Just the IOB. The other factors are more or less where they should be.
Yes, this is important to keep in mind. The IOB is just an estimate! Based on a curve that is an average of the insulin activity of a huge group of individuals - it is nothing more than that. Our carbs on board, digestive process and individual unfolding of insulin is far more complex to be put into an equation or simulator. It still gives a good orientation and prevents over-corrections to some degree. It needs to be questioned though - especially the IOB of 0.
I agree with you that this micro-management pays out. For the Glucosurfer we just have one target value. Thus we always have positive corrections for being above target and negative corrections (neutral carbs) for being below target. It needs at least a pen with 0.5 units to make this practical - but this is no comparison to the 0.1 increments of the pump.
Next time when you have a positive IOB (>2) you could do two calculations with the bolus wizard:
a) enter 250 mg/dl and 30g of carbs
b) enter 110 mg/dl and 30g of carbs
c) enter 50 mg/dl and 30g of carbs
Just post here what the bolus wizard has calculated for these scenarios (carb estimate, correction estimate and total).
Holger, very interesting (and technical) project! Seems like are many PWD's are dissatisfied with bolus calculators and have to mentally adjust their results. I think the calculators need to be more adjustable. Regardless of the terms used (IOB, DIA) what we all want to track is how long any injected insulin will continue to drop our bg.
Bolus Calculators are of no use to me because the IOB algorithms/Duration of Insulin Action cannot match my slow digestion. I suppose the workaround would be to allow the parameters to be extremely adjustable and this allows everyone including those with gastroparesis to to tweak their settings to best suit how their body works. Whereas most people might set their DIA to 3 or 4 hours, I prefer 2.5 hours because my delayed digestion means my body is still actively digesting food. Incoming digested sugars are neutralizing the effect insulin after the 2.5 hour mark versus most people with fast digestion. (This is in fact why we track IOB at all which is to know how long we can expect injected insulin to continue to drop our bg.) I could set my DIA to 2.5 hours but then it would be incorrect for Correction Boluses. Like most people, correction insulin is active in my body for 3-4 hours. Therefore, the ideal bolus calculator for me requires two different DIA/IOB settings.
In my opinion you can set the duration of the IOB to 3-4 hours. For you the problem is that the insulin is overtaking the digestion of the carbs. So the ideal pump program would allow you to set a waiting time before the injection will start. Depending on your pump model you can use the "delayed bolus" or "multiwave" to compensate the slower digestion. Of course the pump needs to be smart here. It should consider the delay for its calculation of the IOB.
The other way around, compared to most people, my digestion eventually overtakes the insulin, thus my IOB needs to be shorter than most! Today's pump calculators are pretty good but they can be even smarter with just a bit more adjustability. The Accu-Chek Combo's Offset Time parameter is one example of how pumps are improving.
I was confused by the words "slow digestion" and "delayed digestion". To me the digestion is the processing of food to get carbs into the blood stream.
So for you the digestion is quicker then the insulin activity. Have you tried the three analog insulin that are available: humalog, novolog/novorapid and apidra? There is a chance that one of those is actually faster than your current insulin. Most endos have vials available so you can test your individual response. Have you tried to inject 20 minutes before you eat? I am just asking because the smartest formula can not change anything about the physics involved here. Most T1 diabetics have no residual beta cells thus we have no pro-insulin. The pro-insulin dissolves to insulin and amylin. It is the amylin that orders the digestion to slow down so the insulin can catch up. It is this missing amylin that makes our digestion so fast. So the injection of symlin (artificial amylin) with every meal is another option to try OR the suggested waiting time.
Tx, Holger. My situation is an outlier but regardless it seems to me that bolus calculators are not yet at the point where most end-users can trust the final result without mentally adjusting it. I am just an extreme example of how bolus calculators could use even more refining.
For sure, I'd need separate IOB for meal insulin vs correction insulin.
yes it is...that is why i have a problem with the Omnipod. IOB takes off a correction (IOB) when bolusing for a meal if still within the timeframe of insulin duration. Say i take a correction at 11am, 2 units...eat at 12pm (need to bolus 3 units for lunch), Pod will still deduct that 2 unit correction from my meal bolus of 3 units. thus, i don't get enough insulin to even cover lunch because it's in the window of the 'insulin duration'.
Hmmm. My pod doesn't do that! It puts enough to cover all the carb!
The only way that would happen is if you get a new BG reading, and it is * way lower* than expected given the duration AND you have reverse correct turned on.
Since you have already corrected, there is no need to tell it to calculate a correction. Just say "no" to use current BG, and enter the number of carbs, since you already have a correction active. You will get the full dose for the carb.
And in any case, it wouldn't take the full 2 units off.... in your example, if you had duration set to 3 hours, it since 1 hour has passed since the correction, only 1 1/3 unit would show as IOB. Depending on your noon BG reading, it will calculate a new correction, adjust that for the IOB, and add in the meal bolus. You can skip that part.
yeah, i think it has something to do too with 'reverse correction' and target ranges. i don't understand this yet, really. but it seems to take off bolus for food if given a correction within 'insulin duration'. i do like the medtronic IOB feature better for this reason. plus, as noted my lifestyle meter - PDM is about 30 points lower then any other meter I have. So confusing...why does everything have to have so many damn variables..ha!
As Holger noted above, I think, many of us use a setting other than 16 to make our PDM match our standalone meters.
It should not be necessary if meter and strip standards were better
The thinking is that if you have insulin on board and are already below the target that this will end in a low. Thus they assumed that the remaining insulin on board can be used for the carbs you want to eat. Of course we know that this really depends on the digestion processes. Quite likely there are still carbs on board as well working against the insulin on board. Can you turn the reverse correction off in the settings?
Both old and new PDMS can turn this off in the settings.
Maybe you can clarify which setting of the PDM and how it will influence the BG.
Sorry, it was Helmut, not Holger who posted about the settings on the PDM in another thread.
Changing the strip code changes the way the PDM reads the BG. 16 is supposed to be the default for the Freestyle strips. Some of us have found that this seems to read low. Higher codes seem to raise the BG reading, lower codes seem to lower it . I do not know if it is linear. The technique many of us use is t ocalibrate with control solution and find the best setting that correlates well with our standalone meters.
Here is a long thread started by Helmut on this issue:
http://www.tudiabetes.org/group/omnipodusers/forum/topics/a1c-surpr...
There have been several recent comments on this issue in the Omnipod users group, by Helmut and others.
For me , code 17 works best, others use code 19, and still others code 16.
it's ironic, really, we get these very high powered, integral pieces of technology and end up over-riding half of the programs, features. well..crap, we just need a cure! :)
I just did the experiment with IOB 2,24-2,28 U:
a) 250 mg/dl (13,9 mmol/l) ->
Carb 3,0U
BG +4,45U
IOB -2,28U
Total=5,15U
b) 110 mg/dl (6,2 mmol/l) ->
Carb 3,0U
BG +0,6U
IOB -2,24U
Total=3,00U
c) when entering 50 mg/dl the pump won´t do the math or give me insulin before I´ve corrected the low, so I put in 72 mg/dl (4.0 mmol/l) instead, which gave me ->
Carb 3,0U
BG -0,5U
IOB -2,24U
Total=0,25U
a) with 250 mg/dl the remaining IOB of 2.28 is not sufficient to reach the target. Thus the wizard is just reducing the correction estimate by the IOB = 4.45 - 2.28 = 2.17 for correction estimate. The carb estimate will cover the carbs fully = 4.45. This looks good.
b) with 110 mg/dl you still have an IOB of 2.24 units. They fear that this might lead to a low because at this point they can not estimate how much food is still processing. This security measure called "reverse correction" can be disabled in the settings. In my opinion the "reverse correction" is too precautious here. In a situation with higher IOB I would just inject after my meal - but the full carb estimate.
c) this is interesting. For the BG it suggests that you can inject 0.5 units less. This is as if you can eat carbs for free to raise the BG to your target. It can be compared to our concept of neutral carbs. However they also subract the IOB as a security measure. I think this is also a result of the "reverse correction" feature. In this situation I would eat fast acting carb and enter them. For me a carb amount of 10g will be neutral so the units for these will be subtracted. If I eat additional carb on top I would like to fully inject the dosage after my meal. It shows again that the "reverse correction" would not work for me.
Thanks, Holger. Just doing the experiment cleared it up up bit for me. There is sadly no option on my Animas Vibe-pump to disable the "reverse correction", so I guess just have to live with it, be aware of this feature and override the pump when I have to.