I assume OA and H events cause the system to raise the pressure on a schedule until apneas/hypopneas decrease (or stop), but then why would it back the pressure down after it finds the level that is working well.
Too many CA or other proprietary signals may cause the machine to decrease the pressure until the OA and H events start to increase.
However, looking at my Sleepyhead data, I can see times when the pressure fails to increase more, even though events continue, and other times, when the pressure falls after being at a plateau for no apparent reason.
My Dreamstation seems to optimize to maintain the lowest pressure, not the most effective pressure at reducing AHIs.
Also, it seems counter-intuitive to start at the lowest pressure each machine start. Why not start at at the mid-point between the upper and lower limits, or even better, the last effective pressure - and then decide to go up or down?
I believe the reason why most people have to keep adjusting their lower pressure limit in APAP (at least for PR machines) is that the machine keeps starting there. From an engineering standpoint starting at the midpoint of the range, or even a random point between the limits, is bound to be "wrong" less of the time than starting at a limit (that's why they are call 'limits'). (Maybe some manufacturer will owe me royalties for one of those ideas in the future.)
Any thought's? Do other APAP machines start differently or do they all start at the lower limit, each restart?