ODP231 Outcomes of Weight-based vs. Constant-based Insulin Infusion for Treatment of Diabetic Ketoacidosis in Non-ICU Patients

Abstract

Background

The American Diabetes Association's guidelines for treatment of diabetic ketoacidosis (DKA) have explicit guidelines on the use of a weight-based insulin infusion to rapidly correct a patient's blood glucose and acidosis. Despite close monitoring, insulin infusion requires close monitoring and carries a risk of hypoglycemia, leading to adverse outcomes. While weight-based insulin infusions are designed to infuse a steady rate of insulin based on the patient's weight, constant-based insulin infusions are designed to infuse insulin based on a "constant" which represents the patient's sensitivity to insulin. Although constant-based insulin infusions are used at many institutions, there are no recommendations from any professional organization on when to use a constant-based insulin infusion. The goal of this quality improvement study was to evaluate our clinical use of weight-based and constant-based insulin infusions on the quality of care for patients admitted with diabetic ketoacidosis.

Methods

A retrospective review of non-ICU patients admitted with diabetic ketoacidosis over a one-year period was performed. Patients were divided into four groups for analysis based upon their insulin infusion protocol. Group 1 (n = 150) received weight-based insulin infusions throughout the study; Group 2 (n = 31) received constant-based insulin infusions throughout the study; Group 3 (n = 114) was started on weight-based insulin infusions and then switched to constant-based insulin infusions; and Group 4 (n = 6) was started on constant-based insulin infusions and then switched to weight-based insulin infusions. The primary outcome variables were hypoglycemic events and severe hypoglycemic events occurring during the initial infusion protocol and after a change from the initial infusion protocol to another protocol (if applicable). Hypoglycemia was defined as glucose levels between 41-70mg/dl and severe hypoglycemia as glucose level <=40 mg/dl.

Results

Both hypoglycemic events and severe hypoglycemic events during the initial insulin infusion were not significantly associated with the insulin infusion protocol (P = .391 and P = 1. 0, respectively). Moreover, hypoglycemic events were not significantly associated with a change in insulin infusion protocol; (P = .145) and no case of severe hypoglycemia was reported on infusion change. Type of diabetes (type I versus type II) was not significantly associated with the insulin infusion protocol groups (P = .784). Patients initially treated with a weight-based insulin infusion were found to have a statistically significant shorter duration of infusion (p<0. 001). This difference persisted on pairwise comparison with each group.

Conclusion

Our study does not provide convincing evidence that constant-based insulin infusions will improve treatment of our patients with diabetic ketoacidosis by reducing hypoglycemic events or reducing the duration of treatment when compared with the standard of care, a weight-based insulin infusion. In fact, weight-based insulin resulted in shorter duration of treatment, which may affect the length of stay.

Presentation: No date and time listed