Conditions Where Insulin Pumps May Not Deliver Intended Doses

One of the most maddening aspects of managing T1D is the erratic and unpredictable variability in the absorption and responsiveness of insulin, which can wreak havoc on glucose control. But a related factor that isn’t really considered very often—if people even know about it—is the possibility that insulin pumps may not always deliver the amount of insulin they’re supposed to.

Investigating medical literature, I discovered published papers on similar situations dating back to 1994, citing mechanical problems with insulin pumps that have long been known, but rarely discussed—or addressed. In fact, the technical inserts included in the product packaging for all pumps provide warnings on insulin delivery variability, though not to the degree of detail provided here, nor with explanations on why.

This article aims to explain the mechanical deficiencies with pumps described in published medical journals, what can be done to fix them, and whether that’s even enough. (Again, nothing in T1D is that simple, so don’t expect miracles.)

But before you gaze suspiciously at your own pump, everything here should be taken into a much broader context: You may well be very happy using an insulin pump, and don’t see any reason to be concerned. What’s more, if your own A1c levels—or those of your child—are consistently north of 8%, then automated insulin delivery (AID) systems may well be a wise choice, despite some of these problems.

2 Likes

I feel like subcutaneous infusion is definitely a weak link in diabetes management. So much variability! It’s been a long time since I did MDI but my memory is that it was similarly plagued by variability as is infusion with a pump. Bad injection sites. Injected insulin leaking out. Insulin seemingly not absorbing for a while then absorbing a lot at once.

Definitely a lot of room for improvement with infusion sets and pumps but I feel like implantable pumps would be the only way to really overcome the problem and they’re probably not likely in the near future.

On a different, but slightly related note, I’m intrigued by the Eversense implantable CGM. I like my Dexcom pretty well, but every ten days it’s a new adventure of about 24 hours waiting for the thing to settle in and start working reliably. A couple days ago I had a G7 that seemed to be going along just fine when about 5 days in I got a message saying that the sensor had encountered some kind of fatal error and it had to be replaced. At 1am! The new sensor didn’t start giving reliable readings until the following evening. A whole day of fingersticks and false lows. It’s exhausting being a cyborg!

5 Likes

exactly – it’s the source of so many problems with T1D, not just insulin, but CGM data. (See my article comparing the G6 and G7 here, which provides a primer on systemic glucose variability and why it’s hard to measure it.)

Subcu tissue is also responsible for the upper limit on how well AID systems can work–the erratic behavior you mentioned introduces sporadic, chaotic, and random variability in insulin absorption that (by definition) cannot be predicted algorithmically. There will always be error within some degrees that can spin out of control if the algorithm predicts one set of outcomes, but the subcu tissue presents a different outcome. (See my article, Challenges Facing Automated Insulin Delivery Systems.)

There are two ways to significantly reduce the subcu tissue problem with insulin: The first is to use an inhaled insulin like Afrezza. It acts immediately and fully. In, out, done. Once it’s in the lungs, it gets absorbed directly into the bloodstream. Of course, if you have insulin resistance from visceral fat or other metabolic conditions, the insulin performance may not be sufficient. In this case, the problem is identical to T2D, but that’s a whole different topic.

The second way to avoid the subcu tissue problem is to use an insulin pen (or old school syringe) and inject directly into muscle, which is highly vascularized, and does not have the same properties that cause lipodystrophy and other skin conditions. This yields almost identical performance to Afrezza.

Intramuscular injections has been highly discouraged by endos and diabetes educators for years, largely because it can lead to hypoglycemia (but only if you take too much). Since the absorption rate and efficacy is exactly the same as Afrezza, why don’t they express the same concern? (Well, because Afrezza doesn’t come in very large doses, whereas you could hypothetically take too much insulin at one time intramuscularly.)

I’ve been on MDI and do intramuscular injections (legs and arms) and get very good absorption and achieve tight control.

Note: you might get bruising now and then, but this goes away within a day or so. It’s a trade-off I’m willing to accept.

1 Like

I don’t know what to do about variable absorption, so I try to eradicate variable delivery. For me, being aware of variable absorption is good enough to maintain control, as long as I control for variable delivery by using reliable hardware. That’s not true for everyone. Some people see significant BG drop every time they step into a hot shower. That sucks.

You should assume (at least) 20% variability when you use any manual BG meter. We have all seen the study showing those devices fall outside of the range mandated by FDA. Diabetes Technology Society
More recent studies may be a better reflection of the current state of the tech.

CGMs have more error than that. Sometime much more - the new OTC CGMs (Abbott Labs) being released aren’t even approved by FDA for T1. Instead of building tech with increased accuracy and precision, they are engineering more inaccurate devices and attempting to market them to nondiabetics. Its an odd strategy because the error in measurement is likely to out weigh any change in BG measurement seen by a nondiabetic. Its snake oil. Its an odd strategy even without that - they are trying to invent a market for a nonfunctional device.

yeah–but for reasons many people aren’t aware of as well. The article cited earlier gets into considerable detail on that: Continuous Glucose Monitors: Does Better Accuracy Mean Better Glycemic Control?

The OTC CGMs are fantastic for non-diabetics. They don’t need accuracy or frequent data. They just need to see extremes and trendlines, while also getting long-term glucose averages. I have had non-T1D friends that have used CGMs to see what kind of foods or exercise does to their glucose levels, and it makes them much more cognizant about their health.

My endo highly discouraged me from doing it but not because of hypos. He referred to damage to muscle and nerves that can occur and to cases he had seen in hospitalized patients. He did not imply that this was by any means common and he said he had patients who did IM without incident. It was, however, enough to scare me off.

I mention this not to discourage you or anyone else from doing something that works or because I have any knowledge of this particular risk. It may be very rare. I just mention it so that people who may have never done it are aware that risks other than hypos may exist and that they should ask their doctor or look into it before going that route.

I spent the last few hours researching this in medical literature, and found no references whatsoever where intramuscular injection (IM) causes any sort of damage to tissues or nerves.

To the contrary, I found considerably more literature citing studies and medical trials demonstrating that IM injection is beneficial: absorption rates are not only faster than subcutaneous injection, but the rate can even be more rapid if you exercise the muscle directly afterwards (e.g., inject in the leg and then walk briefly for 5-10m).

Moreover, while lipodystrophy can still occur, its severity, rate of incidence, and rate of healing is far superior to subcu tissue.

I’m glad you brought this up, as it’s been well over ten years since I researched this, and adds to the bibliography that I’ll be using for my next article on the topic.

1 Like

That’s me. There have been occasions on which I took a hot shower instead of taking more insulin when I was trending higher than I would like, specifically for the BG drop.

I have been pumping for 34 years and doing the diabetes job for 54 years tomorrow! And way back at the start, it was all about rotation! We were given charts for all areas where you could give shots, making it unlikely to hit the same spot for a long time. I was told the same thing when switching to a pump, it’s all about rotation.
I have never had a major problem with any pump I have used. But I have had problems with infusion sets. Infusion sets can make or break a pump experience. I believe getting as many samples as possible to see what works for you and to get a training sessions with correct insertion procedures.
Over the years I have used a number of different infusion sets, and have finally settled on the TruSteel, with shorter needle length. Metal doesn’t bend, so I have no issue with the biggest problem with infusion sets.
Again, what works for me might not work for the next person. And it did take me awhile to circle back to metal. They were much longer at the beginning of my pumping experience. When I was have issues with the Teflon ones, I contacted my pump trainer and got samples of some different ones which lead to the metal one.
Pumps are great if you realize the downfall of a good pump is going to be the infusion set.

3 Likes

I don’t believe they have the accuracy to provide any valuable information for someone who has a 70 - 120 range with no extremes. We already know what the average is. A non-diabetic would require more precision and accuracy to see anything, not less. Plus, what use would the info be to them? Maybe its useful if they are hypoglycemic (where the thing looses accuracy). Otherwise, it feels like they are just imagining relevant data is being returned.

Ok, so I need to clarify what I meant when I said “non-diabetics.” I should have said “non T1Ds.” For reference, there are about 2 million T1Ds in the US.

According to the ADA, there are 38 million T2Ds, though there are other reports putting that number of to 48 million, if you move the A1c level to 6.5%. Also, 97.6 million Americans age 18 and older had prediabetes–A1c’s between 5.7% and 6.5%.

It’s also the eighth leading cause of death in the United States (2021), so treating T2D is a big deal.

While many take insulin, the first line of treatment is usually diet (and exercise, if people are willing to do it). Therefore, these folks need CGMs so they can see the correlation between what they eat and the effects on their glucose levels, and this is the primary target audience that all CGM manufacturers are aiming for with their new OTC (over-the-counter) CGMs.

And then there’s a large number of people who suffer from “metabolic syndrome,” which is a catch-all phrase that can encompass any number of diseases or hormonal dysfunctions that can lead to volatile glucose levels, whose treatment is vastly improved when their glucose levels are monitored.

As for everyone else whose A1c’s are below 5.6, even they have interestingly volatile glucose patterns that researchers are trying to understand. This 2015 study from Stanford University had researchers provide healthy, non-diabetic subjects with CGMs for two weeks, and regularly gave them exactly the same food and activities at the same time each day. Researchers found large variations in blood sugar levels, not just between each other, but the same individuals had erratic patterns. CGMs are being used as a research tool for a large number of metabolic dysfunctions, which will ultimately give us great lessons that may one day apply to T1Ds. (Follow the link for graphs that illustrate glucose patterns and what the researchers had to say about it.)

As for your statement about people having regular glucose patterns between 70-120, I’ll take a random stab at estimating that such individuals are as rare as T1Ds. We live on opposite ends of the glucose spectrum.

“Some people see significant BG drop every time they step into a hot shower.”

Huh. I’ve never heard of this.

Slightly off topic, but I’ve been getting a significant BG rise every time I have a hot shower, especially if I’m already a bit high. I just assumed it was part of my morning cortisol struggle. Could there be another explanation?

FYI. I have a 20 minute hot shower every morning so that I can do a series of stretches to help my fibromyalgia. I’ve learned to give a manual bolus before I step in if I’m around 7mmol or higher. To keep the insulin flowing, I also wear my pump in an AquaPac, a plastic pouch with a belt that protects my pump from the water. This helps, but I’m always surprised by how fast my BG rises.

Insulin absorbs more quickly when exposed to heat. The phenomenon is exacerbated for lantus and other basal insulins, as the heat breaks down the chemical bonds that allow for the slower release. Hot tubs are particular causes of severe hypo events.

People will often get into a hot shower or sauna intentionally to treat hyperglycemia. You can also vigorously rub an injection site (MDI) to accelerate absorption. It’s also possible to apply a hot towel. (Put a wet rag into the microwave for a minute or so.)

Pump users can experience the same thing, but the type (and amount) of rapid acting insulin has an effect on the size and degree of increased onset. Usually, pumps drip more slowly, so there’s a smaller amount may not create as apparent an effect for some as it might for others who use very low basal rates than those with larger doses.

Similarly, faster acting insulins like Fiasp already absorb somewhat more quickly than, say Humalog (true of MDI or pumps, obviously).

1 Like

This post, and article links have, after 33 years of struggle, vindicated my son’s and my dismay with T1, glucose control, instability, and pump/CGM issues. We’ve searched and searched for answers and research papers knowing that it IS nearly too complicated to manage. We have found a few good research articles relating to oxidative stress, etc. Also it seems that most docs (endos) are often busy to keep up patient load … with no time to review the research.

1 Like

Somewhere in this thread, or within the linked articles, I thought I saw input from person who is MD-Endo and lives with T1D. I cannot find that reference today and I am hoping someone will reply with a pointer. Thank you

I found the reference I have been seeking, here it is: Severe hypoglycemia among peop - Diabetes Care "On Air" - Apple Podcasts

Right – I linked to it in my article (that this thread is about). If you asked earlier, I didn’t see it quickly enough.

Back when I was first diagnosed, they used to call an injection straight in as an “intramuscular injection.” I said that once to tech support for my pump, and they had no idea what I meant. :slight_smile:

FYI: "Airplane travel consistently causes insulin pumps to over-deliver a little over half a unit on takeoff and under-deliver a bit less on landing, new research found.

This phenomenon is due to air bubble formation and reabsorption in the insulin caused by ambient pressure changes in the airplane’s cabin. It has nothing to do with the pump itself and happens with all insulin pumps, including those in hybrid closed-loop systems, Bruce King, MD, said at the European Association for the Study of Diabetes (EASD) 2024 Annual Meeting."

I read something about this. I believe we’re in real trouble if there’s a loss of cabin pressure. Almost six extra units! I had hoped the design of the Tandem T-slim might protect from this, but, if it does, the article doesn’t mention it.