THE IMPACT OF A NEW BEST PRACTICE ADVISORY ON THE MANAGEMENT OF DIABETIC KETOACIDOSIS

Abstract

SESSION TITLE: Critical Care Posters SESSION TYPE: Original Investigation Posters PRESENTED ON: October 18-21, 2020 PURPOSE: According to the Centers for Disease Control and Prevention, overall age-adjusted diabetic ketoacidosis (DKA) hospitalization rates have been steadily increasing from 2009 to 2014 at an average annual rate of 6.3%. Further, the direct and indirect annual cost of DKA hospitalizations is 2.4 billion US dollars. Due to the increasing incidence of DKA and the economic impact associated with its morbidity and treatment, effective management is key. Herein, we aimed to streamline the management of DKA in our intensive care units (ICU) by implementing a Best-Practice Advisory (BPA) that notifies care providers when DKA has resolved. METHODS: A BPA was implemented in our ICUs on 9/15/2018 to notify clinical providers when DKA resolves. DKA resolution was defined as a blood glucose (BG)≤250mg/dl, anion gap≤17 and bicarbonate concentration≥18 mEq/L. We conducted a retrospective review of patients admitted to the ICU with DKA a year before and after 9/15/2018. Adult (≥18 age) patients were included in the study if they met DKA criteria on admission and were treated with continuous insulin infusion (CII). Patients admitted prior to BPA implementation were in the pre-intervention group while patients admitted after BPA implementation were in the post-intervention group. Our primary outcome was total time on CII. Secondary outcomes included the incidence of hypoglycemia (<80mg/dl), hypokalemia (<3.5 mEq/L), mortality, length of stay (LOS) and ICU stay. Summary and univariate analyses were performed. RESULTS: A total of 282 patients were included in the study. There were 162 (57%) patients in the pre-intervention group and 120 (43%) patients in the post-intervention group; 143 patients (51%) were females. The mean (±SD) age was 44 (±17) years. There was no significant difference in baseline characteristics such as age, sex, race, BMI, HbA1c, initial BG, initial anion gap or initial bicarbonate concentration between both groups (p>0.05). The mean (±SD) total time on CII in hours was significantly lower in the post-intervention group {14.8 (±7.7) vs 17.5 (±14.3) p=0.041, 95% CI: 0.11-5.3}. Lastly, the incidence of hypoglycemia was lower in the post-intervention group {n=4 (3%) vs 17 (10%), p=0.024}. There was no significant difference in hypokalemia, mortality, LOS or ICU stay between both groups (p>0.05). CONCLUSIONS: The BPA introduced in our DKA management algorithm successfully reduced total time on insulin and the incidence of hypoglycemia. CLINICAL IMPLICATIONS: Incorporating clinical decision support tools, such as Best-Practice Advisory, in the management of Diabetic Ketoacidosis improves key clinical outcomes. DISCLOSURES: No relevant relationships by Ahmad Al-Shyoukh, source=Web Response No relevant relationships by Diala Alawneh, source=Web Response no disclosure on file for Majdi Hamarshi; No relevant relationships by Moustafa Younis, source=Web Response