Closed loop systems

[Catherinegf]

The UK hybrid loop system looks to be released generally at some point. However it needs a lot of manual input. Can anyone suggest a DIY true closed loop system that preferably ties in the Libre 2 and omnipod together? I understand there are no approved systems and nothing in the US approved by the FDA. I want a system where the CGM feeds ALL data to the pod and that determines Fiasp levels. I don't want to have to input carbs. If not a system for Libre 2 and omnipod are there options with the Dexcom and the omnipod? I appreciate these may need secondary apps and I haven't a clue where to start looking.

 

[Deleted member 527103]

The reason you are struggling to find such as system is that Fiasp does not work fast enough.
You need to find an insulin that works as fast as insulin in a healthy human pancreas if you want a fully closed loop system.
Or maybe eat very low carb diet?

 

[Catherinegf]

The reason you are struggling to find such as system is that Fiasp does not work fast enough.
You need to find an insulin that works as fast as insulin in a healthy human pancreas if you want a fully closed loop system.
Or maybe eat very low carb diet?

I don't think there's a faster acting insulin than Fiasp is there? I have read a few articles and there are closed loop systems designed by T1s. Apps developed by T1s to allow the pump and sensor to regulate. Any thoughts?

 

[SimonP78]

I don't use it (I'm on MDI) but have read about OpenAPS's method of handling unannounced carbs, which may be heading in the direction you're looking for, see e.g.: https://github.com/openaps/oref0/issues/297

 

[richardfarnhill]

I don't think there's a faster acting insulin than Fiasp is there? I have read a few articles and there are closed loop systems designed by T1s. Apps developed by T1s to allow the pump and sensor to regulate. Any thoughts?

I think Ljumjev is faster. 5 minutes

 

[Chas C]

There is now a fully closed loop system approved in the US where you only have to advise pump in advance of the meal size (in carbs) low, medium or high, it does the rest itself. https://www.diabetes.org.uk/about_u...e=Twitter&utm_medium=social&utm_campaign=Orlo

 

[Deleted member 527103]

I think Ljumjev is faster. 5 minutes

This is one of the things where we all vary.
I have tried different fast insulins and find Fiasp the quickest for me whereas others find ljumjev faster.

 

[tinyroman]

Chas C. Thanks for that article, really interesting!

 

[richyb]

You might find, speed of fiasco slows down gradually. I was on it for 3 months. It was quick at first but after 3 months was roughly same speed as others. Perhaps it was just me. But I read some others also slowed.

 

[faustus67]

I read about the US closed-loop system the other day. Not sure what the benefit is over my hybrid closed loop system (Medtronic 780g, Guardian 4). I put in the number of carbs in a meal rather than low, medium or high but, other than that, for the most part just let the pump work.

 

[searley]

I don't think there's a faster acting insulin than Fiasp is there? I have read a few articles and there are closed loop systems designed by T1s. Apps developed by T1s to allow the pump and sensor to regulate. Any thoughts?

At the moment even with the fastest of insulin you not get perfect control without inputting carbs.. bg readings are only available every 5 minutes then even if an insulin was able to have peak effect at 15 minutes you’d still be 20 minutes behind so your bg would rise quite a bit before the insulin works

I don’t see full closed loop being approved any time soon even tandem are only able to give a 60% correction to get approval

Another major issue would be the libre2 that does not have iCGM approval as it’s not considered accurate enough. Which is why the libre3 was designed with the approval for pump use