There are several validated insulin infusion protocols to help guide intravenous (IV) insulin dose titrations. However, when continued upward titration of insulin infusion is required to achieve glycemic control, little is known to guide how high the infusion rate can be increased without compromising safety and improving efficacy. The purpose of this study was to characterize patients requiring high-dose insulin infusions (rates >20 units/hour) for the management of hyperglycemia and determine the incidence at our institution. A retrospective chart review of all patients aged 18 years or older who received IV insulin infusion between 01/01/2016 and 07/31/2017 at Beaumont Hospital, MI resulted in 82 patients (2.9%) who received high-dose insulin infusion for the management of hyperglycemia. Average age was 62 years old and majority were Caucasian males. Most patients had hyperglycemia associated with risk factors such as: history of type 2 diabetes (71%), obesity (median body mass index (BMI) was 33.9 kg/m2), critical illness (89%), many were surgical patients (>50%), and a large proportion received vasopressors (79%) and corticosteroids (36%). The highest infusion rate observed was 80 units/hour, with a median infusion rate of 6.5 units/hour. Hypoglycemia (BG ≤ 70 mg/dl) was reported in 29% of patients and 11% had recurrent hypoglycemic episodes. Our findings suggest that the overall need for high-dose insulin infusion rates is very low. However, the incidence of hypoglycemia in these patients was substantially high and therefore concerning. Insulin is a high-alert medication requiring implementation of appropriate safety measures to prevent errors and careful investigation is warranted when used in high-doses. We recommend an insulin infusion “time-out” to assess the infusion site, the IV line and insulin bag before exceeding 20 units/hour. Furthermore, institutional policies for soft- and hard-maximum infusion parameters should be established to conserve use.
A.A. Feleke: None. R.C. Fuller: None. R. Ismail: None. M. Cadiz: None.