In July 2024, a 33-year-old woman with type 1 diabetes was boating on a hot day when her insulin device slipped. By the time she was able to get off the river, she was obviously ill and an ambulance was called. The hospital was at capacity. She lay in the hallway and was given fluids but no insulin, despite her boyfriend repeatedly telling staff she had diabetes. She was discharged while still vomiting. The next morning, her boyfriend found her dead.
This story was shared by a friend of the woman in a Facebook group for people with type 1 diabetes and later confirmed by the friend in a separate heartbreaking post. Although this is an extreme case, it is quite common in the healthcare system for people not to know enough about type 1 diabetes and this can sometimes have serious consequences.
I have lived with this disease for over 50 years and have lost count of the number of times I have had to advocate for myself, correct mistakes, raise issues that were not addressed, and even educate non-specialized medical professionals on some of the basics.
Type 1 diabetes is an autoimmune disease that destroys the insulin-producing cells in the pancreas, requiring lifelong insulin treatment. Type 2, on the other hand, is caused by a combination of insulin resistance and reduced insulin production. Type 1 accounts for only 5% of all diabetics, but with a prevalence of about 1 in 200, it is not rare. And that doesn’t even include the adults who are incorrectly diagnosed with type 2 but actually have type 1.
As a general rule, people with type 1 diabetes are more insulin sensitive than those with type 2 diabetes and are more prone to hyper- and hypoglycemia. Blood sugar levels tend to be more labile and less predictable even under normal circumstances. Recent advances in hybrid closed-loop technology have helped greatly in reducing the variability, but the systems are not yet foolproof. They still require user input (i.e. guesswork), so errors are still possible.
Treating type 1 diabetes is challenging even for endocrinologists. But here are some very important basics that every healthcare provider should know:
We need insulin around the clock
Never withdraw insulin from a person with type 1 diabetes, for any reason. Even if we don’t eat – or vomit – we still need basal (background) insulin, either via a long-acting analogue or a pump infusion. The dose may need to be lowered to avoid hypoglycemia, but stopping insulin will result in diabetic ketoacidosis. And if this continues, death will result.
This should be basic knowledge, but I’ve read and heard far too many stories of people with type 1 diabetes being denied insulin in a variety of settings, including emergency rooms, mental hospitals, and prisons. On Facebook, people with type 1 diabetes often report being told not to take insulin the morning before a procedure, and more than one has described secretly giving themselves insulin in the hospital because they didn’t get any or didn’t get enough.
Although insulin requirements are highly individual, the amount needed for type 1 diabetics is usually significantly less than for type 2 diabetics. Too much insulin can lead to severe hypoglycemia. There are many stories of people with type 1 diabetes who have struggled with hospital staff trying to give them much higher doses than they thought they needed.
The American Diabetes Association recommends that people with type 1 diabetes who are hospitalized be allowed to wear their devices and care for themselves as much as possible. And please listen to us as we tell you what we know about our own condition.
Fasting is dangerous
I cringe every time I’m asked to fast for a test or treatment. Fasting carries a risk of hypoglycemia in people with type 1 diabetes, even when the latest technology is used. Fasting should only be prescribed when absolutely necessary, especially for routine laboratory tests.
Saleh Aldasouqi, MD, chief of endocrinology at Michigan State University in East Lansing, Michigan, has published several articles on a phenomenon he calls “fasting-induced en route hypoglycemia in diabetes,” in which patients who fast overnight and skip breakfast experience hypoglycemia en route to the lab.
“Patients continue to take their diabetes medications but do not eat. This leads to low blood sugar levels and hypoglycemia can occur on the way to or from the lab, putting themselves and others at risk,” explains Aldasouqi, adding that fasting is often not necessary for routine lipid profiles.
When fasting is required, such as during a surgical procedure involving anesthesia, the need for insulin adjustment—NOT insulin withholding—should be discussed with the patient to determine if he or she can do it himself or if his or her diabetes doctor should be consulted.
But that’s a tricky business, even for endocrinologists. True story: When I had my second carpal tunnel surgery in July 2019, my hand surgeon wisely scheduled his first procedure for me in the morning to minimize fasting. (He himself has type 1 diabetes, which helped.) My endocrinologist had advised me, per guidelines, to reduce my basal insulin infusion via my pump by 20% before bed.
But before bed, my continuous glucose monitor (CGM) showed a reading of 170 mg/dL and rising. This was not really surprising since I had reduced my insulin dose before dinner, knowing that if my reading got too low later, I would not be able to eat. I did not reduce the basal rate.
When I woke up, my blood sugar was over 300 mg/dL. This time, stress was the likely cause. (This has happened before.) Even though I had given myself several small insulin boluses that morning without eating anything, when I arrived at the hospital, my blood sugar was still around 345 mg/dL. The nurse told me that if it had been over 375 mg/dL, they would have had to cancel the surgery, but it wasn’t, so they went ahead with the surgery. I have no idea how they got to that cutoff point.
Anyway, thankfully everything went well; my blood sugar levels went back into target range afterward and I healed normally. The point is that treating type 1 diabetes is a crazy balancing act and the guidelines only go so far.
We do not respond well to steroids
If for some reason it is absolutely necessary to give steroids to a person with type 1 diabetes, a plan must be made in advance for the inevitable blood sugar rise. If the person does not know how to adjust their insulin accordingly, ask them to consult their diabetes educator. In my experience, the rise with locally injected corticosteroids is always higher and longer than expected. Fortunately, I have not had to deal with systemic steroids, but I suspect they are probably worse.
Procedures can be annoying
People who wear insulin pumps and/or CGMs must remove them for MRIs and certain other imaging procedures. In some cases—such as CGMs and the Omnipod insulin delivery device, which cannot be re-attached once removed—advance planning is required to bring backup devices for immediately after the procedure.
Diabetes devices can remain in place for other imaging tests such as X-rays, most CT scans, EKGs and ultrasounds. For heaven’s sake, don’t ask us to remove our devices unless it’s absolutely necessary.
In general, surprises that affect blood sugar are not a good idea. I recently underwent a gastric emptying study. I knew the test would involve eating radioactive eggs, but I didn’t learn that there was also a jam sandwich with two slices of white bread until the technician handed it to me and told me to eat it. I had to inject insulin quickly, and of course my blood sugar later spiked. Had I been forewarned, I could have at least given myself a “pre-bolus” 15 to 20 minutes beforehand to give the insulin more time to work.
Another anecdote: Before a dentist appointment where my gums had to be numbed for a thorough cleaning, my long-time dental hygienist told me, “Make sure you eat something before you come in.” I appreciate that she thought about my diabetes. However, this advice would have made sense a long time ago, when treatment involved two daily insulin injections with no dose adjustments, but today it is more complicated.
Today, when we eat carbohydrate-containing foods, we usually take short-acting insulin, which can lead to hypoglycemia if the dose administered exceeds the amount needed for the carbohydrates, no matter how much is eaten. Better to eat nothing at all (assuming the basal insulin dose is correct) or to eat only proteins. And for the doctor, it is best to simply inform the patient of the eating restrictions and make sure they know how to deal with them.
Sure, we already have diabetes
I have heard of at least four cases where pregnant women with type 1 diabetes were required to do an oral glucose tolerance test to detect gestational diabetes. In two cases, it was a “Can you believe this?!” post on Facebook that led the women to rightly refuse the test.
But in May 2024, a pregnant woman reported that she had actually drunk the liquid, her blood sugar levels spiked, she was vomiting, and was trying to lower her blood sugar levels at home with insulin. She had no objections to the test because “my gynecologist knows I have diabetes,” so she thought it was appropriate.
I don’t work in healthcare, but here’s my guess: The gynecologist hadn’t actually ordered the test, but rather neglected to reverse a routine order for a pregnant patient who already had diabetes and clearly should NOT be forced to drink a highly sugary liquid for no reason. If this happens in pregnancies with type 1 diabetes, it certainly could happen in pregnancies with pre-existing type 2 diabetes. Clearly something needs to be done to prevent this unnecessary and potentially harmful scenario.
In summary, I think I speak for everyone living with type 1 diabetes when I say that we want to have confidence that healthcare providers in all areas can care for the patients who brought us to them without adding to the daily burden we already carry. Let’s work together.
Reviewed by Saleh Aldasouqi, MD, chief of endocrinology at Michigan State University.
Miriam E. Tucker is a freelance journalist based in the Washington, DC area. She writes regularly for Medscape Medical News and has published other articles in The Washington Post, NPR’s Shots blog, and Diatribe. You can find her on X as @MiriamETucker.
