ResMed epr algorithm

[bavinck]

My understand is that epr simply acts like a weak bilevel, reducing the exhale pressure based on 1, 2 or 3 cm. My dme told be it works in inhale and exhale, so at 8cm an epr of 3 would exhale at 5 and inhale at 11. Is he correct? If not, does anyone know of a link I can send him from resmed to help him out?

[Pugsy]

so at 8cm an epr of 3 would exhale at 5 and inhale at 11. Is he correct?

Not correct.
At 8 cm pressure with EPR set at 3 the inhale pressure is 8 cm and the exhale pressure is 5.
With the cpap/apap machines and EPR the difference between inhale and exhale can never be greater than 3...
impossible to have 11 inhale and 5 exhale with cpap/apap single pressure machine as that would be 6 cm pressure support and the only way to get that is with the bilevel pressure machines...and they don't have EPR available in bilevel mode.

Sorry, don't have any pertinent links to documentation per ResMed....never needed any.
The clinical manual should be sufficient for most people to understand how it works.

[bavinck]

Thanks, I will check the manual. How would epr work on a bilevel machine? Wouldn't the ability to specify inhale/exhale pressure negate the need for an epr function?

[robysue]

Thanks, I will check the manual. How would epr work on a bilevel machine? Wouldn't the ability to specify inhale/exhale pressure negate the need for an epr function?

The Resmed S9 VPAP's do NOT have any EPR setting because it is not needed. The IPAP/EPAP are controlled by separate settings and the only pressure relief you get is the drop from IPAP to EPAP, which can be set as large as needed.

[bavinck]

Robysue, where did you learn so much has about the resmed/PR auto algorithms?

[robysue]

I've done a lot of reading on a whole bunch of sites and I've done a fair amount of googling academic papers that refer to one or both algorithms. As a professor, figuring out how to find what I want to learn about is part of the job.

[bavinck]

For sure. I am a bit surprised that you found published works on their algorithms as I would have assumed they are proprietary and therefore very secret.

[robysue]

For sure. I am a bit surprised that you found published works on their algorithms as I would have assumed they are proprietary and therefore very secret.

Thing is, they have to prove the algorithms work in practice as well as theory. Hence, there are a fair number of white papers that the companies have released. They're not always easy to find, but they are out there.

But the real gem that I found was an academic paper:

A Bench Comparison of Five Auto-Adjusting Positive Airway Pressure Devices, Response to Apnea, Hypopnea and Flow Limitation, by Robert McCoy BS RRT FAARC, Ryan Diesem BA, Valley Inspired Products Inc. (December 1, 2008)

Unfortunately, the link that I had for the paper is no longer active. While this is an older paper that specifically looks at the Resmed S8 and the Resprionics M-Series (among other APAPs), the basics of both the Remed and PR algorithms and how they respond has not changed all that much with the new generation of machines.

The main changes in both cases are the central apnea detection algorithms, which back in 2010 were receiving a lot of PR exposure on the Resmed and PR web pages, with links to studies in the companies' publicly accessible web pages directed towards clinicians. Unfortunately the US Resmed pages in particular are a lot less informative now that they've been reorganized than they were back in 2010 and 2011. The PR web pages are also being resdesigned and it's becoming more difficult to find useful information on those pages as well.

[bavinck]

That's great to know. I sure would love to read some technical papers on Apap, bipap and Asv algorithms. Thanks for pointing me in the right direction.

[JDS74]

Here is a link to a similar study:

http://erj.ersjournals.com/content/24/4/649.full

Included are references to a large variety of studies, many of which are available in full text.

And here is his resume as of 2009:

http://www.ltotnet.org/BobMcCoy%20Reusme%2002.03.09.pdf

[robysue]

Here is a link to a similar study:

http://erj.ersjournals.com/content/24/4/649.full

Included are references to a large variety of studies, many of which are available in full text.

And here is his resume as of 2009:

http://www.ltotnet.org/BobMcCoy%20Reusme%2002.03.09.pdf

A great big thank you for providing me with the links!

[DKC_apnea]

Couple little things to note with EPR on ResMed S9
EPR can be set to "Ramp Only" , "Full Time", or "Off". The "exhale" pressure relief can be 0, 1, 2 or 3 cms. EPR remains constant during the ramp, but the inhale pressure will climb from whatever start pressure was set to the set pressure once the ramp is completed. Keep in mind the machine will always attempt to deliver at least 4 cm of pressure, so if the EPR is set to 3, it will be unable to deliver the full EPR until the inhale pressure climbs to at least 7 cm.

Personally I'd like to a couple of new options/preferences to be added to the EPR.
1) Pressure can be adjusted in 0.2 cm increments, so why not allow that same fine tuning to EPR?
2) Might be beneficial to let the EPR value "ramp" also (i.e. allowed to gradually decrease during the ramp time). For example, my set pressure is 8, and I use an EPR 3 during ramp. Well at the end of ramp the EPR drops immediately from 3 to off.

[robysue]

Here is a link to a similar study:

http://erj.ersjournals.com/content/24/4/649.full

Included are references to a large variety of studies, many of which are available in full text.

Again, a big thank you for this link. It's a fascinating paper, even though this paper is from 2004. (McCoy's paper was from 2008.)

It's clear that both the Resmed and Respironics auto algorithms have under gone some changes, but it's also clear that both companies have stuck to the most basic part of the their algorithms: The Resmed's rapid pressure increase in response to events is already there as is the more measured PR response. The other thing that caught my eye was the stuff about how the machines handled really large unintentional leak rates of 0.8 L/sec = 48 L/min. The 2004-era Resmed was clearly unable to deal with a leak of that size, but the 2004 Respironics could. No wonder Resmed sets their Red Line at 24 L/min.

I'd love to find a much more up-to-date bench testing of the current algorithms. I'm sure it would have some surprises.

[Sludge]

http://journal.publications.chestnet.or ... FSource=13

http://www.aasmnet.org/jcsm/ViewAbstract.aspx?pid=29104

http://www.rtsleepworld.com/whitepapers ... -algorithm

http://www.rtmagazine.com/2009/08/auto- ... -pressure/

[Sludge]

But the real question is:

Can The Chrome do it?