Transition of Patients to and From Insulin Degludec

A Clinical Challenge

Irl B. Hirsch; Boris Draznin


J Clin Endocrinol Metab. 2020;105(6) 

In This Article

Cases and Discussion

Case 1

A 55-year-old man with type 1 diabetes is admitted with pneumonia and empyema. His hemoglobin A1c (A1C) is 7.8% (61.7 mmol/mol). His insulin dose at home is 20 units of degludec and premeal lispro at a 1:10 insulin:carbohydrate ratio and 35 insulin sensitivity factor. The only basal insulins available on the hospital formulary are glargine and NPH. Insulin lispro is also available. He will require at least 3 days of IV antibiotics. How would you suggest his insulin be managed in the hospital?

Replace degludec with glargine 1:1 22 (3 individuals recommended splitting glargine)
Reduce glargine dose 2 by 50%, 1 by 20%, 1 by 10%
Increase glargine 2
Weight-based glargine 1
IV insulin and transition with NPH 1

Responses to Case 1. Several studies that compared clinical efficacy of degludec and glargine on fasting plasma glucose levels in patients with both type 1 and type 2 diabetes found either nonsignificant differences or greater efficacy of degludec (P < .05) at what some would consider a nonsignificant 8 mg/dL (0.42 mmol/L) decrease in glycemia.[11]

This would suggest that 1:1 equivalency in degludec and glargine doses is reasonable. However, if a patient in a steady state took 40 units of degludec on day 1, and glargine was substituted on day 2, 50% of degludec (20 units) would be contributing to the serum insulin level at the time of the glargine injection (Figure 1C). If 40 units of glargine are administered at that time (24 hours after the last degludec injection), the total insulin concentration would be 60 units: 20 units of remaining degludec still on board and 40 units of glargine. Obviously, transitioning to glargine with degludec on board is fraught with problems of insulin stacking.

Starting glargine in 1:1 ratio to degludec 24 hours after the last degludec dose would result in high levels of basal insulinemia because of the long PK of degludec. With this approach, for the first 24 hours of glargine, the patient will still have 50% of the previously administered degludec contributing to insulinemia during the next 24 hours. It would be reasonable to halve the dose of glargine (20 units). For the second 24 hours, depending on the glucose levels, one could give 75% to 100% of the dose of glargine (30–40 units). Furthermore, particularly in patients with type 1 diabetes, because the duration of the glargine does not always last 24 hours (dose dependent), it would be reasonable to split the glargine dose into every 12-hour dose schedule because this would also give more flexibility. For transitioning back to insulin degludec, it may be wise to give a fraction of the glargine dose with the first degludec injection because after 3 days most of the original degludec will be out of the system.

Case 2

A 65-year-old man with type 2 diabetes, A1C of 8.0% (63.9 mmol/mol), receives 45 units of insulin degludec, metformin, and semaglutide. He is scheduled for a colonoscopy. What changes in his diabetes medications would you suggest before his procedure?

Same dose of degludec 10
Reduce degludec by 10% 1
Reduce degludec by 20% 8
Reduce degludec by 2%–50% 6
Reduced dose of glargine 1
Need more information and/or continuous glucose monitoring data 4

Responses to Case 2 (Degludec).

Continue metformin and semaglutide 6
Hold metformin 21
Hold semaglutide 2
Reduce semaglutide by 50% 1

Responses to Case 2 (Metformin and Semaglutide). The concern for fasting and daytime hypoglycemia from too much basal insulin during colonoscopy is legitimate. Basal insulin is considered adequate when glucose levels are stable overnight (assuming no prandial insulin or food near bedtime). Reducing degludec the night before the procedure will have minimal to no effect on the next day's insulin levels (or blood glucose). Furthermore, because injectable semaglutide is a weekly drug,[12] there is no practical way to change that. Therefore, our recommendation for this procedure is to not change the insulin degludec or semaglutide, assuming the patient does not receive excessive amounts of basal insulin. It is reasonable to withhold the metformin on the day of the procedure.

Case 3

A 23-year-old woman is switching from multiple daily injections to insulin pump therapy. She receives 12 units of degludec with premeal aspart and her A1C is 6.6% (48.6 mmol/mol). How would you transition her from degludec to her pump?

Reduce total daily dose of insulin by 20% 5
Reduce total daily dose of insulin by 50% 1
Insulin pump basal rate 0.4 U/h 7
Insulin pump basal rate 0.5 U/h 5
Insulin pump basal rate between 0.4 and 0.5 U/h 6
Need more information and/or continuous glucose monitoring 6

Responses to Case 3. When a patient on degludec initiates insulin pump therapy, one must keep in mind that, even 48 hours after stopping degludec, almost one-quarter of its steady-state concentration is still on board. Any basal insulin administered via the pump would then be in addition to that degludec.

There are several possible ways to manage this patient. One way would be to withhold the degludec on the day before starting the pump, so that 24 hours after the last dose of degludec is given, serum insulin levels are at about 50% of previous steady state. For that first day without degludec, correction doses of short-acting insulin could be provided. When starting the pump on the next day, only one-half of the degludec would still be available so for that first day on the pump, basal rates could be reduced by 50%, which would be 6 units over 24 hours or 0.25 units/h for that first day on the pump. After that, usual insulin doses could be used. A more aggressive basal insulin reduction would be required if starting the pump on the first day not receiving the degludec. Because in the first 12 hours degludec insulinemia will be reduced by a total of 25% and the second 12 hours by another 25%, it would be reasonable to reduce the basal pump rate by 75% for the first 12 hours (0.125 U/h) or even the entire first 24 hours. Alternatively, one could reduce the basal insulin by 50% for the second 12 hours (0.25 U/h).

Case 4

A 66-year-old man with preexisting type 2 diabetes was admitted for nephrectomy and partial pancreatectomy. For his diabetes, he was treated with IV insulin for 3 days before clear liquids were started. On day 4, he is advanced to full liquids in the morning and that night he is tolerating a regular diet. At home, is he treated with 35 units of bedtime degludec and premeal insulin lispro (A1C 7.0% [53.0 mmol/mol]). It is now 8 pm on postoperative day 4 and he is still receiving IV insulin. The hope is to send him home the next morning. How would you manage his insulin for discharge home?

Degludec 35 units at bedtime 7
Glargine 35 units at bedtime 6
Reduce degludec or replace with glargine 1:1 at bedtime 5
Recalculate degludec or glargine based on IV drip rate at bedtime 4
Keep extra day in the hospital 1
Recalculate degludec or glargine by weight based 2
Need more information 5

Responses to Case 4. As in case 3, it is very reasonable to obtain more information and, if possible, to delay this patient's discharge by another day. Although it is certainly correct to calculate the new basal insulin dose on the basis of the last 6 hours insulin drip rate, it would also be safe to switch him directly to long-acting insulins. There are several possible ways to do this. Because he has type 2 diabetes, there is little to no risk of diabetic ketoacidosis, so simply giving his degludec at the usual time will only result in higher glucose levels, especially the first 24 hours. A better strategy would be to give him a "bridging" dose of glargine before discharge from the hospital. Going directly from IV to glargine, it should be safe to give an entire dose of basal insulin (35 units), as his prehospitalization dose of degludec was. This patient did not receive degludec in the hospital and at the time of his planned discharge on day 5, he would have no remaining degludec in the circulation. Still, many would only give one-half the dose of glargine (eg, 17 units) for that first bridging dose before the next degludec would be administered. This should be an easy solution for an inpatient when the glargine is available.

Case 5

A 45-year-old man with type 1 diabetes and a transplanted kidney is admitted for acute rejection. He will be treated with in the hospital until discharge with 40 mg of prednisone each morning, with the plan to taper the dose as an outpatient. Home insulin includes 25 units of bedtime insulin degludec and premeal aspart. The only basal insulins on the inpatient formulary include glargine and NPH insulin. How would you manage his insulin in the hospital?

Degludec or glargine 1:1 9
Degludec or glargine 1:1, add NPH 12
Increase glargine and aspart 1
Decrease glargine and add NPH 1
NPH alone 6
IV insulin 1

Responses to Case 5. With degludec not available in the hospital, we have the same problem with stacking of insulin, particularly in that first day, as we had in case 1. Based on the PK of the glargine and the degludec, for that first night when the degludec would normally be injected, it would be reasonable to wait 24 hours when approximately 50% of the last dose of injected degludec has been metabolized.

Because prednisone is administered every morning, NPH should be given at the same time to counteract the effect of steroids. One of the easiest ways of estimating the NPH dose is to give it in the following manner:[13] 0.1 U/kg for 10 mg of prednisone, 0.2 U/kg for 20 mg, 0.3 U/kg for 30 mg, and 0.4 U/kg for 40 mg of prednisone. Subsequent adjustments are required during the follow-up. In some patients receiving NPH in the morning, it might be reasonable to wait a day to give the first dose of glargine and use correction doses of short-acting insulin, if needed.