I’ve been managing obstructive sleep apnea for several years and I’ve learned a great deal over time. One of the more recent things I’ve come to realize in the past couple of years is that CPAP machine stats aren’t the only factor that should be used to determine how effective a treatment is working.

If you take the time to browse CPAP patient forums you’ll soon learn that the various machines calculate statistics in different ways. Between two different models in the same line I can estimate there’s at least a difference of two or three units for the value of the nightly AHI numbers. For example, my wife’s S9 appears to be lower than my own S8. Obviously I can’t account for the difference in severity between our cases of sleep apnea but I’m fairly certain, based on forum posts, that the S9’s reported numbers are typically lower. In my opinion, lower values reported by the S9 make it more difficult to pin-point problems because it offers a narrower range in which swings can be detected.

There are several factors to consider. For example, though an AHI may appear low this can be misleading if the leak rate is very high. I consider the AHI value more reliable when I have a very low leak rate.

One should also be mindful of the fact that home devices do not track the same array of data that is gathered in a sleep study. It’s possible that some information won’t reveal problems that might be obvious when compared against data collected in a lab setting (O2 levels, sleep stages, etc).

Sleep stats aren’t enough and simply don’t reveal everything. The quality of the sleep isn’t something that I can track at home. Yes, I can see if there were severe problems with leaks or high AHI values, but my machine can’t really track sleep stages (these can only be inferred to a minor degree) or the quality of my sleep.

I think most experienced CPAP users will agree that statistics are helpful but the most important factor for determining effectiveness is simply how good you feel in general.

With my machine I’ve learned that I’ll feel alright with an AHI below 3 and I typically feel very good if it’s below 2. Anything consistently above a 4 will begin to wear me down. Note that these numbers are well within the “normal” range.

Updated 06/25/2012: But the stats can be very helpful at times as well. If you look at the pressure a machine is using to stop events then you may figure out that your lowest pressure setting should be increased. Over in relevant forums many users have stated that what often happens is that a machine doesn’t ramp up to the necessary pressure in time to stop many events. For example, if your minimum pressure is 6 (with a max of 15) and the majority of your logged events require a pressure of 12 then it’s possible that there are several events at or above that pressure and the machine simply isn’t ramping up enough in time. For example, if the machine is at 6 and only reaches 9 before the event naturally ends (your brain tells your body to breathe, thus disrupting your sleep) then it may not be effective enough. In such a case it may be wise to have your lowest pressure increased to a value closer to the average pressure needed to prevent events.