RE: Please help me understand ......
PaulaO2 does a good job of explaining one of the issues with wide open settings. But it's not the whole story.
There are actually three parts to your question. The first is why would you keep the lower limit at 4 rather than raising it up. PaulaO2 explains this quite well. One other point that was not mentioned is a peculiar effect: when you are used to a higher pressure, a lower pressure feels like there's "not enough air" and you become uncomfortable until the pressure is raised. Most people that become accustomed to their pressure completely stop using the ramp feature on their machines. [ That same ramp feature is invaluable when you are first getting used to the treatment. ]
One reason why you might (temporarily) want to keep your lower limit below the minimum you need for apnea suppression is that it will act like a "sleep-dependent" ramp feature. The actual ramp feature on machines will linearly increase the treatment pressure on a fixed-time basis. This has the disadvantage that, if you take a while to get to sleep, the ramp might finish before you fall asleep, forcing you to fumble for that ramp button to restart it. However, until you actually fall asleep, you won't experience any apnea events, so the auto algorithm will keep the pressure at its lower limit. Then, when you fall asleep, the auto algorithm will increase the pressure appropriate for treatment.
The second part to your question is why you would keep the upper limit at 20 rather than lowering it. And the third (implicit) question is why would you have settings that are below what might be required to eliminate apnea events.
The answer to these last two questions is straightforward: compliance. By far and away the greatest cause of CPAP treatment failure is because the patient stops using it. Now, if you have been newly diagnosed and have a "high" pressure setting (say 17 cmH2O), you have a huge hurdle to overcome. Not only are you going to have to find a comfortable mask (which is a task in itself), but you are going to have to cinch it down quite tightly to stop it from blowing completely off your face. Plus, you have what feel like gale force winds blowing air at you. Not very relaxing. You have a real uphill battle that a lot of people simply give up on.
Now imagine a more gradual response. Forget trying to solve the apneas all at once. Start off with a lower pressure, say 8 cmH2O, and get used to that. Get used to a mask on your face. [ This is actually a huge hurdle for some. ] Get used to exhaling against pressure (yes, you do get used to it). Then, as you become less uncomfortable with that (I would say "more comfortable", but let's keep it real here), increase the pressure. Eventually, you're at your ideal pressure and you can actually fall asleep and stay asleep without too much trouble.
That is why you might want to keep your upper pressure limit less than the wide open setting of 20 cmH2O.
Having said that, once you get used to the higher pressures, if you don't trust your machine's auto algorithm, you might want to put in a fail-safe upper limit on the pressure. If you do trust your machine's auto algorithm, I see no reason why you would not increase it up to the wide-open setting. Either you need that higher pressure (in which case why not provide it), or you don't (and the machine would not provide it).
Hopefully this answers your question.
Now, looking at your specific case, your average pressure values are down at the 4.8 - 5.5 values. There is no point in providing a pressure higher than what is required to prevent your apneas, and 4.8 - 5.5 cmH2O does prevent your apneas at times. Hence, I would not increase your lower limit of 4, or at worst only increase it to 5. That is actually frequently sufficient for you. As to whether to decrease the upper limit, you have actually proven that your machine does not "run away" with too high a pressure, so I would not bother. As time goes on, perhaps the pressure you require will slowly increase (it happens as we age), and this will automatically compensate.
(This post was last modified: 07-09-2013 04:05 PM by RonWessels.)