Debugging CPAP - No longer a patient patient...

butterfly masked 2

My new avatar that I’m using on Twitter and Facebook, derived from a butterfly photo I took at Natural Bridge.

Last night, I spent another night struggling to just sleep and breath at the same time.  One would think that since I’ve been on CPAP for years now and have been trying to fix it so it worked better off and on for a long time, that it would mostly be sorted out.  However, some of the problems with CPAP are not ones I can work around, and my patience is gone.

Now, I do sleep better with CPAP than without -- if I sleep without it, I wake up very groggy, with my lungs tired, and generally yucky (almost feel poisoned).  So, I’m not planning on giving up CPAP.  Yet even with CPAP, I don’t get good rest; I’m perpetually tired, and often wake up with a headache that I fear is from lack of oxygen.  I’m so used to being tired, I sometimes think I’m not when I am -- I catch myself yawning right after thinking that I feel awake.  I sometimes wonder if I know what “awake” really means.


The CPAP bugs are:

  1. Keeping the pressure constant means that when the resistance in the airway increases, the flow decreases -- basic physics.  This is backwards and so flawed, I’m not sure how this was missed before.  I’m guessing that the process of determining what pressure a particular patient needs is actually a way of determining the usual airway resistance and what pressure is needed to compensate.  Still, assuming the airway’s resistance is gong to be constant is wrong -- and in my case, very wrong.
  2. Congestion is not irrelevant.  One of the things patients are told is that there is no need to adjust the pressure for seasonal allergies.  Yet, when congestion occurs during the night it increases resistance in the airway which means airflow is reduced. This is wrong.
  3. People have hair.  Everyone laughing?  Well, when you design a mask to attach to headgear so the CPAP stays in place during the night, you ought to think about the fact that people have hair and that it’s slippery.  Somehow this was overlooked in the beginning.  At this point, you can buy headgear designed for people with long hair, and I actually made my own headgear.  The headgear I was first given did not even begin to stay in place.
  4. Leaks are noisy.  Another thing patients are told is not to worry about the leaks because the pressure compensates.  They’ve actually got the physics correct there.  The leak looks like a drop in resistance and the flow increases to compensate.  Getting the physics right here is a bit of a joke -- it’s like thinking your part of a bill is $19.31 when it’s actually $48.31 and you’re proud you got the 0.31 part right.  They really need to fix #1 and then revisit the leak problem.
  5. Leaks dry out your eyes.  This isn’t every night but you can find yourself with your eyes bothering you the next day without really knowing why.

It’s #4 that’s got me ready to swear off sleeping at the moment and is why I didn’t sleep last night.  I was sick earlier in the week and rested more than usual and somehow managed to get more caught up on my sleep than usual.  Now, you’d think this would be a good thing, but it turns out, I wasn’t tired enough to go to sleep with all the noise from the leaks.  I tried using the ramp, not using the ramp, adjusting my gear, and doing everything else I could think of except for throwing the mask across the room -- and that scares my dog so I didn’t.

And what kind of a day am I having so far?  Well, I feel pretty clear headed and so far, I’m more productive than usual.  I know from experience that it won’t last long enough...  I really do need sleep -- apparently, I also need to be very tired when I go to bed or the CPAP will keep me awake.

IMG 0095

Pepper has a new trick!  Can the docs learn new tricks, too?

I don’t know who designed CPAP machines.  I don’t mean that I want to know who designed the box and put the various bits in; I mean, who wrote the specs?  I had a running joke that my original CPAP was designed by a new engineering grad because you figure an experienced engineer wouldn’t make the kinds of mistakes I’ve listed above.  As I began to realize all of the above, though, I started saying it must have been a class project.  Later, I talked to an engineer who actually does design things and found out that it was probably designed by a doctor who gave the engineers specs that told them to implement the above bugs.  Oh.

Don’t get me wrong; I am glad a doctor thought of this concept.  Making simplifying assumptions like those above in order to show proof of concept makes all the sense in the world.  In engineering terms, it was a good prototype (or first hack -- definition #1).  However, after proof of concept, someone needs to do the real design -- and doctors do not have the background to do that.  “The devil is in the details.”

You have to get the details right.  What do we use instead of the simplifying assumptions?  Is the goal really to keep the pressure constant or is it to assure a certain flow to/from the patient’s lungs?  How do you insure that while still allowing the patient to control the breathing?

Now any doctor reading this is probably thinking, you can’t; you’re asking for the moon.  However, the engineers reading this are already thinking of approaches and ideas that might work.  I’m not saying this is simple, but doctors aren’t engineers and need to know that engineers and sometimes physicists need to be involved way earlier than you think.

I am planning a blog post about why keeping the pressure constant and increasing the resistance leads to decrease in flow; it’s one of those things that isn’t hard to understand once its explained, but it’s not something you’re likely to realize on your own unless you’re a physicist, an electrical engineer, or have some similar background (yes, I did spot this myself).  I wish I could tell you when I’ll have it written up, but being tired all the time makes such predictions useless.  I am also planning a separate post to explain what I do to try and mitigate these problems, like using AUTOPAP.

© liz4cps 2011