Throughout all the talks and papers on CPAP usage at this year's SLEEP 2010 Meeting, it seems like the word compliance is being used less and less, and instead, sleep researchers are using the word adherence. Unfortunately, the practical meaning hasn't changed—it describes how long users are able to use CPAP on a nightly basis. Various researchers use different definitions, but one definition that is commonly used (since it's what Medicare uses) is at least 4 hours per night in at least 70% of nights in a 30 day period. So if there are 56 total hours in a week if you sleep for 8 hours, on average you have to sleep at least 20 hours per week on CPAP to be "adherent." That's about 35% of your total sleep time.
Considering that most people aren't able to use CPAP at all (published "adherence" rates range from under 20% to over 80'%), this is a very liberal criteria for what is considered the minimum time that's considered "acceptable" by the sleep community.
Even if you're considered "adherent," it's assuming that you're receiving optimal CPAP pressures with no leaks. So if your final AHI at your "optimal" CPAP pressure is 16, then if you use your CPAP machine enough hours per week, then you're considered "adherent."
There has to be a better way to measure how well and for how long people use their CPAP machines. We should also incorporate the quality of CPAP use, in addition to quantity.
What do you think about these definitions?