Control IQ Auto-bolus Issue

First off, sorry for all the recent activity, I’ve been asking all kinds of questions.

I have been going back and forth between using 24/7 sleep mode and not. I totally get the reasoning behind an auto-bolus, just think it may not work for me.

My first mistake, cheating my one meal a day routine and eating two donuts brought to me from our counselor as a thank you for administering the ACT test the days prior. I can easily refrain from eating all the goodies when they have them in the teacher’s lounge, but when fresh donuts were brought directly to me this morning, couldn’t help myself! Obvious with not nearly enough pre-bolus timing, I went high. Control IQ autobolused after it didn’t come down quick enough, which caused me to go low.

That was a bit frustrating, but I decided to keep at it. As I was trending down in the high 70s, went ahead and tried to help Control IQ stop the down arrows by eating 2 dates, my food of choice for correction. Within 20 minutes or so, my levels turned around and were around 120 with an up arrow. I know exactly how my body reacts to this though, it spikes high around 130ish, then comes right back down to around 110-120. However, Control IQ saw that spike, and autobolused 1.2 units. I went ahead and ate two more dates to cover that autobolus.

Just think this situation has taught me that I know when I need a bolus correction more than the algorithm. I need all the help I can get from the algorithm when it comes to controlling my basal, but I think I’m going to leave Sleep mode on and take the correction boluses in my own hands. Many of you have suggested this, I just wanted to come to that conclusion on my own, and today is that day. Maybe for a child, or others with their own situations, the normal mode with auto-bolus works and is helpful. I think for me I’m now a 24/7 Sleep Mode guy.

1 Like

Yup, and the lower range in Sleep Mode helps, too. I have heard people like us who use Sleep Mode 24/7 called Sleepwalkers. :slight_smile: I have also recently heard that an update to CIQ has been approved and will be released soon. It will allow us to determine the range we prefer. Don’t know what those range parameters will be. This info came to me from my CDE and I don’t know any more about it than that.


Your way of eating is not really compatible with auto bolus.

When you spike, it will give you more insulin.
And it depends how you enter the insulin too.

If you tell your pump you just ate 50 carbs, the it will bolus the carbs and count it as a bolus.

If you just take 2 units of insulin instead, it considers it a correction. So it will still give you a bolus for spiking if it thinks you will go out of range in the next 20 min.

So I suggest you either enter the carbs for every time you eat
Or turn off ciq and bolus your own way.

The problem with ciq is that if I am running low, like 75 and I eat 25 grams of carb, there isn’t enough to initiate a bolus, so I can’t enter the carbs. This means the pump won’t know I ate those carbs.

Meaning if you eat carbs that doesn’t bring you over 120, then it won’t bolus and it won’t allow the entry. This can very quickly mess up the algorithm

I know I do have a weird eating schedule, but my challenge for one meal a day is not spiking, it’s going high 5 to 6 hours later. I feel like yesterday was a “normal” test case, because I ate “breakfast” and properly bolused for it.

I ate two donuts in the morning. I looked it up, and they are 22 grams each. So I bolused 44 grams, and waited 5 minutes (should have waited 15, but they were fresh…). So I spiked to 215, had about 3 units of bolus left on board, and it gave an autobolus of 0.57, just before it was going to start coming back down anyway. So that extra 0.57 had me coming down.

I had my players coming in for a workout around 1:30, so I didn’t want to chance Control IQ not stopping the low on it’s own, so just ate those two dates. I wonder if I put in that I was eating 20 carbs, but manually changed the bolus to 0, if it wouldn’t have autobolused? I mean, I got a message saying it’s predicting me to go low, so I assumed it knows when it started going up, that I did exactly what it said. So then, to try to match the autobolus with carbs, I ate 2 more dates, and it did the same thing!

Ignore my extremely high basal rate this morning! Control IQ trying to control the rebound rise from a mild low when I woke up at 5am. Just going to ride the wave until it balances out.

I’m 100 percent it’s not the doughnuts causing you issues. Or that it’s super concentrated sugar surrounded by a fried oil shell.

Well, since you went with sarcasm, I’ll respond with, I appreciate your very helpful nutrional breakdown.

Bottom line is this. I ate something. I gave a bolus, a correct size bolus. It auto-corrected, even though more than 70% of that bolus was still working. That is insulin stacking 101. Not a fan. It shouldn’t be allow to auto-correct after a manual bolus for food, especially while a majority of it is still on board.


Something is off, and I suspect it’s your correction factor and a little bit because you set your basal aggressive.

The exact opposite problem is what we usually see reported, that the automatic boluses don’t give enough insulin. It only gives 60% of the correction that is calculated based upon BG, IOB, and your correction factor. And the problem is compounded by the fact that we only get to set 1 correction factor, even though we become insulin resistant when we’re high. That’s why stubborn hugs hugs highs are so hard to bring down. Control-IQ will not deliver an automatic correction bolus within 60 minutes of a self administered bolus, and will only do so if 60% of the calculated correction is greater than the amount of insulin you currently have on board.

You’ve tampered with the IOB variable, since you actually have more on board (the 0.1 unit extra in your basal) than you need. That doesn’t seem like a lot of extra to me, but you’re probably more sensitive than I am, so that could be part of the problem. It’s also possible that your real basal is even less than you think, so that’s a bigger discrepancy than you think. Since the low is mild, this alone might be enough.

The othe variable is correction factor. That’s the big one that determines just how much insulin is given. If you feel it’s giving too much or giving it when u already have enough insulin on board, then that setting is too low and you need to make small adjustments changing it to a bigger number.

I do want to say, though, you’re doing pretty awesome. It usually takes a while to settle into Control-IQ, get your settings all worked out, and learn to adjust your own habits to play nice with it.

1 Like

Good points. I’ll definitely play around with those. Basal testing, and I:C ratio testing I get. What’s the best way to test for the correction factor? Like you said, when my numbers hang above 200, definitely takes a lot more insulin (relatively) to bring them down than a small correction from 160 or so.

I’ve used my 1:40 correction factor for years and it has seemed to work fine, but I know this system is different.

When I’m high (200+) I often increase the correction that the pump wants to give me. Just tap the “units” field on the bolus calculator and enter what you want. For instance, it may show me I need a correction bolus of 0.87 units; I might increase that to 2.0 units, say. This seems to work pretty well to overcome insulin resistance with high BG’s. Trial and error, folks, and be very careful. But that way my correction factor still works well with my lower, more common, BG’s.


Seriously though high carbs mixed with oil or grease is going to make it difficult to dose.

I don’t know many who know how to accurately dose for 2 doughnuts, but I’m sure there is a way

1 Like

60 g carbs, 50% up front.

And the other 50 percent…extended how long? :doughnut:

That would be my approach and it doesn’t always work. The delay, I don’t think I can adjust. So I’d go with the standard. . 2 hrs.
Occasionally I’ll give extra units if I seem to be trending high.
It would be extremely rare for me to actually eat even one donut though!! It’s too hard to dose.

1 Like

Auto blouses make a lot of people low. (I’m a CDE and look at a lot of reports) sometimes raising the correction factor helps. Control Iq looks at rate of change and what it predicts the number is going to be in 30 minutes from now and takes into consideration how much you have on board. The getting a low after treating a low is a real problem because a lot things we treat with raise us really quickly (the whole point) but it also levels out quickly. I turn off control IQ for approximately 15-20 minutes after treating a low until the arrow turns sideways. Then I turn it back on. I think Lyumjev insulin helps greatly with the after meal spikes. I was on sleep mode for over a year and did really great but I wasn’t eating many carbs. When I started eating slightly more I started having a lot of lows because I couldnt figure out a correct carb ratio to use with the extra basal. Also I realized I was really hungry all the time. Like ravenously hungry. Don’t feel that way when not in sleep.


The thing I have found is that not all carbs are created equal. I can bolus 50 carbs from an ordinary well-balanced meal and I have no problem. But that same 50 carbs made up of funnel BBC cake does not have the same result. Glycemic indexing is real.


Yea for me it’s all about the fat that is eaten with the carbs.
My pump won’t account for that. So I manually override the calculation if I’m having some fat w my carbs


If you look at Looping forums where people are turning on auto-bolus for the first time, it’s really common that people have their ISF set too low. Ending up on a roller-coaster is a classic giveaway. You could try increasing your ISF (maybe 20%) and see if that solves the problem.
But if you really want to sort everything out, there’s no substitute to doing proper basal testing first, and only then doing formal ICR and ISF testing (doesn’t matter what order) making sure you use ‘clean’ carbs (eg glucose) and nothing with fat or protein that will raise your BG for hours afterwards.