A benefit of Basal-IQ that may not be obvious.
When BG runs high for whatever reason like 250+, we will often find that a manual syringe will work substantially faster then a bolus. Certainly others do the same and it was on here that we learned that technique.
Basal-IQ will still initiate the basal suspend (and cancel of extended pump bolus) 30 minutes prior to the BG being predicted (via G6 cgm) to go under 80 regardless of whether the bolus was pump or syringe.
@Thomas do you have to manually enter the bolus given by injection (syringe or pen) for Basal-IQ to pick up the manual bolus?
No.
Basal-IQ does not use as input any information regarding bolus or IOB. Hence, this simply is not necessary.
Yup. I’ve come to rely on this. Most of what I do can be handled by my pump, but sometimes my blood sugar has gotten out in front, and the pump is slow. I have very mixed emotions about having to abandon the technique.
However, having come from the automated delivery on the 670G, anything that could affect your blood sugar either way (for instance, insulin or exercise), without the ability to communicate to the pump what you’ve done, could have unintended consequences in terms of the kind of delivery you’d receive over the days to come. In essence, if you did a manual injection to bring down a high blood sugar, the pump would only observe that your blood sugar fell… It might decide at that same time the next day that you don’t need as much insulin— because you might drop.
This was a long comment, but someone in my group mentioned the other day that their educator said it would be “okay to use fake carbs” with Control-IQ. I assumed there was a misunderstanding, but if you’re saying it’s okay to use manual injections—- maybe this is not at all going to work like Auto Mode???
edit: as Chris pointed out, we’re talking about Basal-IQ, not Control-IQ when we say it simply suspends when the trend is heading towards a low, without any hidden parameters or adaptive learning that could be mislead by “fake carbs” or syringe injections.
We are talking about Basal-IQ here, aren’t we?
I also think this is about Basal-IQ… Control-IQ isn’t out yet so we can’t be positive on how it will work
Okay, but then how is that different from Basal-IQ? The piece that is affected with things like manual injections is the output. It’s part of what the algorithm uses to determine the Micro Boluses— I’m talking about in Auto Mode. When you do something like correct your blood sugar without going through Auto, Auto learns you didn’t need that insulin. It ultimately gives you less insulin. So I’m just trying to figure out how you could be doing insulin alongside what your Control-IQ is giving you without it affecting what Control-IQ decides you need.
Does that make any sense at all??
It does make sense. We don’t know exactly how Control-IQ will work but I am hoping it’s similar to Basal-IQ in that it relies mostly on predictive algorithms and current CGM trend. If it is supposed to “learn you” like the 670G does then I’ll be very unhappy about that 
AAAAGGGGHHHHHHH
Sorry.
My bad.
Changing my post.
This was Basal-IQ not Control-IQ.
Very sorry.
ie - This is how we operate NOW. Not what we expect.
Totally my fault.
Don’t even get me started. If I even hear those words… 
But this confuses me. Very much. In my mind, Basal-IQ is comparable to Manual Mode with Suspends. Control-IQ was going to be the equivalent to Auto Mode. To an extent, they’re all using predictive algorithms and current CGM trends. The difference has been that Auto Mode has been the only one to determine insulin delivery. But it definitely uses those predictive algorithms and current CGM trends as part of how it does it. Well, as part of the algorithm. For me, the big question is whether or not Control-IQ can do any better. I also think with a stronger sensor it really does stand a chance, but I’m still skeptical.
I started off saying I was confused then I went all opinionated. It’s more fun than confused.
You went and got me all mad. 
Thinking “fake carbs” really are a more sophisticated approach— and that it was just because I couldn’t handle them. 
I’ll be honest, other than when I read your experiences with Auto mode I know very little about how it operates 
I was under the impression that Auto mode’s behavior is influenced by your historical BG values at a certain time, right? Like I get that it also uses predictive stuff and current CGM, but I was under the impression it ignores your non-Auto basal settings and tries to come up with it’s own, that’s the “learning” it does?
I am hoping that Control IQ does not think my diabetes today is influenced by my diabetes from yesterday, but that remains to be seen
AFAIK - There is no “learning” with Control-IQ. It is not going to bolus you at 6AM because yesterday you did this, that or the next thing. It is not going to kick you out of auto mode because of whatever. It is a different approach. There has been a lot of information already released and some of the people involved in the clinical trial have spoken about Control-IQ. So far, there is zero indication that it operates in any way like the Medtronic approach.
If you are going high and out of range, it will deliver combination of basal and bolus increases. When you are predicted to be going low, it will reduce basal and perhaps suspend. Whereas Basal-IQ (the currently released algorithm) has the basal rate either on or off, the Control-IQ (not yet released) can modulate the basal. When you are running within range then your normal bolus and basal profiles will be in effect. Obvious question is the definition of “in range”. From what I heard, the low target is 112 non-adjustable (save for exercise mode that I think bumps the low up to 150 maybe?)
For us, Basal-IQ works so well right now that I am (at this point) not convinced that our current solution is not superior (for us) to the Control-IQ approach.
Running straight and level through the night at 80 is amazing. Right now, that is what Basal-IQ does for us.
I do expect that for a given pump, once the update for Control-IQ is applied that the Basal-IQ algorithm will be removed and further that the update is non-reversible for the consumer. At this point, I am not willing to give up our Basal-IQ algorithm.
This is yet another reason why I liked the ORIGINAL names for these algorithms.
Predictive Low Glucose Suspend (PLGS) ==> Basal-IQ
Hybrid Closed Loop (HCL) ==> Control-IQ
Kind hard to mix those up. Plus the (original) name tells you what they do.
It’s all a big mystery, but from what I’ve gathered, it uses your insulin history, your BG value history, your ICR, AIT, and current sensor value and trend. Having tested in such a crazy way like I did, I think the “trend” was how it calculated the micro bolus size. But in reality, with all of those factors, it would just get stuck in a crappy delivery cycle and would not be able to get itself back out. The most annoying thing I saw— and I saw it over and over again— is it slowing down my “basal”, causing a climbing blood sugar, increasing the micro bolus in response to its own rise, and then HAVING to slow down the insulin in response to the increased insulin…Overnight was the worst. It’s not that it didn’t give a nice flat line overnight— it’s that it would ramp the insulin up early in response to a rise, and then it would have to hit the breaks because of all that insulin it gave me. Most of the time it would mean heading into my wake up time sitting on top of a couple hours of no insulin… and then I’d hit the roof.
I really like Basal-IQ. I think it’s pretty sharp. I hate the idea of having to ditch it in order to try Control-IQ, but I already know my curiosity is going to get the best of me. I can’t wait to see what it does… And if I hate it all, I’ll head for the MDI hills.
I would not switch because it is new and shiny.
My suggestion is to switch from Basal-IQ to Control-IQ because of specific issues you are having that you expect Control-IQ to solve.