(07-07-2014, 07:53 AM)jcarerra Wrote: (07-07-2014, 06:40 AM)archangle Wrote: ..."Real" AHI as used in treatment is usually the nightly average apnea/hypopneas per hour over your sleep time. "Hourly" AHI is less useful in general, but has some uses. "Hourly AHI" as they show it there is sort of only useful at the last reading within the clock hour. i.e. the 11:59, 12:59, 01:59, etc. readings. The way SleepyHead does it is better, but "THE" AHI reading of interest is still mostly the nightly average.
"THE" AHI ...
You sleep 8 hours.
From 3-4am, you have 40 events.
none the rest of the night.
Your "THE" AHI is 40/8 = 5.0
So, "you're OK."
I say NO!!! you are not ok, even though your AHI says you are. I say that those 40 during one hour are a problem that needs to be looked at--but the AHI average HIDES the fact that there is a problem.
This is not some far-out scenario that never happens;
it exactly describes what often happens with me.
As with many things, outlier data can do all kinds of interesting things.
So to revisit in the scenario posted: An 8-hour night of sleep with 40 events that all occur in 1-hour.
Yes, the overall AHI = 5 (borderline "ok") and but the "AHI for 3-4 am" is 40 (way bad). But the person also GOT 7 hours of presumably high quality sleep both before and after the bad hour on this bad night, so it's not like the person has a AHI = 40 for the whole night either: A person who's
diagnostic AHI = 40 is getting 8 hours sleep with an average of 40 events/hour, or 320 events all night long. And no good sleep.
And so an important question to ask when dealing with really nasty clusters of events is: Does this kind of a pattern occur rarely or frequently?
If this 40-event hour is the only bad hour in a month of PAPing, it's not really all that important: Almost all the sleep is good-to-excellent and PAP is working at its current settings.
But hour-long clusters of 30+ events occur on several nights every week for a month or more, that does indicate that something's not working as well as it should.
And jcarerra writes:
Quote:This is not some far-out scenario that never happens;
it exactly describes what often happens with me
So clearly jcarrea should be more worried about this kind of a cluster than I am: I've had a few clusters of 20+ events in 3 1/2 years of PAP, and they're well separated in time and are usually connected to the flu. Jcarrea, on the other hand sees enough of these clusters to say they
often happen to him.
In my case, there's no point in bringing in Daily data that shows one of these clusters and asking: What's wrong? What should we do about this?
In Jcarrea's case, I think it's critically important to bring in a hard copy of several days of data that show this kind of cluster and directly ask the doc: What's wrong? What should we do about this?
But I don't think it is reasonable to expect a doc to automatically look beyond the overall long term AHI if the patient is not making some effort to indicate there's something wrong.
The main reason sleep docs focus on the overall,
long term AHI is simply time: It's just plain too much trouble to go digging into the nightly data for the dozens of patients you see each day or the hundreds you see each month. So looking at one basic number (AHI) is a bit of a survival technique: It lets you get patients in and out more quickly. I'm not saying I agree with this rationale, but I do acknowledge that it's an important reason that docs are a lot less interested in our data than we are.
But there's another more valid reason for focusing on the long term AHI for the
typical OSA patient: We're all going to have some good nights and some bad nights. And a few very bad nights. But a few bad nights here and there is NOT usually an indication that something is "wrong" and must be fixed. In other words, if we're getting enough apnea-free sleep on the vast majority of nights, there's not much that tweaking of the PAP settings can do in terms of addressing continuing patient complaints of daytime fatigue and sleepiness. Not all
bad sleep is related to OSA and non-optimal PAP therapy.
But again, in Jcarrea's case, the clusters are not confined to a few bad nights "here and there", and so it's worth hitting the doc over the head with the Detailed Daily data and bringing up the issue of what to do to fix the problem.
Final note: I'm not a big advocate of simply telling people like jcarrea: The data says you need to increase your pressure. Hour long clusters of events are NOT common once someone is on PAP. REM-cycles don't last that long. And when the clusters of events really do last 30-60 minutes or more each night, there's a real possibility that part of the problem may be pressure-induce centrals
even if the machine scored events are labeled as Hs or OAs. And teasing apart PAPers who are having pressure-induced unstable breathing and centrals from PAPers who just are inadequately titrated and need more pressure is beyond my expertise as a knowledgeable patient.