(CSEMP074) DIABETIC KETOACIDOSIS SUBCUTANEOUS INSULIN PROTOCOL: A QUALITY IMPROVEMENT PROJECT AT A CANADIAN TERTIARY CARE CENTER

Abstract:

Background: Diabetic ketoacidosis (DKA) is a life-threatening hyperglycemic emergency in patients with diabetes mellitus. The current standard of care in Canada is intravenous (IV) insulin infusion, which poses a significant cost to the healthcare system. There is growing interest in using subcutaneous (SC) insulin as an alternative. The objective of Phase 1 of this study was to implement and evaluate a SC insulin protocol for management of DKA.

Methods: A DKA SC insulin protocol was developed and piloted in the Emergency Department (ED) and Intensive Care Unit (ICU) at a tertiary care center. Phase 1 of this study involved retrospective chart review of patients with DKA treated with either SC or IV insulin to characterize safety and efficacy of the protocol. Patients admitted between March 2022-May 2023 were considered for inclusion if they were 18 or older, non-pregnant, and diagnosed with DKA. Patients who were clinically unstable, had a reduced level of consciousness, or had a weight greater than 160 kilograms were excluded. Treatment involved simultaneous administration of weight-based long-acting and short-acting subcutaneous insulin. Demographics and clinical parameters including time to anion gap (AG) closure, length of stay (LOS) and complications (hypokalemia, hypoglycemia, AG re-opening) were analysed with independent t test and Chi-squared test. Preliminary analyses were unadjusted.

Results: After excluding 27 patients, 41 patients were included for analysis; 10 were treated with SC insulin and 31 were treated with IV insulin. Preliminary analyses showed that the SC insulin group had significantly higher pH (7.30±0.05 versus 7.18±0.14, p=0.008) and bicarbonate (20.2±3.39 versus 13.6±5.99, p=0.002) at presentation compared to the IV insulin group. There was no significant difference in time to AG closure. Notably, patients treated with SC insulin had significantly lower rates of hypoglycemia compared to those treated with IV insulin (0% versus 39%, p=0.019) with no difference in rates of hypokalemia or AG re-opening. There was no significant difference in LOS in ED or hospital admission.

Conclusion: To our knowledge, this is the first implementation of a DKA SC insulin protocol at a Canadian tertiary care center. Interim data show that the use of SC insulin results in similar time to AG closure and LOS as compared to IV insulin without increased rates of hypoglycemia, hypokalemia, or AG re-opening. Next steps include formalizing implementation of the protocol and conducting additional analyses to further establish the safety and efficacy of SC insulin for treatment of DKA.