A little over a year into using Tandem t/slim I’ve grown appreciative of how the mobile app shows its basal adjustments at a fine scale, dialing back to 0 on the low and gently increasing as it climbs the upper end of good.
I am still learning to fight my anger at unexpected highs (13.1! I calculated breakfast perfectly) and not doing my usual reach to over correct when it suggests a measly 2 units.
@cogdog (OMITIf I read the screenshot correctly your basal rate is 1u per hour. With so much variability you might benefit from bumping up basal a bit - maybe to 1.3 or 1.4 and see how that works for you.)
Actually on 2nd and 3rd look forget what I wrote above. I have no clue what your basal rate is. It is the extreme basal variability that shows that your pump needs some adjustment.
I’m not sure where to begin. Basal, meal bolus and corrections all sort of affect one another. How did you come to use the numbers in your pump?
I had fine tuned my dosage from when I was on MDI. That’s went in to the pump initially. Over time those numbers have been tweaked to achieve a better result. It’s all me and not my endo, but he is perfectly happy with the results.
Basal testing is best done after a 4+hour fast. One indication of being on target is if your overnight basal is fairly flat, without a lot of highs or corrections.
Actually I am going to stop here and hope that some of the wise folks here will help to to more stable BG
Thanks @CarlosLuis the vertical axis scale is for blood glucose. You can see some of the basal in the plateaus, I have it set to 2.1 to 2.75 units at different times of day. I have monthly reviews with a nurse educator, and we have been fine tuning basals as well as adjusting correction factors and insulin carb ratios.
It’s all ongoing, but I have gone from about 50% time in range previously on Omnipod to more typically 70-80%. Some of variability is irregular schedule, other is my love of food. But I have felt better since switching to this pump. The other tell is that my retinopathy from laser treatment in the mid 1990s seems better as I had typically had bouts of small floaters in my eyes, and none this year (will confirm with opthamologist in a few weeks).
Thanks, I dont do much tweaking myself with the settings.
That’s my experience. I had a sign of early retinopathy before I started using C-IQ. Both eyes a healthy except for cataracts in both eyes that have been stable fo 15-20 years.
I am not able to see enough history to reach any conclusions. I likely would not try, even if I could. But one thing I do notice and I suggest, is talking to your Educator about your coverage.
How long is insulin active, (I forget if you can shorten it on Tandem) and sensitivity. In the above example you are showing 13.1 with 3u in progress of dosing. Just taking a stab at things, you educator might desire increasing the effective coverage.
Obviously we are all very different and your situation is not mine or vice versa. But I would push more insulin to lower my blood sugar than 3U if I were that high.
Now, this may be perfectly right for you. But it might answer part of the disconnect. It may also be that I am way off base. It is just want I noticed, no more.
rick
I appreciate all ideas. I am not quite clear what setting is “effective coverage” — it sounds like the correction factor. I do get fuzzy over the 3 variables in play.
and my thought is usually- WTF? 3 units will not reduce that 13.1 but I feel like it is accounting for the 2.5 basal per hour too, and it’s aiming for a more gradual reduction over time than a full drop.
I loathe my tendency to over correct both highs and lows. But we look the trends and patterns in detail.
You can customize insulin action time if not using Control IQ, but when Control IQ is active then the action time is set to 5 hours. Which is pretty long for some people! My endo recommends 4 hours for me, but I like CIQ so I deal with it. I do think seeing the IOB for that extended time can help me mentally to avoid overcorrecting too.
Do you use the Tandem Source reports, @cogdog ? They are much more helpful for spotting patterns! https://source.tandemdiabetes.com/ I have never had much luck with the wireless upload from the phone app, so I do recommend just plugging your pump into the computer to upload before looking at reports.
For example, here’s my 2-week trend overview report. (You can see I constantly have my pump in Sleep mode because it says I have been asleep all week haha.)
The spots I highlighted in yellow are my problem areas, where I have tend to spend 10-25% above range. Which totally makes sense to me – these are my postprandial ranges for lunch and dinner. So my first priorities personally is improving my insulin:carb ratios and my pre-bolus timing.
For the insulin:carb ratio adjustments, I try my best to count my carbs accurately and not override my pump’s calculation (except I often do override meal bolus reductions if I’m “below range” since I’m aiming for a lower target BG than Control-IQ is). Then I wait… and if I go high, I also do my best to not correct and just wait. 
A correction will mess up the experiment! If I’m back in range in 2-4 hours, then I leave it as is. (If I went out of range but ended in range, I might try increasing my pre-bolus time to try to stay in range the whole time). If I’m still high after 2-4 hours, then I’ll try increasing my I:C ratio (like from 1:11 to 1:10). If I’m low after 2-4 hours, then I’ll try decreasing my I:C ratio (like from 1:10 to 1:11).
These last few days I’ve had good luck tamping the dinner postprandial range by pre-bolusing by 10 minutes, having some Metamucil before I start eating, and exercising for 10-20 minutes afterwards. If I’m eating something that will digest more slowly already, like a vinaigrette salad and whole grain and bean bowl, I might pre-bolus by only 5 minutes. The lunch pre-bolus timing and postprandial highs are more challenging for me with my desk job… I can stand and march in place while I work, but my typing sometimes suffers!
Thanks Rachel for this insight into your method. Yes I do use Tandem Source Reports, my educator has access and they are the basis to our discussions.
I admit not regularly examining them in this detail. And I don’t pay much attention at all to pre-bolus time. Mostly I am taking boluses right at meal time!
I relate! For the 20 years before I got a CGM, I thought “give your insulin 15 minutes before you eat” was just one of those “things” endos say haha. I’ve learned I really need to understand the “why” for advice like that to stick. Seeing charts of exogenous insulin action time vs. digestion finally helped it click for me.
Pre-bolus timing is a great place to start experimenting! Just keep low treatments nearby – in my hunt for a better bolus, I have waited too long before eating too. 
I’m intrigued to do some experiments, thanks Rachel.,How do you make those fine tune adjustments for I:C ratio? My educator has guided me to a set up with 4 times of day where he suggests settings. Do you use that or just one set that you fine tune?
My long time hang up (no real good excuses) has been being good / diligent at carb counting. I fully understand but I end up getting lazy, and guessing poorly. But I have gotten better since starting the Tandem pump in using the RxFood app to estimate my carbs. It’s gotten me much more inline and even reduces muteness reach for seconds.
but this means I get the carbs and bolus when the meal is ready on the plate— so now I have to delay eating?
Your mention of “15 minute” rule flashed me back to my early T1D days in the early 1970s. In the crude era of urine testing (hear the fizz of Clinitest tablets anyone?) at diabetic camp we had process where we had to empty bladders 30 minutes before testing so it was a bit more “accurate”. In the morning, our counselors would wake use up early with loud yells of “VOID” in the bunkhouse.
The pump is a long way from those times.
Same here. Those small basal changes can make a noticeable difference. It often leads to smoother control without the bigger swings.
I’m so glad the tech has come so far since then! 
The ratios you set with your educator are a great place to start! Then just look for little patterns. Nudge things up a little bit up or down a little bit and then observe for a couple days. Then repeat, looking for just a little bit “better” each time.
I have four different I:C ratios set up by time of day, roughly correlated with the ups and downs of average basal needs below (from Think Like a Pancreas). Personally my I:C factor is highest in the morning (1:7), moderate in the afternoon (1:10), and lowest in the evening (1:13). Then my basal and carb ratio ramp up significantly starting at about 1:00 a.m., but I’m usually asleep and not eating then haha. It’s also definitely not perfect – hence my continual tiny experiments!
This is the insulin action time graph that finally helped the 15 minutes “click” for me! Also from Think Like a Pancreas.
Carb counting is a challenge. 
I rely a lot on apps too. I use LoseIt! and it works well especially for meals I eat on repeat. The app will remember the portion sizes I recorded last time, so for my most predictable meals I can just search and add really quickly. Then when I go to “View Meal Summary” it will give me the meal overview all together.
For this meal I didn’t pre-bolus until I started cutting the strawberries. Not a perfect 10-15 minutes, but it buys me 5-7 while I assemble everything.
Fascinating stuff from Think Like a Pancreas, I might have to get a copy (I did find they have a podcast). I might not understand the first graph for insulin units/hour, thats suggesting needs vary with time of day and age (I am now dark triangle age)
I have my regular meeting with my educator Friday and plan to ask him about bolus timing. I trust his experience.
It looks like my peak areas (marked red) are mid afternoon and also bumps overnight (sometimes that’s 10pm snacking that maybe I am not bolusing enough).
We’ve done a bit of fiddling over the last year, I think to land on good I:C ratios 1:3.6 and correction factor 1:0.6. I 'd bet we adjust bolus amounts again. I have to admit I get the effect of the first two confused.
Another interesting observation is that when I turn off the pump for my shower, it tends to rise slghtly 1-2 mmlol/L, it generally self corrects. I just thought it interesting that it responds that way to a short stop in basal.
Again, thanks for the ideas here.
This discussion caused me to pull my copy of “Using Insulin” by John Walsh teal.
Reading the chapter on TDD your roadmap to better control. I made champion my T:Slim pump profile of I:C and CF. Currently leaving basal as is.
TDD divided by 500= carb factor. Correction is variable TDD is divided by 1600 for the lest aggressive up to 2200 for the most aggressive.
Using a pump we have a history of TDD. Average several days will put you in the ballpark. For MDI it takes some record keeping.
TDD can be adjusted 5 no more than 10% if BG is consistently running high or low.
Pumping Insulin may be good but I don’t have it.
https://www.diabetesnet.com/using-insulin/?srsltid=AfmBOoqhCpBwoUweBsVObJtMkq3iF3eb6zYz_1hyh9-tt7fp4-qagVf2
I wasn’t looking for it, but my unread copy of “Pumping Insulin” 2024, it was just staring me in the face in a bunch of unrelated books. Right next to Karol Copek’s RUR (Rossum’s Universal Robots) 1923 and the Complete Book of Dogs.
Anyway I am reading chapter 8 “Total Daily Dose & BC (basal/bolus) Settings Often Need Improvement.” This continues through chapter 14.
Quoting the first paragraph - “Good habits in diabetes work hand in hand with accurate and well-adjusted devices to produce desired glucose levels. Bolus before meals, count carbs or size meals, wear a CGM, and respond to high or low glucose readings these actions lower your average glucose and minimize glucose fluctuations.”
From the section on Carb Factor- Giving boluses 15 to 30 minutes before eating compensates for insulin’s slow action compared to digestion.”
I am not an expert on reading scatter graphs, I just don’t seem to really visualize what they are conveying, This one comparing pump and meter users in 2007 to AID pump and CGM does make some sense to me. The non AID users are all over the place while the AID uses not so much.
I wish there was a 3rd graph of those on MDI with meters.
I will say that when I was 1st diagnosed as Type 2 DM 36 years ago, I received excellent education and support. It was emphasized that we the diabetic are the captain of our team.
To be the captain one needs to know how things work to get them to work together.
I hope this is not boring everyone. I know I can get obsessive and that I am a wanna be teacher. I will continue reading and learning and sharing to the willing. I know that this may be old hat to some, but probably not all.
Not bored at all, and also I am not leery of changing settings. I’m armed with some questions to ask Friday.
Yep, needs typically change throughout the day! The general up and down in this graph is more helpful than the “units/hour” on the side, as insulin resistance can vary greatly. Part of the up and down throughout the day is hormonal. For example, the dawn phenomenon or foot-on-the-floor / FOTF phenomenon, where a lot of people (like me) experience a significant rise in insulin needs early in the morning as hormones kick in to wake you up. Another part of the up and down could be related to activity levels.
It looks like your educator has set you up with a relatively steady basal rate throughout the day, rising as the day goes on. But midnight to 8:00 a.m. is your period with the lowest basal rate, whereas for me that is the period with my highest basal rate. With your highs overnight, it might be worth showing your educator that “dark triangles” graph alongside your screenshot of Tandem Source CGM trends. Just to see what they think about the possible correlation of the “general basal needs by age group” graph from Think Like a Pancreas showing a rising need over those hours, and your highs correlated to those hours.
YDMV, but for comparison, I’m a “dark square” on the graph, and these are my current pump settings. My basal doubles between 12:00 a.m. and 6:00 a.m., which just means my body needs a lot more insulin to keep me steady. In general, a change in basal rate will affect your blood sugar 1-2 hours after it changes. So my 6:00 a.m. change is actually to prevent a 7-8:00 a.m. rise.
You can see my carb ratio going up and down roughly correlates with the basal. More basal = more aggressive carb ratio, less basal = less aggressive carb ratio. (How to adjust correction factor still confounds me a bit… but I need my correction factor less since my variability has gone down a lot this year!)
I also notice this! I think the heat from the shower might make the CGM sensor read higher than it really is.
The division number seems super high, so I’m not sure I’m doing the math right… My total daily does is about 37 u / day. But divided by 500, it suggests my carb factor would be 1:0.074, which would be tons of insulin! My carb factor is more like 1:7 to 1:13 depending on the time of day.
This makes me want to read Pumping with Insulin for the first time. Basically everything I know came from Think Like a Pancreas, Pregnancy with Type 1 Diabetes, or this FUD community!
I think I wrote that backwards it is- the number 500 and divide it by the person’s current total daily dosage of insulin.
That gives 13.5 for you
I have to leave so maybe I need to reread both calculations ICF and CF to edit if wrong. As with any such calculations what works for the diabetic is correct.
Aha, that makes muuuuuch more sense! The math is mathing! Thanks for the clarification. 
