G6 bolus adjustments based on trend arrows

Day of of G6. I want to know how to adjust bolus using trend arrows (I have looked at Think Like a Pancreas)

So breakfast at 9:30. I bolused about 45 minutes before eating. Now two hours later I am at 187 with a sideways up arrow. My Omnipod PDM says I should NOT take more insulin and I have plenty IOB.

Using my correction factor and Appendix D of Scheiner’s book, I took a small additional bolus of .60. My only other choice is to wait until my BG goes high enough for my PDM to decide I need more insulin.

Other options???

I meant to say day ONE of the G6

Assuming you entered your insulin to carb ratio, insulin working time, and insulin correction factors correctly, I would lean towards not adding additional insulin if your PDM says not to. At two hours after bolusing you likely still have IOB and I would be careful not to stack boluses. Otherwise you could face a hypo in another couple of hours.

Just my opinion.

You on omnipod? You can’t override that? I can. I dont use the bolus calculator. I don’t think it works…for me.


I DO take the trend arrow into account when I make bolus decisions. While nobody should follow my recommendation because I am not a trained professional and because, to be candid, nobody will point to my CGM profiles and say: “Wow, this guy may be old, but he sure has his T1D under control …”

Note: I use a Dexcom G6 in concert with a Tandem t:slim X2 pump.

Let me try to use a very detailed example. For this I will assume that my correction factor is 50 mg/dL per unit and that my target is 100 mg/dL. Let us further assume that I have IOB of 1unit.

Let’s consider 3 different situations:

  1. Current glucose reading of 200 mg/dL with a flat trend arrow.

  2. Current glucose reading of 200 mg/dL with an “upward diagonal” trend arrow. In other words my glucose has upward momentum of 1-2 mg/dL/min. Note: I don’t know if “momentum” is an ADA-approved term, but that is what I call it.

  3. Current glucose reading of 200 mg/dL but with a diagonal downward trend arrow.

I don’t know about Omnipod, but Dexcom has a bolus wizard to help with the math.

For case 1 the bolus wizard will say that I should add a bolus of 1.0 units. Why? My current BG is flat at 200 mg/dL, but the 1.0 unit of IOB should drop me to 150 mg/dL. So, the bolus wizard calculates that I need 1 more unit of insulin to drop me from 150 mg/dL to 100 mg/dL. Agreed?

For case 2, I’m at 200 mg/dL but rising. If we assume that it takes 30-ish minutes for a bolus added now to “begin to work” my glucose reading then will actually be something closer to 230-260 mg/dL. The bolus wizard will once again say that I need a correction bolus of 1.0 unit based on the current reading of 200 mg/dL … but I would likely increase that to 2.0 based on the fact that my BG is likely to be close to 250 mg/dL by the time it “begins to work”.

Finally, for the downward trending reading of 200 mg/dL, the bolus wizard will once again calculate that I need to add a bolus of 1.0 unit to the 1.0 IOB. But, after the 30-minute “activation time” of insulin, I know that my BG will have already dropped to something like 140-170 mg/dL … and that the IOB that I already have is likely to drop me close to 100 mg/dL. In that case, I would likely ignore the additional 1.0 unit that the bolus wizard recommended and give myself NOTHING additional.

Of course, for me, if the arrow is vertical up or vertical down … or worse yet double vertical up or double vertical down … I would likely make more extreme corrections to what the bolus wizard suggested.

Lest you think that this is just some “stuff” that I made up, here is a reference to a Dexcom document (in their health care provider section … not their normal user section) that more-or-less supports the approach that I am suggesting:

As I say, I’m not an endocrinologist or a CDE, so I suggest you check this out with your diabetes team before you try this. I keep hoping that the bolus wizards in our pump/CGM integrated systems will begin to do something like this for us.

Finally, if we assume that newer insulin’s will get faster and faster, worrying about the momentum of rising or falling BG becomes less of an issue.

At least that is the way that I look at and use trend arrows.

Stay safe.


1 Like

At 187, rather than look purely at the trend arrow, I like to look to see if I’m at the top of the “hump” of the after meal curve.

If it looks like I’m at the top of the hump then I hope that I’ll be coming back down soon, I will usually not take an additional bolus at 187.

If it looks like the slope is getting steeper upwards - like I completely underestimated the number of carbs in the meal - then I will take a correction bolus.

For me 0.6 units is pretty dang tiny. I will usually have 5-8 units with a meal and if I’m above 180 and not yet coming down after a few hours I will take several more units.

If I hit 140-150 I take a correction bolus. I don’t care what my Omnipod says. I either set my Omnipod for 2 1/2 hours or 3 hours for duration time depending if I’m aggravated lol. 2 1/2 is really too little, but 3 hours I am overriding it more.

But I aim to hit below 140 by timing. So if I am hitting that number, my theory is I guessed or dosed wrong. Most of my insulin works between 1-2 hours, it’s max effective time is around the 1 1/2 hour mark. So while it might trickle work some after 2 hours, it’s not at it’s full strength on board. Maybe more like 10-20% on strength left on board. Which is not enough in my book to bring me down from a 140 plus.

That’s me though, everyone handles it differently. It becomes a guess to know how much to take at that point.


For me I worry less about that arrow and look at the shape of my graph. If I already dosed and I can see that my meal is peaking and will come back down then I don’t correct it.
However if it’s moving up steep and wide I will correct it. You can generally figure out where you are going by looking at the inverse of the climb you already have.

If you see here I’m already recovering. But even 10 min ago you could see the curve was going to come down the same as it went up. About a half hour and I’m back to pre meal.
I hope that makes sense.

1 Like


I agree. I would argue that using current BG is a zeroth-order correction, adding trend arrow is first-order, and your use of the shape of the curve is a more sophisticated second-order correction.

Stay safe!


Thanks to all for your advice. I guess my real question is, I see people posting that they stay between 70 and 180 (or equally fabulous numbers) “92% of the time.”

HOW? As soon as I eat, I go up. I try and bolus at least 30 minutes before a meal. I search and check carbs (but am sometimes wrong) but how does someone keep those numbers?

I think I will post this new question. I need tips.

Hi @PamS Right now I am at 97% TIR between 65-160. Usually I pride myself on being 99%, but DP has been playing with me.

I find it’s a matter of timing, how good you are at timing can determine what levels you go to. I am probably more aggressive in treatment than most but I am retired now and I feel that has made a huge difference. Plus a pump and CGM after I retired helped tremendously. I also use my exercise bike promptly to control spikes when I have them.

It also helps if you are familiar with the foods you eat, a lot of new foods or foods you have to guess at hinder knowing how you are going to react. So I have certain meals/foods down pat that I don’t go above 120. A common one for me is a bowl of veggies with tofu etc and I bolus half a half hour before and the other half when I eat. This also allows me to adjust the second dose to what I actually ate. I stay under 120 for this.

So trending up to 187, 2 hours later can be because of what you ate and potentially time of day you ate? And timing?

A delay to rise 2 hours later can be from some fat in the meal or very high protein and not bolusing for that? Or I guess not enough insulin to start with. High fat or in some cases even some fat can delay absorption of carbs, so that by the time you digest your food, your insulin is starting to go away? So in a meal with some fat, say a piece of dark chocolate, I can bolus when I eat it and hardly get a rise in BG levels. High fat and I am adding a bolus amount after I eat, some a half hour later, some an hour later. If it’s very high protein, sometimes you have to take some additional insulin for the protein too and that doesn’t hit as quick as carbs.

And then there’s the time of day of bolusing. In the morning I mostly don’t eat as I have a very hard time bolusing for it. If I drop too low and even eat 4 carbs for an adjustment I can go up 50 points. Sometimes, so I try my best to not have to go near food in the morning. I have a lot higher ic ratio before 11 am.

So it might take some experimenting, but I would try to evaluate the food you are eating first. Maybe add in some bolus later for some of the foods that have some fat or higher fat or very high protein. Hope this helps some. All of us can be so different, it does take some experimenting to figure it out.

Depending on what you ate, the bg spike might not match the insulin activity. Sometimes placement of injection/pod will influence how fast, and activity before/after delivery can impact it.

If your food is mostly carbs, try adding something with fat or protein. I rarely eat straight carbs, always a mix to dampen the rise.

I manage to stay in range 90+% of the time and some of it is just luck! However, I tend to eat high fat, high fiber, medium carb meals and the result is that I tend not to get a big spike early on. I may begin to sail high at 3 or 4 hours but by then I don’t have much insulin on board so it is easier to correct using the bolus wizard.


I target 95 percent but I often hit 97 and 100 rarely.
It’s all about timing. You need to calculate when the carbs hit you and dose so the insulin hits the same time. Normally for me it’s 30 min. But when I walk to lunch for work it hits me faster like 15 min. So I just tine it like that. I don’t always hit it right so sometimes I’ll dip before or it will go high a tiny bit.
In my graph you can see I dosed too early for lunch and a bit too late for dinner but neither went out for too long.
You just gotta find the sweet spot.

1 Like

Thanks to everyone. I am bolusing earlier pre meal and am doing better.