Impact of I:C ratio

My daughter uses a Tandem X2 Tslim with Control IQ. She’s pretty insulin sensitive. We have her on Sleep Mode during the night.

Several days ago, our endo changed the I:C ratio from 1:19 down to 1:18. At the same time, she lowered the nighttime basal rate, saying that Control IQ will regulate the BGs.

After making the changes to the profile, our daughter has had some extreme highs on the heels of some extreme lows (during the night). In other words, the graph looks like a roller coaster.

Is the change in I:C possibly the reason?

I can’t say my daughter’s graph had been a flat line before but we rarely saw the extremes we are seeing now.
I’ve put in a call to the endo to get help but meanwhile I simply want to have a better understanding as to why this might be going on.

The I:C ratio would only change the size of the bolus she got for a given amount of carbs. If she ate and bolused not too long before the extreme lows (less than four hours before the low, say) then it’s possible, although the change seems modest. If the low came in the middle of the night long after her last meal bolus then it seems unlikely the change to the I:C ratio had any effect. And if the nighttime basal rate was lowered that wouldn’t account for the lows either.
Do you have alarms set that go off before the lows get extreme? If the lows come out of nowhere they could be compression lows (in other words false lows caused by the sensor being squashed.) Are you double checking the Dexcom with a fingerstick test?

The doctor lowered the nighttime basal rate. If the basal is not enough, your daughter will start to go high at night. Then, when the pump responds to the highs, your daughter goes low. You are in the rollercoaster.

I think an insufficient basal amount can potentially cause a lot of problems. If her BG numbers have gotten worse since the change, you might want to consider changing it back.

Maybe, Eric, but Daniela describes the highs as being “on the heels” of the lows suggesting the lows come first.
It’s possible there is no causal relationship between the doctor’s modest changes and these bad nights, just coincidental correlation. The roller coaster nights might be caused by something else — bad infusion site, compression lows leading to suspension of insulin and unnecessary correction, hard to know — but I wouldn’t immediately blame the doctor’s changes. Though, of course, you could go back to the old settings and see if that fixes things.

1 Like

Sometimes when I start a new Dexcom sensor it will display extreme lows for the first several hours, (like 40 or 45 low). When this happens the pump suspends all insulin. If you’re not really low, that can cause you to go very high. I have my pump alarm set to off when I get down to 80. I’ve actually turned off Control IQ when this happens just so I can continue getting basal insulin. Then of course, I have to remember to turn it back on in a few hours (set a phone alarm). I don’t seem to have this problem when I presoak the sensor for at least six hours.
So much fun!

2 Likes

Thank you all for the helpful input.

It’s interesting to me how much faith doctors (and Tandem reps too) have in Control IQ taking care of highs and lows. The doctor felt strongly that the slight change made in the nighttime basal rate (Was around .375 and was lowered to .33) wouldn’t make that much of a difference. He said that previously my daughter’s basal had been automatically suspended quite a bit thanks to CIQ.
believe his thinking is: with a lower basal rate the suspending wouldn’t happen as often, and perhaps therefore the numbers would stay more readily in range….
I just wish I had a clearer understanding of the impact of basal and I:C changes.
I’m also wondering if the BG she has 4 to 5 hours after eating has a long term effect on what happens at night. Might it be better to be slightly high (e.g. ca. 150) 5 hours after dinner? Or stable at 112? At that point our daughter is asleep.
??

1 Like

This is possible if Tandem is making changes “behind the curtain”.

For example, if she starts to go low, Tandem turns off her basal. And then it can result in going high later.

On the other hand, if she starts to go high, Tandem increases her basal. Perhaps too much. And then it can result in going low.

Everything is connected in some way. It is tricky to get everything dialed in correctly.

It really takes patience and just making very small changes, and only changing one thing at a time!

You are doing the right thing. Asking questions and willing to test things out.

If you post a T:connect screenshot so we can see how the algorithm is behaving and what the various settings are doing, we can explain better. The pump screen doesn’t have enough info. That doesn’t show us anything about Control-IQ behavior in the previous hours.

There are 2 pieces of info you need to understand before we can talk settings and what they do: One, insulin lasts a LONG time in the body. Any individual dose of insulin has potential to lower BG for more than 5 hours to come. It is most powerful/noticable in the first 3ish hours, though, and then continues to work more quietly in the background as it fades away. But this slow activity means you need to look at what happened in the 2 or 3 hours BEFORE any issue arose. Secondly, Control-IQ works on the basis of “net IOB”. Basal insulin is supposed to match the sugar our livers crank out to fuel our constant energy needs. It’s constant. That never really stops. (At least not for T1s without beta cells.) So that basal insulin is null and void. Control-IQ ignores that and only really accounts for insulin that varies from the programmed basal rate. Any insulin delivered that is in EXCESS of the programmed basal rate will be displayed as IOB. IOB is all about how much potential the insulin in her body has to lower BG over the coming hours. IOB may also be negative, in the case of insulin suspension and basal reductions, but we don’t see that. IOB will just read “0” when it’s negative… But new boluses will seemingly disappear faster, because it had to cover the basal deficit. Every adjustment Control-IQ makes is to balance out that net IOB with where her BG is predicted to be in 30 minutes.

To answer your questions about what the settings do:

Basal settings are the very foundation of Control-IQ. That’s the framework the automation is built upon. Setting up good basal rates is step one. This is why testing basal rates is so strongly encouraged. It’s the very first piece of the puzzle.

In the loosest and most general explanation I can offer, with a basal rate that is too high, a person will trend low, or at least below target, and Control-IQ will frequently reduce/suspend insulin to get them back up to target. (It is not the ONLY reason you might have too much insulin on board, though!) It also screws with the IOB displayed, because there is actually more insulin available to lower BG than the pump knows about. This means there will always be a little bit of extra insulin added to adjustments/corrections. Conversely, if a basal rate is programmed too low, that person will struggle with being above target and Control-IQ will constantly be fighting to bring them down. And there is actually less IOB available to lower blood glucose than the pump knows about, because some of that is actually needed to cover the inadequate basal. It’s not really that simple, though. The other settings can mask/complicate things, which is why we start with fasting basal tests and adjust the other settings after. You don’t have to starve her, either. You can break it up into a bunch of mini-tests that are much easier to handle. I actually do believe a good Endo can recognize the patterns, though, and make appropriate adjustments without the fasting, but it’s a slower process. It’s the difference between months of minor tweaking, trial and error, and getting basal rates locked down in a few days with the fasting tests.

Once basal rates are established, i:carb ratio is one of the simpler things to check. This setting is not used by Control-IQ at all. It is only used by the bolus calculator to figure out meal time doses. If she shoots up after a meal and doesn’t come back down, the i:carb setting is too high. If she plummets like a rock, the number is too low. If her blood sugar looks a bit like a rollercoaster where she winds up pretty much near where she was before the meal 3ish hours later, then you found the sweet spot. At that point you may even be able to figure out insulin timing (how long to pre-bolus and does this meal require an extended bolus) to better match food absorption. That last bit is a skill that comes with experience.

And then you didn’t even mention Correction Factor. This is the setting that Control-IQ is using CONSTANTLY to make insulin adjustments and corrections. It’s a biggie. Like, real big. It’s the aggressiveness dial. It sounds to me like the doctor made good adjustments based on the aforementioned patterns, but now the CF setting is causing problems. Because he set up the pump to make more insulin available to the pump with other settings, there might be TOO MUCH insulin available for corrections now. The equation isn’t balanced.

You know how blood sugars are constant rollercoaster of ups and downs? This is the setting that determines if you’re playing on a boring kiddie coaster, a family-friendly ride, or a dangerous extreme thrill ride with giant peaks and plunges. Whenever people talk about the bad rollercoaster, you’ve got to look at that CF setting. A CF that is TOO LOW (smaller number here = more aggressive) means Control-IQ will dose too much insulin when she’s high, driving her low. And do the same exact thing when she’s low, withholding too much insulin and driving her high again. And then the rollercoaster repeats…

On the other side of the equation, a CF that’s too high will essentially cripple Control-IQ. That takes away any power to influence BG. She’ll be high and Control-IQ’s hands are tied. It just can’t help. The CF setting won’t allow it to give anymore insulin. Same with lows. If she’s predicted to go low, it will THINK it’s withholding enough insulin to prevent it, but it really can’t. She’ll just keep trending down until she’s low enough to trigger the full suspend. (Full suspensions are 100% based on Dexcom predictions, and pump settings don’t mean squat at that point)

So the real questions are: Have you done basal testing? Did the meal and inappropriate bolus make her high/low, or did Control-IQ do that?

Also, you’ve gotta ask, are you exacerbating anything by being impatient? Not a judgement here, I’m totally guilty of this myself! It’s hard to sit there and do nothing when this all moves so slow. If you feed a low, you’re double-correcting on top of the insulin Control-IQ already withheld, and you should expect a wicked rebound high. That’s a behavior issue, not a pump setting issue. Control-IQ will actually tell you to eat carbs if it thinks it can’t handle the low. “Control-IQ has predicted that will drop below 70 mg/dl in the next 15 minutes. Eat carbs and test your BG.” And if you do need to treat with carbs, she really needs way less than she would have before Control-IQ. Common recommendations are 5g instead of 15. But of course, that’s subjective.

I’m guilty of comfort eating carbs if I’m low, because it’s the only thing in my power to fix that horrendous feeling. It’s a self-soothing thing, not a medically-necessary thing, if that makes sense. I wind up having to bolus for those carbs just like any other snack!

2 Likes

This was my thought too, but I don’t use Control-IQ so I’m not overly familiar with how it works. But also not mentioned is, how (or how much) are the daughter’s lows being treated? If she previously had a higher basal at night – which was resulting in lows, which is why the doctor lowered it – that “excess” insulin might have partly covered for the high that followed treatment. Now the basal is lower, those recoveries go higher, and CIQ kicks in, hence the rollercoaster. But, again, perhaps I’m missing some important aspect of Control-IQ.

So helpful!

I’m embarrassed to admit that I’m not sure of how to attach a T connect screenshot (although I now have one in my photos) here onto my Forum reply.

Robyn, you mentioned patience—among so much other useful information. The last 2 nights have been better. Not exactly sure why though.

Our daughter’s CF at night is 1:100.

As for treating lows, 5 carbs of glucose can sometimes be too much when she’s low. Of course it depends on how low. But even when she drops to 50 and I give her 5 carbs, her BG will often go really high afterward.

My experience with C-IQ is about 30 months. My pump is set to sleep mode for bedtime. In the following screenshot there are 4 times basal was suspended between 3A and 5A. BG 78mg/dl was the lowest. Not enough for an alarm.

Generally unless BG is < 65mg/dl I just acknowledge and go back to sleep. C-IQ Will suspend basal and BG will rise. At this point basal is resumed.

I am cautious on treating these bedtime lows, sometimes 1/2 a glucose tablet for 2g. This keeps me off the adult roller coaster for an easy ride on the kiddie coaster.

This works for me. Your mileage may vary.

As to the change of I:C that shouldn’t affect while sleeping. Basal is a different story, but C-IQ can vary basal from zero to a bit higher than the basal rate. You can see that in my screenshot. When the pump is delivering full set basal it shows as a straight line without the 5 minute divisions. See where the pump delivers more than basal rate.

For me C-IQ works very well, my problem is more with bolus for meals. This may be because I’m type 2 with varying insulin resistance.

I hope this has been helpful.

Ha Ha forgot the screenshot.