auto algorithm makes that decision because I almost always have my BiPAP in such a tight auto range: The difference between my min IPAP and max IPAP is only 2 cm most of the time and hence when Kaa finds an improvement in the inhalations, the increased IPAP is at my max IPAP and there's nowhere to go except for testing for decreases. (On a BiPAP, the "search" algorithm is applied only to the IPAP pressure.)
With my sleep doc's permission I was doing some autotitration earlier this fall where I did up the max IPAP and max PS settings to see just where the machine wants to take my pressures and whether using a bit more pressure would lead to better sleep. The sleep doc was skeptical that I'd find a better setting than my current one, but told me to go ahead and satisfy my curiosity. In the end the net result was indeed that higher pressures don't necessarily bring my AHI or FL down any further than my current range; higher pressures DO aggravate the aerophagia; higher EPAP pressures aggravate the aerophagia more than higher IPAP pressures do; and overall allowing the machine to increase my pressures further did nothing to increase the quality of my sleep.
But in going through that data I can find some really good illustrations of just how the PR search algorithm works. I'll put those in a separate post.