Basal Rate Adjustments on Pump

Question I've restarted back on POD this AM and meet with CDE tomorrow. I've already had to cut basal down to .25 as I'm going low 7 hours after last bolus or food. Question, if one adjusts their basal rate down (I decreased .05), how long does it take for this adjustment to take effect. How does one know if this is OK. If I keep going low with just basal do I do another adjustment down to .20. I just become so d@mn insulin sensitive on the pump compared to MDI - Levemir...but, I never got my levemir doses correct so, at this point, it's a guessing game. THANKS much!

I noticed this before, Sarah; I think you are thinking of lowering basal rates as a form of correction for a low. It just doesn't work that way, at least not effectively. If you are too low, then treat your low with whatever you ordinarily use in the right dose to bring you back up to target. I think you set yourself up for a roller coaster if you are continually responding to a single low (or high) with a basal change. My suggestion is to correct the low then keep careful records. I use blue marker for lows and yellow for highs. Then take a look at three days records. If you see that you are repeatedly low at a certain time of day than reduce the basal for that time period only by the lowest possible amount. Then observe again for another three days before making more changes.

I know you are nervous about lows, but with regular testing you can always catch and treat them. Don't use the basal for that purpose.

If one starts a pump, he/she often doesn't know what the correct basal rate is and we need to adjust. If i'm not eating or bolusing and going low, it's too much basal, it's the same thing that was happening late at night, far after dinner and last bolus...because it's too much basal. I'm fully aware that we treat lows with glucose, highs with corrections. However, I'm just once again starting out and if I'm dropping from too much basal (obviously) I need to adjust it. I have to 'start' someplace FIRST and then we'll figure out how to adjust for different basal rates to work with my individual needs. The basal is the foundation and has to be set right. It doesn't matter if I treat a low, if it's from basal, I'll just continue to drop regardless, if it's too much. Same thing if it's not enough basal, one can correct but once correction burns out, BG's will climb right back up.

Hey Sarah:

I've been experiencing similar frustrations with my Omnipod in getting the right basal settings. I've found it can vary based on what site I use and I've read posts about how well the pod is applied and/or the cannula is deployed can affect absorption. Just when I thought I got it right for one pod, I'd have to re-think it for the next pod. Ultimately I learned that in my case I was getting the best consistency from the backs of my arms.

Further, I found that the type of insulin I was using made a dramatic difference too. Apidra would work well for me for the first 24-36hrs, but after that it seemed to lose it's efficacy - I was doubling up on basal and bolus just to get the same results. When I switched to Novolog this issue went away completely.

You mentioned frustrations in getting your Levemir dosing right, but I have yet to find anything else that can give me such a consistent, flat basal rate. When on MDI, I split my 24hr dose in half and inject once at night (9PM) and once in the morning (7AM).

Because Levemir works so well for managing my basal, I've discovered a hybrid approach that gives me the benefits of both MDI and a pump. I inject 2/3 of my basal requirements with Levemir using the approach above, and do the balance of the basal as well as the bolus with the pod. That way I still have some latitude for basal adjustments on the pod for things like exercise.

Now if you want to take a pure pump approach, I would first isolate those sites that give you the most consistent results. Then I would experiment using your approach of changing the basal rate in very small increments (0.05/hr) and give it 24hrs between each change to see if you're trending one or the other direction after your bolus' have worn off. You should be able to settle on a rate that keeps you relatively stable.

Hope this helps.


I usually wait like 3 days before I decide if I keep a rate or ratio change. I just use pre-meal #s for basal and post for bolus ratios. It's not precise but I think it's more important to watch it and engage with it and that success can follow that sort of pattern. I agree with Zoe that if you're meeting one low with a basal shift, it might be hard to discern whether or not the change is working. Sometimes, I seem to really run low the next day, not like super low but just sort of 60s on rails the day after a change but then it calms down and gets used to it. It's also very hard to separate "overlap" out of stuff. 7 hours for sure would be a basal. I always liked running my basals a bit hot and snacking to keep up but getting it flattened out is really ideal.

They have pie chart reports w/ the Medtronic pumps that are neat for seeing "ok, I get a bunch of lows after lunch, I'll turn that down" which, in turn, can illuminate say highs at dinner time (correction hangover...) which, in turn lead to more lows sometime later. Sometimes fixing an ur-low can have a ripple effect and lead to discoveries down the road which is a good feeling.

I would avoid labeling yourself as "insulin sensitive" as if it were a negative thing. Whatever it is, the whole thing is about balance and keeping an eye on things. Take enough insulin to make your BG behave and, if you take a little more and run low, you know you're in the right spot! I also recently tried an experiment, which I've blathered about elsewhere, where I did every hour 1/2 hour .8U/ hour and 1/2 hour .775U/ hour, to "fake" a .7875U/ hour, splitting the difference. So far, it's worked really well. I think I may end up turning up some ratios that were "cooked" low to cover or be covered by the .8U which seems to be a bit too much but .775 and I see a rise. I dunno exactly what you can get out of the Omnipod in terms of low rates but that sort of increment seems to be helping me. Happy hunting!

Another thought crossed my mind: If you feel you're too sensitive to the insulin you're using (I assume you're using a popular analog like Humulog or equivalent), have you considered experimenting with regular human insulin (Humulin R/Novlin R)? I've never tested this in a pump, but I'll use it for MDI for certain foods or situations because of it's milder onset. Or alternatively, I think there are diluted formulations of popular insulins for pediatric uses (U50, I believe).

I also don’t make basal changes until I see about 3 days of patterns. I will use my temporary basal rate for unusual situations. After I make a change I usually wait about 3 days and do some basal testing to verify setting changes.

It has been at least 2 years since I needed to use a long acting basal insulin and then it was only two days but if I remember correctly I needed significantly more Levemir unit for unit than I did rapid insulin in my pump. I want to say about 10 to 20% more and even then it wasn’t enough and I was happy to get my replacement pump.

Hi Sarah,
when you changed the basal rate, for example at 6pm, you would expect the change to start to take affect about 2 hours later. But to make the change effective for tomorrow, at 6pm, you would change the basal rate at 4pm to be less.

I assume you started the pump after your last Levemir dosage ran out ? Otherwise your low may be that you still had some Levemir.

Are you starting back with the basal rates you previously used ? Sounds like you used the PODS, then switched to MDI, and are now switching back ? Or are you starting with completely new basal rates ?

Keep good logs of your rates, BGs and the changes you're making. Consider doing food or bolus corrections for the first couple days, noting the time. Then go back and tweak the basal to match the time of the corrections.

Refer to the books Think Like a Pancreas or Pumping Insulin if you have them. They go into alot of detail on basals.
Or you can check this link,
Basal Testing - Integrated Diabetes link

How does one know.... You just have to keep tweaking and testing, and repeat.
As others have mentioned, trying to see a pattern of 2-3 days before making changes will help reduce the tweak/test/tweak cycles.

so, if you just started on a pump and was dropping, dropping dropping at 10 - 11pm at night and you knew it was only basal, as you hadn't bolused for over 5+ hours, and were eating trying to bring your BG's up but they kept dropping, you'd go to bed knowing you probably have too much basal on board and not change it? Or, the just start on a pump, a random set basal rate...go to bed at say 120...and wake up to 400 the next morning, clearly not ENOUGH basal programmed're seriously gonna wait 3 days to change that..OK? yes, most people need much more MDI basal then they do basal w/pump. It's my understand, when first starting out on pump...if you're dropping too much you don't wait, you change the basal rate. it's not like MDI where it takes 3 days for the levemir - lantus to take effect and get into system. this is pumping with fast acting, it's pro's and con's are fairly straight forward right away. However, to set different basal rates, etc...yes...we're supposed to look for 'patterns'.

My question long does a basal rate change take effect? Does it take an hour - 2 hours a 1/2 hour for changes to take effect, either to lower or raise blood sugars? thanks! yes, mos

It would take a couple of hours, like the insulin curve thingy, to take effect. If you are certain you’re low, at 5 hours post-parandial, I would run the basal down another click or two since it’s bedtime.

thanks, mmxx! I was told an hour..but I'll ask CDE tomorrow. yeah, probably do still have some levemir on board, too. i log everything, have since day 1, also have a CGM. I've also read book those books, too. THANKS!

thanks, AR! I'll see what I'm at a bedtime..unless I'm really high I probably won't correct and just check it overnight...I have my CGM, too.