discrepancy between therapy hours and blower hours?

I am fairly new to this. I have only had my cpap since 11/9. Generally, the therapy hours and blower hours are roughly the same, but I have had 3 days where the difference doesn't make sense to me. I always turn the machine off before I take off my mask when I get up for any reason. You all have so much experience and knowledge so I am hoping someone can explain this to me. I imagine there is probably a simple explanation. Thanks!

TH / BH
8.54 / 9.60
6.06 / 7.80
7.06 / 9.40

I don't have a PR-1 machine. I have a Respironics M-Series AutoCPAP with AFlex. A PR-1 machine reports differently (for example, it offers a local time offset option) so what follows may not apply to you.

For my machine, the Therapy Usage hh:mm/Night values displayed on the LCD screen are 1) for the machines Noon-to-Noon day which operates under GMT/UTC time (London time) and 2) are rolling 7-day and 30-day averages. Therapy Usage hours may, for example, differ from the EP/EV/EPA software reported hours (7-day and 30-day averages) as the software reported hours have been adjusted to report on a Noon-to-Noon day based on local time. On my machine, the Therapy Usage Hours can be reset to 0.0 which allows a user to see a nightly average for AHI, etc. Please note that, with the reset, my machine's LCD reporting day is still Noon-to-Noon GMT/UTC time.

On my machine, LCD displayed Blower hours and Therapy Hours (not the 7-day and 30-day average #s) are cumulative hours. I have observed on both of my machines (my backup is the same model) that my Therapy Hours slowly lose "time" compared to my Blower Hours. The machine that I am now using (I switch machines every 3 months or so) shows TH = 886.4, BH = 890.4, and Sessions = 110. So over 110 Sessions, my TH accumulated count has lost 4.0 hours compared to the BH count or 0.0364 hours/night which is equivalent to 2 minutes/night and a discrepancy of 0.4%. I have not investigated this small difference. I suspect that the difference is simply that my Respironics machine uses two differently written/designed/implemented approaches to accumulate the TH and BH Hours. If someone has looked at this discrepancy and reached a different conclusion, I would appreciate learning the analysis details.

Two possibilities spring to mind:
1. Some machines don't count the first 15 minutes or so of blower time as being "therapy time" because you're probably not asleep yet. PB/Covidien machines definitely do this and I think Respironics also. So that would mean your blower time would be 15 minutes longer per session than therapy time.

2. If you pull the mask far enough off to be really leaking badly, that might not count as therapy time.

Two possibilities spring to mind:
1. Some machines don't count the first 15 minutes or so of blower time as being "therapy time" because you're probably not asleep yet. PB/Covidien machines definitely do this and I think Respironics also. So that would mean your blower time would be 15 minutes longer per session than therapy time.

2. If you pull the mask far enough off to be really leaking badly, that might not count as therapy time.

1) Respironics machines do count the first x minutes of blower-on time as Therapy Hours - please note that my machine only shows an average of 2 minutes/night discrepancy. Many nights do not show any discrepancy (0.0 hours difference) between Therapy Hours and Blower Hours. I do not use the Ramp feature but suspect that it does not play a part re TH and BH accumulation.

2) I have low and flat leak rate lines and only experienced "large leaks' during my initial two weeks of CPAP-enabled therapy (late March & early April '09). At a large enough leak rate (likely will be reported as Large Leak by the reporting software), I suspect that Therapy Hours will not accumulate for the time spent in Large Leak (thus one of the purposes of our machines reporting each category - the other one being if the machine does not detect breathing - we are not connected to the hose). Since I do not experience even High leak rates let alone Large Leak rates, I have never observed this directly.

I occasionally see issues like this from my antique. I always assumed it was a bug in the code, reading data from a previous day that didn't get cleared out correctly.

Thanks for the ideas.

The machine came with a 20 min. ramp set up but I haven't used it since the 2nd day. I felt like I couldn't breathe well enough at the lower pressure. I set the machine to local time after the first night. There was a 2.6 hour discrepancy the 2nd night that I initially attributed to the reset.

I reset the data on a daily basis (after popping out the data card temporarily) so it isn't a cumulative difference. I have 17 days of data. Some days the #s match exactly. Some days the difference is minimal, and then some days the difference is much larger. The days with large discrepancies are scattered among the others and are not offset by corresponding differences in the other direction on the following days.

The machine has reported 0% system leaks all along, but I don't have access yet to any info about other leaks. Maybe that is what it is.

I do pull the mask away from my face to readjust for small leaks created when I move, but don't think it could add up to more than 2 hours in one night.

On at least two of these days I took a nap and wore the mask then too. Could it be that the nap period was relatively short so it doesn't count as therapy?

I have a follow-up appointment with the sleep doctor in 2 weeks. I am going to have them give me a printout from the data card. Maybe there will be more information on it that will explain the discrepancy.

In any event, I am feeling better since I have been using the machine, so I probably should stop obsessing over the numbers, but......

I suspect that your machine reset, time offset, and napping are interdependent contributors to the discrepancy that you are observing. You may want to run an experiment.

Proposal

Don't reset the machine (leave the time offset alone and continue to nap when you need to) and observe the cumulative numbers for more than a few days. Report your numbers in a subsequent post. Thanks.

PS: Write down the starting and ending times for your sleep sessions, both your nightly sleep sessions and your naps. Include this information in your post.

cinco777 wrote:I suspect that your machine reset, time offset, and napping are interdependent contributors to the discrepancy that you are observing. You may want to run an experiment.

Proposal

Don't reset the machine (leave the time offset alone and continue to nap when you need to) and observe the cumulative numbers for more than a few days. Report your numbers in a subsequent post. Thanks.

PS: Write down the starting and ending times for your sleep sessions, both your nightly sleep sessions and your naps. Include this information in your post.

OK, I'll try not resetting the data for a while. So far, my sleep times have corresponded to the blower hours. I will include them in the future post with the cumulative hours as requested. I'm not sure how long it will take since these discrepancy days are sporadic. For clarification let me say that I napped on another day and did not have a discrepancy and The GMT offset setting change only had to be done once and has remained the same ever since. I don't see how this would impact the numbers on a sporadic basis but I'm willing to give it a go.

Of course I will still want to know the daily info so I will just have to do more math to manually break the numbers out of the averages. ah well...

There is a possibility that your machine is broken. However, before reaching such a conclusion, it is worthwhile to investigate - thus my proposal for an experiment.

Thank you, Bailachel, for your fascinating report. We don't have much experience yet with the PR-1, so you're helping to explore new territory.

Several thoughts come to mind as possible explanations for "blower time" exceeding "therapy time". (It would be much more problematic if the discrepancies were in the other direction.) Again, since I have no experience with the PR-1, these are just WAG's (Wild-Ass Guesses).

Your mention of occasional naps leads me to wonder: might the "therapy time" be updated on a "noon-to-noon" schedule, but the "blower-on" time be always immediately current? It would make some sense to do this: daily "therapy time" needs to be flexible to include all sessions within a "day," while there's no reason for cumulative "blower time" not to be always instantaneously accurate. Could you be looking at data at a time of day when the "therapy time" for your nap has not yet been added to the total, but the blower time has been? If this is the case, then one might expect a discrepancy in the other direction the following day, though perhaps not due to the "reset".

I don't know whether "ramp time" counts toward "therapy time" but in any case, you say you don't use it.

You mention 0% "system leaks." I don't really know what this is, unless you might mean what with prior machines has been termed "large leaks". This too may not apply to you, but is potentially relevant generally, since a "large leak" situation (or any situation in which the machine can't detect or recognize a normal "breathing" pattern - e.g. mask off, or using the machine to dry out tubing) would be counted as "blower time" but not necessarily as "therapy time."

Although you do not seem to be using monitoring software, you do describe having a data card in place. There are many reports with the older Respironics systems of losing entire nights of data, and I have experienced losing "fragments" of data during a single night. It is not clear what causes such data problems. You seem to describe several on-off cycles during the night: it's between such shut-downs that I have occasionally lost "pieces" of data. This may be another possible explanation of your situation, in that your machine occasionally loses "therapy time" data, and that this is reflected not only on the data-card / software systems which have been previously described, but also within the machine memory and LCD display subsystems as well. I sure hope this isn't the case; I have never really looked at this aspect.


Cinco777: My supposition is that "blower time" is taken directly and immediately from "state change" flags which are labeled to the nearest second. Therapy time, as you point out, is often a few minutes less each night; my supposition is that this data is taken from the 30-second time-stamped data; perhaps "therapy time" begins to be counted only when the 30-second mean pressure actually reaches the therapy minimum set pressure, and stops only when the mean falls below it, and is accurate only to the nearest 30 seconds. If several starts and stops per night are experienced, this difference might accumulate and account for the small nightly discrepancy. I must confess to never having looked at the LCD data in this regard, having worked only from the Encore source data.

Regards, Velbor

Well, now I have a new wrinkle to add. a little while ago I looked at the data from last night. When looking at the individual categories on the LCD, the way I normally do, therapy hours = 7.0 and blower hours = 9.1. I slept from approximately 11:30 til 8:35, just over 9 hours. When looking under "phone in" data, which I don't normally do, both blower hours and therapy hours = 9.1. (BTW no nap)

Velbor, I do have a data card but there isn't any software available yet for the PR System 1 so I will have to wait until I can get someone at the DME or sleep clinic to download it for me. I am guessing the "system leaks" and "large leaks" are the same thing, but this is the only machine I have ever had so don't know for sure. You may be right that the reset is erasing the offset in the other direction. I am going to follow cinco777's suggestion of not resetting for a while, so if that is the case, should I expect that therapy hours should have a 2.1 hour "head start" over blower hours tonight based on last night's numbers?