Advances in blood glucose (BG)–monitoring technology have the potential to decrease blood sugar excursions, which is especially important in patients utilizing insulin pumps. Insulin pump manufacturers use proprietary algorithms for their BCs to adjust insulin doses based on carbohydrate load (CHO) and the amount of insulin that is currently on board (IOB). However, little is known about how the various BCs used by the different insulin pump manufacturers affect insulin bolus doses.

Investigators conducted a simulation study in which they tested three commercially available insulin pumps' BCs to compare recommended bolus doses. Using the Insulet Omnipod UST 400, Medtronic MiniMed 670G, and Tandem t;slim X2 insulin pumps, the researchers programmed each device with the identical settings. The insulin pumps were used in simulated manual mode in which BG and CHO values were entered to determine dosing recommendations since they were not applied to patients. Each device was set to the same insulin-to-carbohydrate ratio (ICR), target BG, insulin sensitivity factory (ISF), and active insulin time (AIT).  

The ICR represented the units of insulin per grams of CHO. The ICR was set to 15, which meant that 1 unit of insulin was administered for every 15 grams of CHO entered. The ISF was the unit of insulin for every mg/dL of BG that was above target. The ISF was set at 50, indicating that 1 unit of insulin was administered for every 50 mg/dL increase of BG above the target goal. AIT was how long bolus insulin was expected to remain active in a hypothetical patient's body; this was set at 3 hours. The target BG was set at 100 mg/dL. A "reverse correction" was also programmed if the BG was below goal in to prevent hypoglycemia.

The following scenarios were simulated: BG of 100 mg/dL with administration of 60 grams of CHO, BG of 200 mg/dL with administration of 30 grams of CHO, and BG of 300 mg/dL with no additional CHO. These scenarios resulted in a dose recommendation of 4 units of bolus insulin for each pump.

Differences were observed between the pumps when, at 1 and 2 hours postbolus, BGs were set for 60, 75, 150, 250, and 350 mg/dL with varying CHO loads (i.e., 0 grams, 30 grams, and 75 grams) administered. One and 2 hours postbolus were chosen because insulin would still be on board from the bolus since the AIT was 3.

A total of 108 suggested bolus doses for each BC at the two timepoints were observed. Using the parameters mentioned above, the researchers were able to calculate the proprietary formulas used by insulin pump manufacturers for their BCs.

The investigators found that when the subsequent BG was above the target BG, all three pumps varied slightly in their bolus doses. Similarly, the differences were not great when the subsequent BG was a target. The largest variations between recommended bolus doses were seen when the subsequent BCs were below target, with differences as high as 10-fold (015-1.5 units). Finally, there were differences in recommended insulin doses by the BC depending on the presence of IOB.

While this study has limitations, including that these BG patterns were simulations and that assumptions were made about the insulin decay model, this study represented the first published head-to-head comparison of specific dose recommendations from pump BCs based on simulations. It serves to alert pharmacists that excursions in patients' BG may be related to insulin pump BC differences. This may be especially important if a patient changes their insulin pump manufacturer, as this would necessitate closer monitoring by the pharmacist.

The content contained in this article is for informational purposes only. The content is not intended to be a substitute for professional advice. Reliance on any information provided in this article is solely at your own risk.

« Click here to return to Infusion Pharmacy Update.