(1253) Evaluating the Implementation of an Electronic-Based DKA Protocol in the Emergency Department

Introduction: Glucommander™ (GM) is a computerized patient-specific glycemic management system for insulin titration. Patients presenting with diabetic ketoacidosis (DKA) to the emergency department (ED) at the study institution were historically initiated on a paper-based insulin infusion protocol (PBP) and transitioned to GM software upon admission. As of October 2021, patients presenting to the ED in DKA are initiated directly on GM. The purpose of this study is to compare the efficacy and safety of PBP to GM in the ED.

Methods: This is a single-center, retrospective study, comparing patients that received an insulin infusion from October 2018 through March 2021 (PBP group) and October 2021 to May 2022 (GM group) at a large, urban academic medical center. Included patients were 18 years or older diagnosed with DKA and initiated on a DKA insulin infusion protocol in the ED. Exclusion criteria were incarcerated or pregnant patients, left against medical advice (AMA), greater than a 12-hour gap between the end of the paper protocol and the start of GM, and duration on either protocol of less than one hour. The primary outcome is time to DKA resolution from ED arrival. Secondary endpoints include ICU admission, duration on an insulin infusion, incidence of hypoglycemia, and hospital length of stay (LOS).

Results: Total of 227 patients were included. Treatment with GM was associated with a faster average time to DKA resolution as compared to PBP (12.9 ± 9.4 vs. 20.6 ± 12.7 hours, p< 0.001), a decrease in the percentage of patients admitted to the ICU (10% vs. 34%, p< 0.001), and a shorter median duration on an insulin infusion (17.5 vs. 28.2, p< 0.001). The use of GM did not have a significant impact on the incidence of hypoglycemia, although lower in patients initiated directly on GM in the ED vs. the PBP group (0.05% vs. 0.1%, p=0.56), and overall median hospital LOS when compared to PBP (3.7 vs. 3.2 days, p=0.63).

Conclusions: Implementation of Glucommander™ software in the ED to manage DKA provides patient-specific insulin titrations from the start of DKA treatment, resulting in a faster time to DKA resolution, fewer ICU admissions, and shorter durations on insulin infusions.