Omnipod Closed-Loop Study Results at ATTD in Paris

I just came across this press release that came from Insulet (the guys behind the Omnipod system) about the success of a study using a closed-loop algorithm.

From the press release:

The study demonstrated that the Omnipod automated glucose control algorithm performed well, was safe during the day and night for adults with type 1 diabetes, and was very effective at night with minimal hypoglycemia and excellent fasting glucose.

With the Medtronic announcement and then Omnipod’s study I’ve been getting more and more excited by the idea of a closed loop. This has been a technology that I thought was a bit too distant to get excited by previously. Does closed-loop tech excite you? Why or why not?


Closed loop alogrithms make our decision making easier. It still requires lots of work and diabetes knowledge but they absolutely make life easier. I have not tried Omnipod but I´m sure some friends will find this algorithm in omnipod a marevelous news…right ? @Mila

1 Like

What is the definition of a “closed loop system”? Pattern recognition in BG readings thru time?

I’m not excited about it.

For some people it will be great. It will remove a lot of the continuous monitoring and adjustments they need to make.

But for me, I speculate that the problem will always be that whatever algorithm the FDA will allow to be approved will be far too conservative, weak, and slow compared to how aggressively I adjust my insulin manually.

As an example, I believe the new Medtronic algorithm will only let you pick 120 or 150 as high adjustment targets. That’s extremely limiting.

I don’t believe there is anything that makes changes to the algorithm based on the slope of the increase or decrease. Sometimes you need to take more insulin than other times because you are climbing faster.

And suspending insulin for a low is not nearly as quick or effective as eating a cookie. So the suspend feature doesn’t have much appeal to me.

I’d be much more interested in a system that let users customize the algorithm more, and used glucagon to treat lows rather than just suspending basal.


My main worry is that results will be dependent on good sites for both the Sensor & the Infusion set. My control is very dependent on this, sometimes I will great insulin absorption & then for no reason it goes all wonky. How will an algorithm cope.

I’m very excited! We’ve started the process of Looping recently and it’s been fascinating. Caleb and I have been sugar surfing for years, before it was even dubbed as such. I cannot keep up with the surfing of the Loop as we start with open looping. It’s amazing to see a machine do the thinking rather than us. Having an electronic big brother watch over Caleb when he sleeps and to help minimize our emotional responses during the day, I expect will have a notable impact to our quality of life. Thanks for sharing the OmniPod news!

It’s good to see this news of more progress in automated insulin delivery. I’m sure the OmniPod enthusiasts will take pleasure in this announcement.

I’ve been living with the hybrid artificial pancreas Loop since 11-14-16. This is an experimental non-FDA approved do-it-yourself system. I’m learning more each day I use it and would not want to go back to living without it. Loop makes adjustments to my basal rates and moves insulin delivery up or down based on various inputs. The integration of Loop into my life reminds me of the improvements I felt after starting the Dexcom CGM in 2009 and thinking, “I don’t want to live without this!”

Looping for me has been a boost to my quality of life. I am now getting very good results with less effort. I supplement my Loop insulin with Afrezza as well.

Loop looks at my blood sugar level, insulin on board, carbs on board, and other parameters and then makes and executes a decision every five minutes. I’m dedicated to good control but I’m not going to monitor things every five minutes! I look forward to all these commercial automated pancreas systems progressing as they will help the widest group of people. I will continue to participate in the do-it-youself movement because it makes the patient’s preferences plain with its commercial competitors.


I think the do-it-yourself stuff will always be superior to the one-size-fits-all, straight out-of-the-box algorithms.


As long as we survive, we’ll have a front-row seat to this show.


The loop refers to the continuous cycle of feedback information: the blood glucose level changes; the change is detected by the continuous glucose monitor (CGM); the CGM sends information to the insulin pump, which adjusts its insulin output; and the blood glucose level changes again in response to the insulin.



Interesting. I’ve been using OpenAps for almost 8 months now. Quality of life in an abosolutely personalized algorithm has helped in my diabetes management enorrrrmooouusllyyy! I was diagnosed 31 years ago, and I´ve seen lots of technologies change but THIS has amazed me and empowered me as no other tool :wink: Interesting combination with Afrezza @Terry4

1 Like

Interesting point. Algorithms should always have the option to be personalized. Not all of us are comfortable in 120 or 150. Diabetes is so different in every one of us. One size does not fit all in diabetes technologies. :wink:


That’s awesome @Lorraine, congratulations! In my experience, it takes some time to learn how to make use of the Loop, and it also takes some mental effort to just let go and not worry about bg all the time :slight_smile: Once you get it going, it’s pretty amazing how well the Loop can work (btw, in my experience, the same holds for OpenAPS). But, as usual, individual experiences can and will be very different. Wish you and Caleb best luck and success with the Loop!

1 Like

The sensor is usually ok as it can be calibrated. But the infusion site, at least with the medtronic pumps makes a huge impact, as even without the loop.

Are you on Loop or OpenAPS?

Yes, indeed. We’re open looping, still using Caleb’s OmniPod. I’m studying the recommendations of Loop and trying to see how his current settings influence them, how they compare to choices we typically make, and trying to understand what settings would need to be refined based upon the recommendations that might be less than desirable. It’s a lot to process!!

Are any of the commercial systems currently in development going to allow this, or is your point that the FDA won’t regardless of what the commercial systems may request?

Once this system is released, it is only a short matter of utilizing the OpenAPS system with this. The “Autotune” feature is especially exciting as it will automatically tweak your sensitivity with each site, sickness, ect. I, too, believe a custom system is going to be more beneficial that the “one size fits all” approach.

[quote=“Bradford1, post:18, topic:59425”]
it is only a short matter of utilizing the OpenAPS system with this
[/quote] Do you mean you expect Insulet’s Horizon system will be able to connect to open APS?

There is a group creating “AndroidAPS”–using OpenAPS algorithms (and others) with an Android device. They are already using this on the “DANA” pump. Since Omnipod Dash will be using an Android platform, the transition is more feasible. I believe only the pump drivers will need to be configured.