Resmed S8 EPR

[-SWS]

Yeah, but that's not "event handling", that's waiting for an event to occur, ....

Even if that's all that's happening, then I'd still define the EPR portion of the algorithm as executing the following sequence: 1) event detection, followed by 2) pressure response, and therefore meeting the definition criteria of "event handling" (at the very least for ensuing cluster-based/associated events). But I'm not sure that's all that's happening at this preliminary stage of analysis.

However, toward the question of whether event-handling occurs in EPR mode versus being reserved for the traditional A10 portion of the algorithm, here's a text excerpt of cpap.com briefly describing EPR (not AutoSet) as having its own dedicated "event-handling" circuit:

Sleep Disordered Breathing Event Handling. EPR has an Event Detection Circuit. When a sleep disordered breathing event is expected or has occurred, EPR stops until the event concludes and normal breathing resumes. C-Flex is of such short duration it is thought that is has no negative effect on sleep disordered breathing events and does not need to stop.

In the above excerpt, cpap.com correctly associates EPR as having its own event handling circuit. And in that same publication, the excerpt below correctly states that event-handling EPR does not occur in APAP modality (I would note that in APAP modality evolved A10 event handling occurs in lieu of EPR's independent and unique event handling routines):

# Auto Mode Pressure Adjustments. C-Flex works when the machine is in the "Automatic Pressure Adjust", or "Auto" mode. EPR only works in "Constant Pressure", or "CPAP" mode. EPR will not work when an S8 Vantage machine is in the "Auto" mode.

I admittedly haven't had a detailed look at the EPR event-handling portion of the algorithm yet. However, I reserve the possibility that EPR BiLevel suspension may be more than the short-termed response that you described in the quote at the very top of this post. I reserve that possibility because I believe it is technically feasible for EPR suspension to also utilize trend-based pattern analysis and prediction. And that would certainly add a whole new respectable dimension to EPR event-handling.

I agree that all comments in this thread are pretty much theoretical and speculative in nature. I think EPR is going to be a very interesting platform to watch, regarding both medical publications and patient anecdotes. I don't see any EPR white flags just yet. But it's nice to have analysts in the crow's nest of "Ship Apnea", keeping an eye out for the rest of us... .

[StillAnotherGuest]

Well. algorithms are slick, but trying to make EPR work successfully, that's gotta be translated into practical application. And at the point where ResMed has decided to target this adjunct (initiation of inspiration) you need two things-- speed and sensitivity. If you can't provide that, and provide it in the face of common variables, you're gonna have an issue.

So anyway, back to that "obvious" thing. It struck me that the higher the flow rate of the machine, the less likely that it would be able to turn on a dime. It has to stop (or "freewheel") to allow exhalation and then start (re-engage) to respond to inspiratory effort. If it lags in the freewheel phase, you'll get expiratory overshoot (remember the "why do I have a higher than set pressure" question?) If in lags in sensing inspiratory effort, there may be excess negative pressure generated and compound airway instability.

So what governs machine flow rate. It's leaks. Leaks for any reason (whether it's a designed exhaust leak or leaks from ill-fitting equipment) will cause an increase in machine output in order to compensate (since the ultimate goal is to maintain set pressure). And if a large total leak is causing a large flow rate from the machine to be occurring, what has to happen at the precise moment that exhalation begins is that the flow rate has to undergo a rapid and substantial reduction (the freewheeling phase). If this doesn't occur, then there may be a pressure overshoot as the flow rate reduction lags behind the pressure that is created by not only the flow rate itself, but now the additional flow from patient exhalation.

OK, looking at the pressure waveform of a RM FFM with half the exhalation ports blocked and no extraneous leaks (for a net "lower than normal" machine flowrate) shows no pressure overshoot at all. Pressure during exhalation comes to 10 cmH2O. Uncovering the occluded pressure ports ("normal" flow rate), however, reveals a pressure overshoot of 1-2 cmH2O. CPAP 10 cmH2O, EPR 3:



Finally, adding in some mask leaks, and the blower cannot enter freewheel phase fast enough to adequately reduce flow and prevent a substantial overshoot, now nearly 4.0 cmH2O:



Now I don't believe this is news to ResMed. Reviewing the patent it is noted

By disabling the freewheeling technique at the end of the ramp phase the patient is provided with maximum comfort while trying to fall asleep (i.e., during the ramp phase) without the freewheeling technique interfering with the clinically determined therapeutic pressure during sleep (i.e., during the therapeutic phase).

Ooh, almost forgot-

Would a prudent approach be to experiment with different masks and pressure settings to see if AHI improves?

If there may be a clinical issue with EPR during sleep, then just it during ramp. Once you fall asleep, who cares what's happening? Otherwise, what kind of setting could you use? If you're set at 10 cmH2O CPAP with EPR 3 cmH2O, so you say, "OK, let's set it at 13 cmH2O", now the overshoot could very well be up to 16 cmH2O. And is it the nadir pressure causing the issues or the overshoot? I think there's too many variables to try to figure out what EPR is doing in sleep, and trying to "make it work" using single channel analysis is going to be tough, unless, as noted above, if EPR is a user-defined parameter, then try to ascertain if the modality itself is the source of the problem or not.
SAG

[kurtr]

SWS and SAG,

Thank-you very much. I appreciate it.

Kurt

[DSM-Guest]

SAG,

Now you have piqued my interest.

If EPR can cause a 3 cms overshoot when the machine is set at 13 cms, that seems problematical. I am now keen to get one to try the feature out.

The EPR machines have the S8 blower which (as for the S7 based range) uses variable speed to change CMS - in the S7 blowers (as fitted to S7s & VPAP IIIs) I don't see this kind of overshoot so am wondering why it would happen with the lighter S8 blower and its pressure sensing circuits.

I would be very surprise if this level of overshoot is happening but we are always learning.

Cheers & Happy New Year all

DSM