How would you make Control-IQ (a bit) more aggressive

Diabetes Technology
1

I would like the Control-IQ on my T-slim to be a bit more aggressive in heading off and bringing down high BG. Would it be enough to just lower the number on the correction ratio of my personal profiles, ie input that one unit of insulin brings my BG down 50 instead of 60? Or, should other settings be changed? For example, my total daily insulin and weight in the control-IQ menu are currently accurate. Should I alter them a little to say that I weigh more and use more insulin than I really do?

In the recent past, at some times of day my basals were a little high, exerting downward pressure on my BG. Meaning, on the one hand, that I sometimes had to eat just to feed the basal and, on the other hand, I had a little bit of insurance if I was trending high.
Now I’m happier with my basals overall in that they generally keeps me steady and I don’t need to feed them, but I’d like my basal to increase a bit more aggressively if my BG is high or rising.

2

To my read, it seems that your correction factor might be a bit “off”, at least for certain times of the day. So I’d try exactly what you suggested: 1unit changes blood glucose less than it used to (in part because you’ve tweaked your basal rate).

However, I would start with a smaller change, say 5% or so, rather than 20%, and see how it goes. (So to your example, 55-57 rather than 60)

1 Like
3

If your BG is high just after meal, you may need to bolus sooner, or do 2 boluses to align to your food choices - carbs vs fats and proteins. Some are taught to count carbs only, but some do better with counting some proteins in addition.
Depends mostly on meal choices, but may also be impacted by exercise, stress, or other, before or after meal activities.

2 Likes
4

That’s what I do, I think of it as a manual extended square wave bolus. The extended bolus function on the T:Slim does not work well for me. I’ve tried various percentages, but just doing a 2nd bolus works much better if I see BG climbing again.

An example is if I eat a high fat/protein/carb meal such as the occasional 2 tacos from Torchy’s Tacos I will prebolus 4u and maybe 1 to 2 3 hours later.

If it means anything, my last HbA1c was 5.6% a drop of 0.2%. The first year with T:Slim I was rock steady at 6%. In this the 2nd year of 6 months my HbA1c had dropped to where I was on MDI. This is without the nocturnal hypos.

1 Like
5

You can set the collection factor – how much your bG drops per unit of inulin, and the carb ratio – the amount of carbs (gm) a unit of insulin will cover. I set those as close to the right values as I can. Since I can’t change the duration of the insulin, I add extra to my weight and total daily insulin. I also have a profile with a basal of 5X normal. This is handy when my bG’s get “pegged.”

6

Continuing on my earlier thoughts on extended boluses, I would like a smart extended square wave bolus that would use the 2nd wave to counter a late rise of BG after a meal.

7

My problem is that the algorithm starts backing off my basal too soon in order to prevent me going low (I guess). E.g., this morning my BG was slowly heading down as lunch time approached, which is exactly what I WANT it to do. It got to 115 about an hour before I planned to eat, so I thought I was in ok shape. But when I was ready to bolus for lunch I was up to 130. I looked at T:connect and saw that immediately after I hit 115 it started backing off my basal and I started rising again. NO! Exactly not what I want! What parameter should I adjust? I like this pump a lot but lately it really seems to be fighting against me rather than assisting.

Arrrgh.

IMG_E1B4DAFBFA49-1
IMG_E1B4DAFBFA49-11125Ă—2436 107 KB

1 Like
8

Do you feel like it’s behaving well the rest of the time and just not before meals? If so, set up a time segment a few hours before you generally eat and raise both the basal and correction factor (set it to a higher number) a titch. Raising the basal sets you up for insulin the pump is way less likely to take away unless you’re trending low, and raising the correction factor takes power away from Control-IQ so it can’t withhold as much to prevent what it thinks is an impending low. Raising the correction factor also alters the program’s conception of his IOB is going to affect you. If it’s reducing insulin at 115, it’s because it thinks any IOB you have is going to tank you, but with a higher correction factor it thinks that IOB has less power to affect your future BG.

Control-IQ really isn’t as complicated as people think it is. It’s all about wave dynamics, how big/small the waves are and where the baseline falls . The basal rate sets the baseline and the CF sets the amplitude (how big the highs and lows are). If you look at your graphs and the middle of the peaks isn’t where you want it, turn the basal up or down for that segment. If you see big swings up and down, or your don’t like the way your BG is rebounding after Control-IQ intervenes, then you need to take power away from the system by raising CF to a higher value. Oh the other hand, if you BG seems stuck and you want to give them system more power to intervene, then you need to lower the CF for that segment to a smaller number.

3 Likes
9

Ah, I see. This was particularly helpful:

A few weeks back, about 6 weeks after starting with the system I wanted to tweak it a little closer to what I was looking for, and I must have mis-read or misremembered something and did the opposite of what you suggest. This explanation of the underlying logic makes clear why I should have tweaked in the other direction. I’ve reverted to my original, higher CF settings and will see how it does over the next 24 hrs or so then reevaluate.

Thanks for the lucid explanation, as always.

3 Likes
10

So is the correction factor all I should be looking at if I want control-IQ to be more aggressive? Does my total daily dose and weight play any role?

11

Control-IQ is pretty much always delivering your programmed basal rate or a modified version of it informed by your BG predictions and correction factor. TDD and weight don’t play into that equation at all. I can’t personally tell you what factors need changed, though. Changing your basal can also make the pump feel more aggressive. But generally speaking, yes, CF determines aggression. Since the settings work together, you very often have to change both basal and CF. As you nudge one, the other might start acting up. That’s pretty much what all the “tweaking” talk you see on these boards is referring to. That constant nudging of one or the other until you find the sweet spot.

TDD and weight are just used to impose safety limitations on the algorithm. They have nothing to do with DELIVERY, and everything to do with restriction. (Unlike other systems that go and try to invent a basal rate from those parameters) I don’t know how the programming works exactly beyond that. I’ve never seen weight and TDD discussed in further detail than that. Those two factors combined tell the pump how sensitive to insulin you are for your weight. The safety limit is set at a much higher bar for a 200 lb person using 100 units a day than it is for an 80 lb person who’s only using 20 units a day.

Say you’re stuck at 300 mg/dl for a while. It’s time for the next automatic correction bolus, and you don’t have enough IOB already to cover the high. You would expect the pump to give more insulin then… But it might look at your recent delivery history (I don’t know how long it looks at) and decide that based on your weight and TDD, it has already given you more insulin than the programming is comfortable with and NOT deliver anymore at this time. It’s not a situation you’re likely to encounter, though. It’s just an extreme backup protection. Just one of many safety parameters built into the system.
Even the DIY options have this baked into the programming.

1 Like
12

I think you have some misconceptions.

Control IQ is a relatively dumb tool designed to help maintain a level of control that is initially established manually. Its “feedback” loops are intentionally designed to be moderate in effect to prevent insulin overdosing. It works using parameters that are determined manually by testing your body’s response to carbs and insulin.

Those parameters are NOT constants. They change rapidly as your BG moves outside the “target” range, or increase the carb loading. Their effect changes with exercize and illness. (If you have lots of peaks and time above 225mg/dL Control IQ isn’t going to make as big a difference as lifestyle changes will. )

What you need is for Control IQ to work as well as it can. That is partly influenced by the pump and CGM. The rest - maybe the biggest part- is up to you. The more you do up front, the easier it will be over time.

Control IQ can help a person dependent on insulin do a better, easier job, by doing routine calculations many times a day, but like any dumb system, it needs good directions to get good results.

Control IQ can work well hands-off - with the right parameters, when you are well, not doing anything strenuous, not eating large or imprecisely measured amounts of carbs or fats. Change any one of those and it won’t work as well.

While Control IQ will increase basal delivery rate to try to keep you under 160mg/dL and reduce it to try to prevent you from going low, the fact is that it’s only a "trim"adjustment. It reacts only to rate of increase. If the other factors or your food measures aren’t very close, and your BG rises or falls too much too quickly, the basal rate reaction will be too little too late for close control.
The same is true of the auto-boluses. They are intended to handle small errors NOT compensate for large ones If your manually-directed meal bolus is too little, too late, the auto bolus will be worse.

Your question suggests that either you haven’t yet established good manual control under controlled conditions - or your body has changed due to a change in lifestyle or health. (If you’ve lost weight or started an exercize regimen, your old pump settings won’t be accurate.)

If you want to get better results from Control IQ, the place to start is with periods of consistent behavior where you can measure your body’s response during fasting, after meals and after a correction bolus - with Control IQ turned off. You need to know these.

If your basal rate is too high or low it will make it very hard to determine the other effects. Basal is easiest to measure at night. During a prolonged fast (more than 8 hours not including 4am or strenuous activity), a correct constant basal rate will keep your BG constant for at least 4 hours. Study the effect for 7-10 days, where you didn’t go low or high enough to need to intervene, average the rise or fall and divide the correction by the time interval to get the basal adjustment amount.

With that basal rate determined you can observe the onset time of your dawn rise. That’s the end time of your night basal profile. Knowing that, you can turn night management over to Control IQ with less danger of night hypo and start measuring your day profile settings.

Knowing your dawn rise start you can calculate the increased basal rate during the early morning needed to cancel it before breakfast.

Unless you eat the same composition and size meal for breakfast, lunch and dinner you may need different ICRs and CF for each meal. The way to “tune” ICR is to measure carbs very carefully and with Control IQ off calculate error from target 5 hours later. Do this for a week with each meal - trying to keep your peaks below 200mg/dL by adjusting the carbs in each meal.

Each day record the meal carbs and the correction sugested by the pump at the end of test. Add that correction to the number that the pump made for the meal bolus, and divide the number of meal carbs by it to get an estimated meal ICR.

At the end of the week, average the daily ICRs. Unless all your dinners are very close in composition, the day ICRs are likely to be very different from the average, and average will be the best the pump can do. The same is true of breakfast and lunches. The meals that produced very large corrections are ones that you will need to correct for manually.

As you calculate the new ICR for each meal, you can create a profile time interval for it, and let Control IQ take over calculating that meal, with you spot-checking its result an hour later.

You can calculate your CF during a fasting period of stable BG. Disable Control IQ, take 7 dextrose tablets (28grams) wait 10-15 minutes for your BG to rise. Measure it when it peaks with a fingerstick, then bolus 1 unit. Four hours later measure your BG again. The error will tell you how far off the 1:30 correction factor was.

If you provide accurate numbers for ICR, CF and meal carbs to Control IQ it will do as good a job for you as it can.

13

I wish I could adjust the insulin duration and the bG levels at which it takes various actions. I understand their not want to be liable if a user gets those wrong, but I think it’s nonsense to force everybody to use the same values. They just aren’t the same for everyone.

2 Likes
14

I’d like to set my own target. 110 bg is too high for me. I’d also like to be able to keep the temp basal option. With BIQ, I don’t have the lows and I can e as aggressive as I want to be.

2 Likes
15

There a few things you can do.
You can tell your pump that you weigh more than you do you can change your carb ratio and your correction factor to be more aggressive.
If you want to bolus and you don’t want your pump to shut off basal, you can go into the fill cannula menu and deliver up to a few units thst will not count toward your onboard.
Lastly you can live life in sleep mode where you will target 112 instead but it won’t give bolus corrections, which you don’t need if you are real good at meal boluses.

I’m not suggesting any of these things, only pointing out that they are available.

1 Like