221: OPTIMIZING A DKA PROTOCOL: BALANCED CRYSTALLOIDS VERSUS NORMAL SALINE

Introduction: Diabetic ketoacidosis (DKA) is a common complication of type 1 diabetes mellitus. Traditionally, normal saline (NS) has been considered the preferred intravenous fluid. However, NS is known to cause a hyperchloremic acidosis, which may prolong DKA resolution compared to lactated ringer’s (LR). Emerging data suggests faster resolution of DKA with the use of LR as compared to NS. Methods: The purpose of this quality improvement project was to improve the time to DKA resolution in patients admitted to the intensive care unit (ICU) and decrease time to insulin drip discontinuation utilizing LR as compared to NS. The primary outcomes measured were time to DKA resolution and insulin drip discontinuation. Secondary end points included length of stay, mortality, AKI on discharge, and new hyperkalemia or hypokalemia. The project was conducted from June 1st, 2021 to February 10th, 2022 and included adults ≥ 18 who presented to the emergency department with DKA, which was strictly defined as: a plasma glucose greater than 250 mg/dL, serum bicarbonate ≤ 18 mEq/L, an anion gap ≥ to 12 mEq/L, and the presence of ketones on urinalysis. During the first 90 days, the NS-based protocol was utilized. Subsequently, the DKA protocol was modified to use LR for 90 days. Data was collected from the electronic medical record and reviewed. Results: 124 patients met inclusion criteria with 71 in the NS group and 53 in the LR group. Time to DKA resolution was significantly longer in the NS group (41.76 hrs vs 29.398 hrs, p = .017). Values ≥ 200 hrs were excluded. There was no difference in the time to insulin drip discontinuation (42.549 hrs vs 44.765 hrs, p value= .800). There were also no significant differences in length of stay (5.39 days vs 4.83 days, p value= .567), AKI on discharge (11.3% vs 13.2%, p value= .743), new hyperkalemia (4.2% vs 1.9%, p value= .635), or mortality (1.4% vs 3.8%, p value =.575). New hypokalemia occurred in 18.9% of the LR group compared to 5.6% of the NS group (p value= .021). Conclusions: The results demonstrate that changing DKA protocol to use LR instead of NS decreased DKA resolution time by approximately 12 hours. This suggests an opportunity for decreasing ICU length of stay and improving patient-centered outcomes for patients with DKA.