Blood Sugar Increases Following New Pump Infusion Site Insertion

Since we got OT in another thread on this topic, I thought I’d start a new one. I was looking for a scientific study about this, but I failed to find one. However, I did find this, which verifies what several of us have reported anecdotally:

Quoting a phone interview with Jennifer Smith of Integrated Diabetes Services, the author of that article writes, “When you switch to a new site, there’s a lag in absorption. You have to experiment with site changes because some sites work better than others.”

This agrees with my experience. (I do not have an Omnipod, I have a Tandem. I also saw this with my previous Medtronic pump.)

From my experience, when it happens I start going higher very slowly. It usually doesn’t exceed 180 for me but just keeps creeping up from my usual 100-120 morning reading to 160-170. Small doses of insuline (say 2 or 3 units) seem to have little impact on this slow rise. It lasts for about 3-4 hours. Then… it drops and continues low for 6-8 hours. (Not <70 low usually unless I overdose on insulin, but an unusually tenacious 80-90 low) It’s as if all the insulin was being stored in the tissue, dribbling out for 3 or so hours and then suddenly the flood gate opens. Of course, YMMV. I do find less frequently used sites are usually (but not always) less problematic than well used sites.

The OT discussion can be found in this thread: (Which, by the by, is a poorly titled thread because it’s really about the acceptability of prefilling cartridges and personal experiences of being able to rely on insulin in the pump exposed to temperatures above recommendation. But that would be a ridiculously long title.)

Here are suggestions from the Diabetes Educator website for dealing with this problem. (As an aside, this tacitly recognizes the problem as such being recognized by people who are in frequent contact with multiple T1s.)

This happens to me often enough that I always leave the old site in for a few hours so I can do a quick switch and knock a rising blood sugar down. I had a problem with the old site yesterday, so I had to remove it and I couldn’t do that.

1 Like

I don’t experience this but it sounds fascinating (as an outside observer- personally I hate having to deal with this kind of ‘weird diabetes’ stuff myself!)
I wonder if perhaps blousing anywhere from 30 minutes to immediately before changing the site would help curb that rise? Vs blousing after changing the site?

Another suggestion - would massaging the site before insertion help? Or after but with caution so as not to disturb the cannula.:thinking:

I wonder if it isn’t something that affects those of us who have been using pumps longer? Back when I was first on the pump, we didn’t have CGM feedback, so it was more difficult to accurately track the progress of blood sugars without doing lots of finger sticks. I didn’t get CGM until my fourth or fifth pump. So I don’t really know how prevalent this sort of response may have been when I first started on the pump. Only three of us are reporting the problem (me and @Dennis, both of whom have been on pumps for a long time, as well as @joe, who reports other issues that seem to affect his sites).

I’d blame overuse of sites, but within the past few years I started using previously unused locations for insulin injections and I have noticed the same problem in those locations. (Well… I used to use some of them back in the bad old days of syringe injections, but your body regrows new skin every 7 years, so I think I can safely call them unused.)

I’ve had t1d 37 yrs, pumping about 5 yrs, and my 13 yo son has been pumping 3 yrs. We definitely notice these issues. I get a bit of anxiety at site change time, wondering how it will work. It’s a wait and see, esp for my son. We rotate sites pretty well…and do his site chnages every 2 days, as he seems to lose absorption day 3.
I’ve had another, similar question, that his endo somewhat suggested…can there be pooling and dumping at times? When we get a stubborn high…dose a few units, no response, dose a few more…that cycle can go for 2-3 hrs. Then, get active, and slam! It’s like the insulin wasnt effective for hours, but literally, in 5-10 min of something active, will crash down. Is the insulin pooling at the site and the increased movement releases it?

1 Like

@T1Dx2 ,
I’ve noticed that happening too, but didn’t connect it to site absorption issue.

@MarkCK I’ve noted similar issues when changing from my abdomen to my left arm. I use Omnipod Dash and have been using it about 18 mos or so, so don’t think it’s a “long users” thing. I tend to rotate pod location from my abdomen’s left side to the right side of my navel (I usually use similar location twice, once with the cannula to the left, then the right; or first “up” then “down”), then the inner left bicep area, then to the back of the left bicep area (tricep?). I note a significant rise, as much as 40-50 pts, in BG’s for much of the first day when switching to or among the left arm locations. It’s like the flow from interstitial fluids needs to get established. I often, but not always, need to due corrections more aggressively during that time as well. I’ve read several articles that state the abdomen tends to best the absorption location, it certainly seems true for me. But it also seems to be one of those things that varies by individual as well as how active people are.

I also recall Jennifer Smith has said she boluses a 1-2 units + when changing pods/sites. I’ve not found that necessary on abdomen locations, but I’ve experimented doing it when moving to my arm and it seems to help somewhat. Ref @wadawabbit’s suggestion of massaging, I do that routinely on my arm locations; it seems the insulin is “stuck” in my arm, and massaging should get it to absorb better (if that is indeed the issue), sometimes it seems to help, sometimes not, nothing consistent.

This is what I sometimes notice. (It’s not always like that, only sometimes.) It seems quite probable. Massaging would seem to be helpful.

Back in the days of injections, I would sometimes get a little knot of insulin under the skin that you could actually see and feel when injecting NPH. That definitely required massaging to work out.

I have definitely experienced this with Omnipods. I usually bolus before changing pods, and/or set an increased basal rate for an hour after changing. I assumed it was due to tissue trauma from the insertion. Curious if this phenomena happens with other pumps, or if it’s unique to Omnipod.

No, as I mentioned before, I have experienced it with both Tandem and Medtronic pumps. I actually did ask my diabetes educator about this when I got my last pump and she said something like the body sometimes fighting an unrecognized foreign element (meaning the insulin) at the new site. (That’s rather clumsily stated and I don’t recall exactly how she put it which is why I didn’t mention it in my previous posts. Plus it didn’t completely make sense to me. Why at one site and not at another?)

I’m lucky like @6yGodsGr, this issue rarely happens to me. When it does, most times I find that I’ve nicked a blood vessel which didn’t bleed enough to be noticed under my Pod until I remove it at 72 hours. My theory agrees with the pooling/absorption; either the clotting bleed or the vessel itself (crimping the cannula) causes just enough restriction to raise blood sugar, but but enough to trigger an occlusion alarm.

However, my diabetic wife has hyperglycæmia after most site changes. She’s 4 years younger than I, and was diagnosed 6 years after I was. We both started pumping the same week in '93 (we met in our pump training group, actually.) Currently she’s T:slimming and I’m Podding, but even when we both used Tandem, the Site-Change-Hyper gets her every time and almost never affects me.

When she used Smith Medical’s Cosmo pump, she had some success using that pump’s “disconnect duration” feature. As I recall, Cosmo delivered a percentage of her current basal based on how long she thought she’d be disconnected from the pump. (As an added bonus, the “stopped delivery” alarms didn’t keep sounding the entire time–only alerting after the estimated disconnect time and disconnect bolus had passed. I don’t think any other pump has ever had this brilliant feature [\mournful sigh]…)

I really appreciate @MarkCK discussing this. I’m going to share your screenshot from the diabetes educator’s website with her as soon ss she’s not working. I know there’s some techniques there that she hasn’t tried.