Omnipod: auto-calc by PDM or manual bolus calculations?

Do you mean you don’t enter a BG value, you only enter carbs? Or you don’t use the calcs at all?

I don’t have calcs turned on normally. I just did it to answer Beacher’s question.

1 Like

We never enter a BG number for the basis of a correction, and we never enter carbs either. We always calculate the bolus ahead, then enter it. So, if I understood the post correctly, it is always possible to extend the bolus, and we never see this discrepancy.

I’m unclear on what you’re saying here. So, let’s say @Kaelan’s CGM reads 245. You would then use the PDM to do a finger stick blood draw. Let’s say that number that’s returned is 260. You wouldn’t use that number as the basis for the PDM to determine the correction dose to be given? To us, once ISF is dialed in, we use it all the time to administer corrections.

1 Like

Yes, that is correct. K’s correction coefficients vary with his basal which is up and now every few days, sometimes every day with sports, and it’s too much of a pain to change them all the time. But, also, the PDM can be a pain in the neck and not let you do what you want, so we enter every bolus directly.

That’s likely why we have never seen this. I guess here is a good reason to enter all boluses by hand :slight_smile:

1 Like

Also, we hardly ever do a fingerstick to confirm a reading before we inject when he is high. We figure the uncertainty is too high anyway so we won’t learn anything. We trust the Dexcom on these highs. We only do it when he is low, and even then not always, it depends on the circumstances.

We inject probably 20 times a day or more often, many micro boluses, we could never test for every one, and we don’t feel it is necessary either.

Do you have bolus calcs turned off?

1 Like

Also forgot to mention, K’s ISF varies across the day, it is much higher between 1:00am and 10:30am.

Wow, I wish our CGM was that accurate for us.

This is earlier this evening…now granted, this is a new change-out (replaced yesterday morning) – but this kind of thing isn’t unusual for us.

First, notice between 1500 and 1800 hours. His CGM read all the way down to “LOW” for a while. His BG was actually 80.

Then after a sugar tablet, his CGM had him shooting all the way up to “HIGH” and stayed there for 15 minutes. His BG was actually 195. I had to eventually “double calibrate” to reset that number because it was still “double arrows up” at “HIGH”.


We have a basic I:C for MOST foods…foods that we know will require more insulin such as pizza, we adjust for. But most foods…if it’s 20 carbs…we bolus for a straight 20 carbs…we don’t differentiate carbs from one food -vs- carbs from another food (unless, as I said, it’s something high starch content or will digest over a longer period of time and typically spike for us.)

So we check BG’s, then, if they’re in range, we just program X carbs, and have the PDM adminster whatever the correct dose is based on the variables we’ve established for him (I:C, DIA, ISF, etc.,)


thread split?

1 Like

Very interesting.

We frequently have a bad, too low reading, when we eat after turning the corner while going down fast. Many times the Dex will stay 20-30 low for 1/2 hour, then catch up, so we don’t correct it there.

On the other hand, we have never had a fake high. The CGM may be off by 30-40, but I never know if it’s the CGM or the BG meter. And the discrepancy sorts itself out when he comes back down, in general.


On our Tandem (and our previous Animas) the pump would calculate the Bolus but it was just as easy to override that as it was to accept it. No difference. So it would not save anything for us to turn off the calculations (assuming that is possible on the Animas or Tandem as we never looked into that).
We (almost) always use the calculator as our starting point. It gets us in the ballpark. We then adjust the number based on data points which the pump is not aware of.

Agreed. Same. Except for a large bolus based on a high right before bed. Then I will double check. During the day, we have both “physical symptoms” as well as cgm alerts to let us know of a low. At night, there are NO physical symptoms (based on experience, sleep right through, no waking up) so one avenue of “redundancy” is gone and we have only the cgm alert although I will usually supplement that with a 2 or 3 hour fingerstick depending on the circumstance (like how old the sensor is and my trust at the moment of its reading). (I will not stay awake waiting for 3 hours later but rather set an alarm clock for 3 hours later, go to sleep then wake up for a quick fingerstick. Not alway but for a large nighttime bolus - I just feel more comfortable.)

This was curious. Does not your pump allow you to have an ISF which is variable across the day? Both the Animas and the Tandem do and we do take advantage of that. Which certainly helps for the calculator to be closer to what is required.

One item which is absolutely NOT in the calculator due to not being a configurable parameter in the pump (neither Animas nor Tandem) is the variable part of the ISF which corresponds to the current BG. The higher the BG is, the more we need to bump up the Bolus as the ISF ratio changes and requires significantly more insulin per BG to get a drop as the BG goes higher and higher. So this, we certainly manually adjust once we see what the calculator has come up with.

And of course we manually adjust for things such as food differences when we want to extend or excercise we are aware of or sickness or already had 2 lows today or something coming up in the next few hours that will impact it or when going to sleep or or or or…


The Omnipod can certainly do this. I’ve had up to four different correction factors during hairy periods, but mostly I use just one throughout the day.

Not in the Omnipod system either, and I sure wish it was. Surely it’s a fairly simple pump configuration.

1 Like

And it’s even more complicated than that. For me it’s not ISF as a function of my BG right now, it’s the ISF that will apply at my predicted BG in the future, at the time the insulin reaches significant effect. I don’t calculate this, but when I see my BG rising moderately fast and I’m also over 140, I take lots of extra correction insulin.


We rarely see completely fake high – but the CGM can easily be off by 150+ points. Just last week it read HIGH and finger stick said 245 – still high but nowhere near 400. On the flip side, the vast majority of LOW readings are usually higher by a good 20 to 30 points. Not sure if that’s because Samson was *never low or because, once we treated, he was already on the way up and tests at 65 mg/DL.


Have you or anyone else experienced that the longer you go without calibrating your CGM after the blood drop appears, the further off the readings become? I know I’ve read some people who ignore that blood drop for hours or days after it appears, but for Liam, if we don’t calibrate, the longer we wait the bigger the difference is in his actual BG and what’s on his CGM.

1 Like

When I always calibrated twice a day (with G4), the CGM was almost always off from my fingerstick by 1 or 2 or 3 mmoL or even more. Lately, once the numbers are accurate, I ignore the calibration reminders. It will stay accurate for a day and a half or two days, then the numbers start drifting, which is when I calibrate again, wait till it’s accurate, then don’t calibrate until it starts to wander.


I’ve gone a couple of days and had my CGM within about 10-15 points of the meter, which I view as more or less accurate. My suspicion is that if it’s working accurately in a steady way for you, number drift is minimal, but if it’s inaccurate from early on, then maybe that drift compounds more. So probably most of the folks who find Dex reliable/accurate could get away with calibrating once every day or so, but clearly there’s a good amount of people who can’t (but also can’t rely on it that well otherwise).


I would say on average, yes. But we just swapped out “the little sensor that could” because the sticker finally fell off and we were not calibrating more than once a day. It was basically never more than 10 points off – even when we were measuring during steep ,steep drops, etc.


We also use the Tandem calculation as the basis for everything. It is so easy to just click the suggested bolus, adjust it for the “other factors” that have taken place or will take place right on that screen and let the bolus fly. Quick and easy, and hardly any additional button presses.