What an interesting comment! It actually ignited some reflection in me.
My son has a lean body right now, and we often have a hard time getting a sensor to last. When I think of accuracy, we have several cases that come up:
About half the time or more, our calibrations are within a very small number of units of actual. It often is associated with a specific sensor, where strings of calibrations always seem good—but not always.
About a third of the time, a calibration is about 15 off, and the calibration result will halve that. It does happen quite often that these come in strings too, where, everytime with calibrate a sensor it is 15 off. These strings appear to progressively drift away, but others don’t always drift continually with time (although the all do, just not with a steady rate)
A small number of times, we are way off. In that case, I will often do two calibrations within a half hour or so. It is quite rare to be that far off many times on a row.
I have experimented a few times, and forced a calibration on a sensor (twice the same number) when it is somewhat off, by averaging two or three fingerprint measurements and force-entering the result. It has several times resulted in a tack-sharp follow-up calibration. But my number if samples is too small to be reliable.
What is your starting point, if you don’t use the PDM’s bolus calculator? Are you calculating a dose according to BG, carbs, IC ratio, correction factor, etc., in your head or on paper? Of course the PDM can’t take into account things like past or forthcoming activity, BG trend, type of food to be eaten, activity of insulin today, general health today, and so on and so forth, but given the right numbers, it can suggest a dose, which you can then easily raise or lower as needed. Which seems to me easier and simpler than starting from scratch each time. Or maybe I’m missing something here.
I don’t use any calculator or formulas or anything. All of those are very elementary, and they do not consider all the factors you can use in your head.
And who wants to measure every little bit of food. If you go to different restaurants, they don’t all give you the same size servings. Do you want to bring a measuring cup and measure rice at a restaurant?
Doing it in your head is much easier, and it takes into account a lot of things bolus calculators don’t. So that’s how I do it.
We normally do it in our head, based on the carb count (although my son is typically able to estimate a plate more or less). We start by looking at his normal ICR, modified by the time of the day. We compensate for the BG at the time with the CF of the time of day, then we tweak based on whether we are trending up or down on the day’s basal, how much sports effort we had that day, and the activities that may be coming after the meal. I am probably forgetting some factors. Sometimes the tweaks are almost nothing, but sometimes they amount to 20-35%. It is rare that it is a lot more than that, except right after practice.
Occasionally, my son uses his phone calculator to figure it all out. A lot of it is numbers-based, but there is almost always a gut tweak at the end that my son does.
The problem with the PDM is that it won’t always let us do what we want, for instance, if we enter BG. So, after trying a few times early on, we stopped counting on the PDM.
The other aspect is that, to be honest, I want my son to be able to keep all of this in his head. I figure that my job is to give him the tools to be successful on his own. I don’t want him to depend on the PDM. For instance, if he can’t do it in his head, he may not figure out when there is a PDM dosage problem. In general, I would like his brain to be his tool more than his PDM, as a matter of training.
We inspired ourselves from @Eric in the way we look at it. Although we are still fairly numbers-based compared to him. We use the spirit of his approach!
@ClaudnDaye I haven’t noticed this issue. Like @Beacher, once I notice readings are stable I tend to calibrate less often. And as @cardamom noted, when it’s good - it’s good, when it’s not - it’s not.
So over the weekend I experimented. I calculated my dose in my head and then checked against what the PDM calculated plus my tweaks. I wasn’t far off, in one case by as little as 0.35 units. But I wouldn’t call it easier for myself, and it certainly took much longer, and was generally a nuisance when I knew the PDM could have done most of the “work” for me in a flash. Whether you start from the PDM’s suggestion or do it in your head, you still arrive at Dose X and then adjust for The Variables In Life. I found using the PDM is not necessarily more accurate or otherwise “better,” just a lot quicker.
So while it’s nice to confirm I can do it in my head, we surround ourselves with tools to make life a little easier, and the bolus calculator is, to me, just another of those tools. I’m totally on board with needing to know how to calculate a dose without the PDM (or any other pump), and I think it’s really smart that @Michel wants his son to know how to do this from the outset; it’s a necessary skill. But to dispense with the tool altogether? I can add, subtract, multiply and divide, but that doesn’t mean I throw away my calculator.
With a complex meal for which we are taking into account protein and fat (obviously carbs also) and pulling the data for multiple foods, this complete carb calculation is certainly what we would spend the most time on.
For many (most?) people, meals are a fairly common collection of foods. I have had steak and baked potatoes more than once. I may eat steak with other things, or baked potatoes with other things, but they are frequently in the same meal.
When people eat buffalo wings, they frequently have blue cheese or ranch with it. So the fat in the dressing and the semi-sweet wing sauce are often in the same meal there. And if you’ve eaten it once, there is a reasonable chance you’ve eaten it more times (unless you hated it).
People generally have favorite restaurants that they go to a few times a year, or maybe they even go there very frequently. And restaurants do not have an infinite number of choices.
There comes a point where you can recognize meal combinations and patterns. I guess my point is that you don’t need to keep adding everything up all the time.
Additionally, someone’s IC ratio is possibly different from one day to the next based on a thousand things - those ratios are not set-in-stone. And different carbs have different glycemic indices. BG’s are not always the same, which affects your correction numbers. BG’s are not always flat at meal time, so they can have different slopes up or down. So carbs calcs are just a best-guess.
There is a good chance a person’s “best-guess” will actually be better than the carb calculator’s “best-guess”.
For sure. Most meals are as you describe. Recognized or a couple seconds of mental for the carbs. It is meals we do not frequently have which involve multiple food items with high levels of protein and fat for which I will pull out a scrap of paper and figure the total effective carbs out by hand.
My approach is to get the carbs (including effective carbs from protein/fat if significant) as close as possible. Guestimate the insulin from carbs and BG. Throw the carbs and BG into the calculator and see what it comes up with. If my guesstimate is off by more than 2 units than I am going to take some time and figure what is off - the numbers are not that hard to mentally figure. In any event, once we have the number from the calculator then we adjust that based on all the other stuff that the pump has no information about.
This is an interesting thread for me to read as someone who has never in my life used a bolus calculator of any sort. I don’t know if it would be better or not. One issue I see is that the times when it’s hardest to figure out are usually when I don’t have exact numbers on the food anyway. I don’t think it’s the math that’s the limiting factor for my ability to figure out boluses so much as lack of relevant info, which would be the same either way. Also I usually factor in a lot of hard to quantify factors, like, a vague sense of “how insulin sensitive do I seem today?”, how active have I been/am I going to be, and how disproportionately annoying to cover, relative to its presumed nutritional info, was this particular food last time I ate it, etc. I think if I could quantify it all enough to enter into a calculator meaningfully, and the calculator had the equations to factor that stuff in, it would in theory be superior to my own estimates (mechanical judgments (e.g, PDM) pretty much always outperform clinical judgments (someone’s head), when all the relevant data can be utilized in the mechanical judgment). But otherwise, it doesn’t seem that useful/have much appeal to me. But then, this is one of the many times that @Eric and I are just Team Old School, and I imagine it’s really different depending on how you learned to do diabetes care.
Don’t all pumps have bolus calculators? So far this discussion is limited to the Omnipod (except for the meerkat’s Tandem), but I’m curious how many people using other pumps don’t bother with the pump’s calculations. (I can’t change the thread heading.)
What’s your starting point for an injection? Do you have a scale that says if you’re between, say, 70 and 100, you take x units, and if you’re between 101 and 140 you take x+1 units, and so forth, and then further adjust for carb amount?
That sounds like a (somewhat modifiable) sliding scale. I carb count. The difference is a sliding scale starts with a baseline mealtime dose at that 70-100 number that you could adjust. So maybe dinner is 4 units at 70-100, 5 units at 101-140, etc, and if you had “extra” carbs, you could add more. Carb counting on the other hand starts with 0 units at 70-100 (or whatever your ideal range is) and then adds insulin for both corrections and carbs/food, with no presumed set “mealtime” dose.
So I have a correction ratio (1 unit per 30-50ish, depends on where I am to start and other factors affecting insulin sensitivity, sometimes I alter it further) and a carb ratio (1:10). I usually only factor in fat/protein if it’s a low carb and heavy fat/protein meal, and to be honest, I’ve had good success eyeballing that, since I’m much less capable of estimating fat and protein counts on the fly (part of why I like eating low carb). So if I’m 70-100 and eat 50g of carbs, I would take 5 units. If I’m 140, I’d probably take 6 units. If I were trying to correct down from a much higher level though, I’d probably use a stronger correction factor.
Sounds like essentially what I did on MDI, and it incorporated carb counting. The starting doses presumed a certain carb intake. So if I tested 100 at lunch, my suggested bolus would be 7U assuming I would eat 50 g carb. If I was only eating 10 g, then based on a 1:10 ratio, I’d subtract 4U.
All of which is what a bolus calculator does for you. You then further adjust for all the things a bolus calculator can’t account for (activity, sensitivity when higher, and so on).
I have an omnipod but I just use the bolus calculator on Android APS. It takes into account carbs on board and total IOB which is way more math than I can manage in my head. I find it helpful because I can be really aggressive with corrections and becasue it logs everything. I like to look at numbers sometimes With that said, I will ignore the bolus calculator when I think I can do better than it or when there is exercise involved.
For my son who is on MDI, I just do it in my head.
No, it was because I’m not swift at math and I’m rusty doing this in my head and I had to figure in the IC ratio and then the correction factor and then estimated IOB to arrive at a starting dose, and I wondered why I would opt to spend that time doing it the long way when it only takes me a few seconds to enter my carbs into the PDM to arrive at basically the same (pre-tweak) number.
That would certainly make things tons easier. My breakfast is pretty consistent, but in the past two weeks, not a single dinner has repeated or been that similar to another. So at dinner especially, I don’t really have frequent combinations and patterns to recognize, and so I do have to add it all up. And even with my two slices of toast at breakfast, I eat different breads, and depending on their density and how I slice them, in the past two weeks those two slices have had a carb value anywhere from 21 to 40 g – a significant range that leaves me uneasy guessing.
I often use the t:slim bolus wizard for the carb calculation, but normally adjust that for BG and other factors myself. Entering BG into the t:slim wizard is way too tedious, so I rarely do that. Instead, sometimes I lie about the amount of carbs to compensate for any BG adjustment I wish to make, and sometimes I just change the suggested bolus amount after they calculate it. And sometimes I skip the wizard entirely and just directly enter my guess for an appropriate number of units. When I start LOOPing I’ll have to use the wizard and be truthful, but rumor says that the user interface for LOOP is designed by people who actually use it, so I’m hopeful it will be easy to do the right thing.
With that said, I will ignore the bolus calculator when I think I can do better than it or when there is exercise involved.