I recently read the attached article in Diatribe. I found it very interesting and wondered how many of you are using a fully closed loop system such as the Beta Bionics iLet System or other similar systems.
I am also curious about the open source software systems that are available. I know nothing about them and would like to learn more. I have been using the Mobi pump since September and my biggest gripe is how slow the response from the algorithm is after bolusing, basically eliminating any form of spontaneous eating. The various algorithms used in the fully closed loop systems seem to be able to address this.
@funk The definition of “fully closed loop” used in the article seems to only include those not entering carbs for a meal bolus, so I doubt you’ll get many responses due to the few systems capable of it.
I’ve used Loop (still keep it available, updated, and read the ZulipChat) and don’t believe it qualifies as “fully closed” given the definition above. I’ve also used iAPS and believe it can qualify as “fully closed” depending on how the user has it set up and uses it. I know several people continue to use it, but it fell out of favor with me due to the frequency of updates released, the testing applied before release, and mostly the lack of documentation of how to use it…too many references to AAPS, Loop, and others docs with lack of info on how to apply them. I view it as scientist developing a system for himself; he/she knows what it does, others involved may know what it does, those not involved are left to their own research; fine for those with the time and willingness to research but not for your average T1D. I’ve been using Trio, an off-shoot of iAPS with more testing of the algorithm and features (some would say too much so) and it can also be setup for use as a “fully closed loop” system. I’ve stuck with Trio because it also allows for fat/protein entries and dosing which I’ve found useful. I’m currently one of the beta testers for what is supposed to be the Trio 1.0 release version. The developers are being intentionally methodical in their efforts and this comes off as “slow” at times (“soon” has become a running joke in the discussion groups of the apps development).
While I have UAM settings (Un-Announced Meal) in my app, I have not intentionally tried not announcing carbs/meals. I’m still slowly adjusting settings for my normal meals, then plan to try UAM. My results are normally in the 90% plus level of TIR, sometimes dipping to the high 80’s on a daily basis. I use a G7 (normally pretty good, just enough failures to make me not have faith in it), Omnipod Dash insulin pumps, and Trio tying them together.
The article you cite is interesting and it’s good to know some major manufacturers of CGM’s/pumps are involved. I find the first case it presents somewhat troubling because it appears based on a single individual’s experience with a fully closed loop system (AAPS?). T1D is so variable, this individual could easily be an outlier with very predictable reactions to insulin dosing. I’d want to see a much larger population study, and that’s the rub. FDA demands them and manufacturers are not want to spend the money required to get the data from them without assurance of recovering the expenses…never a sure thing! Hopefully, the DIY community (AAPS, iAPS, Loop, Trio, et al) will force the issue as more and more people become aware and are willing to make the attempt get action. Any progress and easing the needs of T1Ds and T2Ds is welcome!