Early Versus Late Administration of Long-Acting Insulin in Adult Diabetic Ketoacidosis

Background: Evidence is inconclusive if early administration of subcutaneous (SQ) long-acting insulin (LAI) in management of diabetic ketoacidosis (DKA) improves outcomes. Objective: This study compares early versus late administration of SQ LAI in time to DKA resolution. Methods: This single-center, retrospective study included patients with DKA who received >= 12 hours of continuous intravenous insulin (CIVI) with LAI overlap. Patients were compared based on LAI administration time to CIVI initiation: Early (<12 hours) versus Late (>= 12 hours). The DKA resolution is defined as blood glucose < 200 mg/dL and 2 of the following: anion gap < 12 mEq/L, pH > 7.35, or serum carbon dioxide >15 mEq/L. Outcomes included time to DKA resolution, length of stay (LOS), CIVI duration, and adverse events. Results: A total of 27 patients were included in each group. Baseline characteristics were similar between both groups. There was no difference in time to DKA resolution, Early = 17.6 (13.9-26.8) hours versus Late = 19.2 (17.1-32.1) hours, P = 0.16. The Early group had shorter CIVI duration (Early = 19.5 +/- 10.3 hours vs Late = 25.6 +/- 8.4 hours, P = 0.02) and received less intravenous (IV) fluids in the first 36 hours (Early = 4.04 +/- 2.12 L vs Late = 5.85 +/- 2.24 L, P = 0.004). No differences were identified with adverse events, including hypoglycemia, or LOS. Conclusion and Relevance: Administration of SQ LAI < 12 hours did not decrease time to DKA resolution or LOS. Patients in the Early group had received a lower dose of LAI, shorter duration of CIVI infusion, and required less IV fluids within 36 hours of admission. This study supports the need for further research to determine the potential benefits of administering SQ insulin early in managing DKA.