I'm more interested in the physics of what happens when you go up and down. In a plane, the most obvious change when you go up and down is in the cabin pressure.
I'd heard that reduction in pressure can cause dissolved gases to come out of solution, and displace insulin in a pump causing extra insulin to be infused. It also seemed obvious that any bubbles in the cartridge would expand and do the same thing. Going back down you'd expect it to end up where it started, so it would need to suck back down the line.
So when flying to Queenstown this weekend just gone with my wife and T1 diabetic son on the occasion of our 10th wedding anniversary, I knew to disconnect the pump when the plane was going up, and also when it was going back down.
What I had no idea about was the magnitude of the effect.
So while I was sitting there holding my son's pump, I thought I'd take some pictures.
This first one shows what you get when you go up over 30,000 feet with about 50U in the cartridge. It looks like just over 1/2 a unit (wild guess based on previous experience with syringes etc). On descent we disconnected again, and I watched entranced as it sucked about a 1 1/2" air bubble down the line, imagining what would be being sucked out of my son if it were still connected.
So of course when we landed it needed priming. I was interested in how much volume we needed, so I used the "fill cannula" feature to give small doses until I could see something come out the end. I reckoned there was just under 0.7U worth of bubble there.
Now 50U in the cartridge is not much at all - my son uses only about 12U / day. For an adult, the cartridge could easily be holding 4 - 6 times as much, and I'd expect the amount of effect to be dependent on the amount of insulin in the cartridge. I resolved to test again on the way back north on our return.
During our stay, we needed to change the site due to an infection, so I put a little more in the cartridge so there would be about 100U left in there when we headed back north. I also decided just for kicks I would load up another cartridge and line with 100U and not put it in the pump but just see if I could see any of these bubbles etc.
So this is the cartridge I loaded up zoomed a few times. I'd never really looked at a cartridge under magnification before, and I was pretty shocked at all those bubbles. No amount of shaking or tapping or flicking the cartridge would budge those little suckers. So this means we need to always consider there will be some bubbles in the cartridge.
When it was time to take off, we disconnected the pump, and I set temp basal to OFF so that the droplet of insulin wouldn't contain any amount that would normally have been pumped (his basal is only 0.225U/hr so minuscule anyway).
Then we took off. At 27,000 feet several things happened. Firstly the captain announced we were at 27,000 feet which is how I knew we were at 27,000 feet. Secondly, I took a photo of the droplet on the end of the pump line (cartridge line droplet was pretty much the same size). Thirdly my wife wanted to give my son some insulin for an in-flight biscuit. Hmmm, but what about my experiment? Let's just say my wife was a little bit down on it by this stage. Anyway, here's the photo.
This droplet is actually bigger than the one above (less magnification). I reckon about 1U.
The fourth thing that happened is that I accidentally flicked the droplet off the cartridge (not the pump line). This made me reluctant to give the pump over to my wife to dose my son since it was the only one with the droplet. In the end I decided that science would have to come second to my son's immediate health requirements, and that part of the experiment was over.
This is the cartridge at about 27,000 feet. The bubbles do look a bit bigger, but it's hard to tell with the different magnification.
So anyway, knowing that the plane is pressurized to 5,000-6,000 feet, I didn't expect to see much change from 27,000 to our cruising altitude of 37,000 feet. Wrong again. Between 27,000 and the top, the cartridge gave up this much more insulin.
Just under 1/2U I think, so all up I reckon we got about 1.4U out of it. With my son's ISF set to 15mmol/L/U (270mg/dl/U), this would be a dangerous dose if we didn't know about this effect.
Then was the descent. As before, it sucked air back down the line. My wife wanted to bolus my son again for the sweets you get on descent to help with recompression, so the pump grew a smaller bubble than the cartridge. We think that bolus was all air though, and so all up we figured that line sucked in about 1.1U of air. But the cartridge was a star! By the time we got home and in a position to put a ruler against the bubble, it was just over 4" from the tip of the cannula to the end of the bubble!!! Could you even conceive a 4" bubble in the line?
So it seemed close to double the effect of the flight down where there was only about 50U in the cartridge. Maybe the cartridge had more bubbles in it from the start.
There are quite a few conclusions to draw from this, firstly you really can't ignore this issue, and I hate to think of what would happen to a small child on a full tubeless pump that they couldn't disconnect. I guess I'd have to advise minimising the amount of insulin in the pump. Or buy a hypobaric chamber and get rid of all the dissolved gas in your insulin before loading your pump.
Also, it's clear that the effect of the change in pressure takes time. If you think that once you're above 5,000 or 6,000 feet you're good to re-attach I'd suggest not. You're still going to get probably half the effect at least. I wonder whether it just takes time for the bubbles to force the insulin out, or whether it even just takes time for the cabin pressure to come down.
It would be great if we could estimate how much to allow for this, for example there might be cases where you wouldn't mind that insulin going in, you just need to know how much. It looks like it's about 1.5% of whatever is in the cartridge.
Going down could be the bigger problem. If you can't disconnect I don't know what to suggest. I guess with a tubeless, the amount of the problem would be limited by the capacity of the actual tube before anything sucked in hits the cartridge. Any tubed pump that uses a rigid reservoir / plunger (such as Animas and Medtronic) needs a prime.
Happy flying!
