My CIQ suspended overnight for an hour but 15 mins after the suspension lifted, my sugar started to rise and eventually sky rocketed , it went from 75 to 180 straight arrow up an hour later.
How to avoid the after highs? Does that mean my ISF was too strong during the suspension period?
I should add, my basal should be correct. I am typically a relatively straight line over night.
If you figure it out, let us all know! The rebounds after suspension are one of the biggest issues with Control-IQ. Nothing in your settings matters. The suspension is entirely based on of where your glucose is predicted to be in 30 minutes, not your settings. There’s some inherent lag with CGM systems anyway. Then add Dexcom’s smoothing algorithm into the mix, which computes data from the past 15 minutes for every number displayed, and it means Dexcom is REALLY slow to show the turnaround. And then on top of that, we’re largely inactive while sleeping, so the interstitial fluid the the Dexcom measures is barely moving, so this whole thing is even slower yet at night.
It means Control-IQ almost unilaterally withholds too much insulin and you’re going to spike, and if you happened to have a natural glycogen response to the low, it’s going to be even worse. If you ate carbs in there, even if it’s a half-asleep unconscious behavior, it’s going to be worse because you’ve double treated the low. And if it was a false compression low and you didn’t notice and your insulin was suspended for a long time, you’re going to get sucker punched.
A lot of people think Control-IQ had learning abilities because these problems get better over time, but the reality is that it’s us who’s doing the learning. You’ll most likely pick up on these nuances and adjust your own behavior. For instance, the best way to avoid the rebound is to avoid the big low in the first place. You might become more aware of when you need to run a reduced profile. You might become more aware of late night snacking and how it makes management more difficult at night. You might become more aware of sensor placement and potential compression lows. Your personal response to nighttime lied might change. Etc…
Almost half that screen is suspended insulin and we don’t know what was happening the 3 hours prior. Do you know why? Figuring that out, addressing it, and learning from it is the best way to avoid the rebound
The other learned behavior is learning to bolus to compensate for Control-IQ’s weak points. For instance, I often eat carbs when I’m low, even if it’s a gentle descent and I know Control-IQ should be able to handle it… Because I feel like 
when I’m low, and the sugar is psychologically comforting. It’s hard for me to sit there and do nothing! But I know now that I have to bolus for it. Right away, or maybe as soon as I see the first uptick. I also bolus wherever I see a long suspension, because I’ve learned it’s going to bite me in the butt later if I don’t.
It’ll be exciting to see how this changes as the updates come out!
YES, ditto!!
Others have mentioned they take injection of long acting insulin, which means you still have insulin in system during suspend. But pump doesn’t know about it. I’ve tried this a few times, but mostly CIQ works good, or like Robin, bolus to make up after long suspend. My liver likes to over-correct, which CIQ has no control over.
Thanks for that guys.
I read somewhere, it may have been a YouTube video that CIQ uses the correction factor and the weight of the person to determine the basal adjustment. So it was suggested if you want it more aggressive, you can enter your weight more than what it really is. I haven’t in this instance, but if the correction factor plays a part, that could mean that my correction factor was too strong so it overcompensated.
My basal was generally stable the rest of the days last week which suggests to me that my basal at the time was set correctly. It could be hormonal changes which caused the suspension but that can never be proven.
That’s a good point about bolusing after a long suspension, I’ll keep that in mind. Thanks. I wish we could override the suspension like we could in BIQ, that would have been nice! Not that it would have mattered in this case as I was asleep when the suspension happened haha.
I mostly use Control IQ and have never found “perfect” settings. But over the last 2 years I have become more comfortable with somewhat higher BG levels than pre-Control IQ.
Ultimately I think that we need faster insulins so that the algorithm can be more agile in responding to post-suspension highs. Unfortunately Tandem doesn’t play nice with Fiasp and I got huge welts from Lyumjev. Plus even these faster insulins aren’t fast enough.
I definitely bolus during and after long insulin suspensions. I constantly play around with lowering insulin use to prevent the suspensions but then I go higher than I want. I periodically add an injected basal insulin with the untethered regimen to “trick” Control IQ. But I am not convinced that it helps that much.
I have some frustrations with Control IQ but I have to remind myself of three things. One, my A1c has not gone up with Control IQ although I think my numbers are higher. Two, Control IQ does a fabulous job of preventing lows and it is wonderful not to eat a whole tube of glucose tabs every day. Three, Type 1 diabetes (okay, Type 2 also!) is really tough and success means fighting the battle every day and not expecting perfection.
My settings these days must be better than a couple of years ago. This is an example of a long insulin suspension…. 
Before I started with CGM and pump I was having worrisome nocturnal hypo.
Those technologies did not solve my problem, learning what they needed to work well and geting thst information did. What the tech does is pay closer attention and give me a bigger buffer zone, but it only saves me time and effort because I made some minor changes and gave iut good numbers.
Every system follows one basic rule. It doesn’t matter how good the tech is, how it works or what it tries to do, GIGO. Garbage in. garbage out.
I’ve been watching what CIQ does intently since starting it and have not seen any evidence of “skyrocketing BG” after a suspension ends. (imo, If you do, then start by setting the CGM alert levels to more within range and the rise/fall alerts more sensitive. Thse annoying alerts should motivate you to change behavior to reduce them and give more time to react appropriately… )
BUT, I never trust technology to do what I can’t do pretty well without it, using similar, reliable information. it’s a good thing that I did because practically from the start I had problems with sites. I used a CGM without a pump for more than a month to get my diet under halfway decent control before I started with a Tandem t:slim. During that time the CGM was untrustworthy about 30% of the time as I mined my body for unspoiled sites.
Since both Tandems algorithms depend on good approximations of BG data, ICR and basal rate profile. I was changing insulin from Novolin to Novolog, so I had to determine those numbers a lot closer than my endo’s generic guesstimate.
I chose to not start with CIQ because it would have made determing my numbers more difficult as it constantly adjusted the basal delivery rate. Instead I started using BIQ for its alerting of nocturnal drops, and turned it off as I did my testing.
For a few weeks I followed the directions in “Think like a Pancreas” with the pump in manual mode to get really good estimates for my basal profile ICR, and CF. I had gotten to zero time below range for almost a month. I was only getting night time alarms well above true 80 mg/dL - from compression.
It helped that I’m older, a make, very careful about my diet, I’m aware of my food intake.and my new ICR is around 1:9. I used Xdrip+ to progressively alarm as I approached and passed out of range. That helped me quickly adjust the ICR to a reasonable approximation. Everyone is different, but not that different. Younger women would have a periodic monthly metabolic change. Adolescents would have growth spurts. The first is a cyclcal change and the other has a trend. These would require time to track but are not totally random.
An unplanned, erratic lifestyle would have too many variables to identify useful correlations and patterns. It’s impossible to manage randomness well… A person with those should be satisfied with an improvement in TIR using CIQ.
imo, Someone who is already using Lantus and a “fast” analog and getting reasonable result should have an easier time doing what I did, because their total basal and average ICR should already be pretty close. Mine were way off from decades of injecting Novolins.
After I had good numbers, with Basal IQ only at night, it rarely ever suspended, During the day it only suspended when I anticipated it, and its only issue was if it dumped the balance of an undelivered extended bolus. CIQ supposedly doesn’t have that problem. Now having ALL my parameters tuned, without CIQ, I expect to stay in range +90%.
Now, the only times I would expect CIQ to suspend delivery would be at night from compression, after meal dosing from a low level but before a delayed meal taking effect, and following unplanned heavy exertion. Because I’m not dependent on the pump, but use it as a convenient assistant, those times should be rare. All the subsequent rises should be expected and moderate.
There’s a saying that I’ve adopted but I try to live my life by . it’s that doing the same things over and over and expecting a different result is willful ignorance.
Not always. I’ve had CIQ suspend at well over 80 without substantial insulin on board. I don’t see any need for quick carbs in that situation.
Words of wisdom, Laddie, every day is new. Yesterday’s perfection may not be repeated today.
I had really tight control with MDI, with a minimum of 6 injections per day. With C-IQ not as tight, but less hypos and I like being able to do small override corrections if needed. With type 2, for me, insulin resistance varies over the day.
Tandem requires you have a good manual settings in place. It makes small corrections based on that. So if you are running high at 3 am you prob need to increase your basal rate at that time.
Same if you are going lower than you want.
I almost never correct lows because I know my pump suspends.
Unless I’m on a hike or some major exertion.
One of the great things about my tandem is that I don’t have to constantly throw glucose at it. I’m sure you get a similar effect from Medtronic and pods.
Eating carbs is a habit that is hard to break. Now I don’t take any rescue carbs unless I’m under 60.
Control IQ in night mode suspends well above hypo and will increase basal up to the basal limit set under pump setting to keep it from rising. Check that setting it’s probably too low.
I was looking at the original post again. and what I see looks like sustained too-low BG at night from too much insulin, but it could also be caused by compression and false CGM readings .
The way to determine which, is to turn off Control IQ and not react to low alarms unless you test by finger stick. If you get up and walk around to do this you can see whether it was a compression alarm.
You might also consider setting your low alarm to be slightly above the cutoff level., or using a cell phone app like xdrip+ to alert you. before cutoff. These alarms might be a nuisance short-term but can help you figure out what’s going on.
If your IOB before bedtime is very low and your basal delivery is set correctly, then your blood glucose should stay relatively constant until 4:00 a.m. If it drops too much then your basal rate at night is too high.
Btw the original post was before the 7.6 Control IQ update. The changes that it included at least for me seems to work better at keeping me in the night time control range without rebound problems. But I have alarms set at levels starting above the bottom of the control range, and more agressive ones as I approach hypo.
