600: RETROSPECTIVE REVIEW OF OUTCOMES IN PEDIATRIC DKA: EARLY VERSUS LATE LONG-ACTING INSULIN

Abstract

Introduction: In patients with diabetic ketoacidosis (DKA), the American Diabetes Association guidelines support administering long-acting insulin (LAI) at the time of conversion from IV insulin to subcutaneous rapid acting insulin (RAI). Challenges with LAI timing in children may complicate this transition, and LAI may be administered earlier in the DKA course at some institutions. Limited evidence currently exists assessing the value of this early LAI administration while the patients are still in DKA. The purpose of this study was to assess the benefit and safety of early LAI in pediatric DKA. Methods: This single-center retrospective cohort study included patients < 21 years old admitted to the PICU in DKA on an IV insulin drip. Patients meeting inclusion criteria were divided into two groups: early LAI (administered >4 hours prior to conversion to RAI) and late LAI (administered within 2 hours of RAI). Each group was subdivided into DKA severity (moderate vs severe) and new onset vs known diabetes. The primary outcome of the study was to compare the time to DKA resolution, and secondary outcomes evaluated the total IV insulin doses received, and incidence of adverse effects and complications: hypoglycemia, hypokalemia, cerebral edema, and recurrence of DKA. Results: A total of 372 encounters were included in the preliminary assessment (224 in early LAI and 148 in late LAI groups). Baseline characteristics were similar between both groups. The time to DKA resolution was significantly longer (13 vs. 9.6 hours, p< 0.0001) and total dose of IV insulin significantly higher (65.7 vs. 44 units, p=0.0018) in the early LAI group. Rates of mild hypoglycemia and hypokalemia were significantly higher in the early LAI groups (33.5% vs. 21.6%, p=0.0134 and 42.4% vs. 14.2%, p< 0.0001, respectively). The trends in longer duration of DKA and rates of these adverse effects for the early LAI were also seen in the subgroups of new onset patients and with both moderate vs severe DKA. No differences were observed with severe hypoglycemia or hypokalemia, rates of cerebral edema or recurrence of DKA. Conclusions: Use of early LAI in pediatric DKA demonstrated no reduction in the time to resolution of DKA or total IV insulin dose regardless of severity of presentation, and increased risk of mild hypoglycemia and hypokalemia.