Testing the Artificial Pancreas

My friend Kyle Rose had the opportunity to participate in part of the European artificial pancreas project. This is a guest post from him over at Diabetes Mine, sharing his experience.

Here's a pic of the phone device that was used for entering BGs and carb counts, etc. during his time on the pump.

I think I have to agree with him that I’ve been in control of the decision making processes since I was diagnosed 17 years ago, so the thought of letting a machine make all of the decisions makes me a bit apprehensive!

Assuming that time and technology advancements will help minimize the bulkiness of the apparatuses, what are your thoughts on being hooked up to one of these?

The thought of giving over control of my diabetes to a machine also makes me nervous. What if something goes wrong? Of course, my diabetes management depends on a piece of machinery now--my pump. Many people benefit from a pump/CGM combo; the idea of linking them and allowing them to talk to each other makes gobs of sense. But only if the people wearing them know how they work and what to do when they go wrong. I don't think any gadget will allow us to stop thinking about diabetes completely.

No way. Sounds appealing on paper, but I can't imagine how algorithms could replace the multitude of daily decisions we make, provided they could be developed. The cost will be staggering & widen the gap between haves & have-nots. People clamor for pumps & many are declined.

I think the thing to keep in mind is that the AP is just a way of automating current technology, although I agree with others that I would have a REALLY hard time turning over control to a machine. I always tell people that, with T1D, BG management is an art, not a science. There are so many things I have to factor in when administering insulin and there is no way that an algorithm could capture all those things. Sometimes, my dosing decisions are based more on intuition than science, mainly because of the running tab I keep in my head of what is occurring in my life at that moment (stress, hormones, exercise, weather, future exercise, etc).

I have concerns about the resources we are putting towards the AP and how it's being touted as a cure. The AP is NOT a cure. The AP will come with the same issues we currently experience now with both CGMs and pumps (failed infusion sites, insulin gone bad, inaccurate CGM readings, needing to calibrate, etc). Yes, if it works as intended it would make things easier, but it's not a cure. Not even close.

As Gerri points out, my other concern is cost. We know that the CGMs and pumps that are currently available do wonders for many people in terms of improving control. Yet so many folks cannot afford these technologies. Heck, even with an income that is way higher than the national average and excellent health insurance, I have, to some extent, had to make decisions about what technologies I use because of costs. If people are unable to access what is currently on the market simply because they cannot afford it, why aren't we putting the money towards that issue?

I guess I just have a hard time reconciling putting money towards the AP when there are so many T1s out there who can't access what is available now. Why isn't anyone addressing that problem??

I completely agree guys. You summed up a lot of the reservation I was feeling. The "intuition>science" is key. I think there are too many other factors to rely on the limitations of one, or two, or even half a dozen algorithms with regard to insulin predictions.

Without the integration of glucagon the closed loop will not be close enough. Insulin depended diabetics always have to inject more insulin than they need for good control. It is a balancing act of controlled carbohydrate intake and insulin dosing. Often balanced with carbs to prevent the excessive insulin from creating lows. An AP without glucagon can not fight any low. Does that automatically mean it will calculate the dosage more conservatively? I expect so - and further that all the legal issues around these devices will result in very conservative performance. If people get blind it is their diabetes to blame but if they are killed the manufacturer is in the focus. I am sure the manufacturers will decline under this pressure. Thus I expect that the first death person under use of an artificial pancreas will lead the manufacturers to aim for an A1c of 7.5 or even higher.

I suspect that in our lifetime type 1s will have to continue to monitor BG. But, hey if the machine can do it and I only need to keep track of the machine (and BG on occasion) Why not? I agree that the cost issue is real, but that cost is nothing compared to the cost of treating folks who will not or cannot control their BGs.

I've been a type 1 for 40 years and seen the addition of BG meters, injections, pens. long acting insulin, pump therapy, constant monitoring sensors and each improvement has helped me to feel good and live longer. I would say that it really looks like noting radically different than my pump and CGM system I now have...

I would be neither the first nor the last to use one - after it got downsized to chip size. Hooked up will not be the method. Chip & balloon implant and vastly improved human-derived-insulin will be a back-up sensor to one in the chip causing a fail safe operation. I don't trust anything.

The whole idea of giving your entire existence over to a computer algorithm is definitely a bit scary, but if you think about how many times you already do that on a daily, weekly, monthly, and yearly basis, you realize it really isn't anything new.

I'd rather have people working on a device like this with a chance of having a successful implemntation, and then worry about trying to get it to the masses who could benefit, then not have a device like this for lack of trying.

Success is it's own reward I suppose.

The sensors are not really accurate enough for me to trust the computer to judge my insulin needs.
When the sensors go south they can flat line at say 99mg/dl when I am really at 60mg/dl or even 200mg/dl. Only because I have sense enough to double check or just realize that my sugars cant be so flat after eating etc.

I bet that I can control my pump better than the closed loop system can. Maybe in a few years when the get it down better,

I would LOVE to forget about my diabetes for 3 day runs, but somehow I doubt that will really happen. This artificial pancreas still requires quite a lot of fingersticks.

Pressing a few buttons several times a day is not the worse part of diabetes,