2025 Southeastern Residency Conference

Title: Evaluation of DKA Resolution with a Revised DKA Protocol in a Community Hospital Setting


Authors: Nancy Henin; Vanessa Velazco; Tracey Bastian; Valerie A. Van Vickle


Background: Diabetic ketoacidosis (DKA) is a life-threatening complication of diabetes that has resulted in an increased hospitalization rate leading to higher healthcare costs. Early and appropriate management of DKA is important to prevent mortality and decrease the risk of complications, such as hypoglycemia, hypokalemia, and  cerebral edema. Implementation of a standardized DKA protocol, including fluid resuscitation, insulin therapy, and electrolyte repletion, is important to ensure optimal treatment outcomes. The objective of this study is to assess the time to DKA resolution after modification to the institution’s DKA protocol in a community hospital setting. 


Methods: This is a single-center, retrospective, Institutional Review Board (IRB)-approved study aimed to evaluate a revised DKA protocol in patients admitted to a 337-bed community hospital. Fluid therapy was adjusted to Lactated Ringers with special considerations for patients with congestive heart failure (CHF)/end-stage renal disease (ESRD). Insulin therapy was modified, including a distinct titration for type 1 vs type 2  diabetic patients. Phosphate supplementation criteria was also included in addition to potassium for electrolyte repletion. Utilizing the order set usage within the electronic health record (EHR), a retrospective chart review was performed to collect relevant data to determine the effectiveness of changes to the DKA protocol comparing data from 2018 (pre-initiation of protocol) and 2023 (post-initiation of protocol) in a 1:1 ratio. The primary outcome is total time on insulin infusion. Secondary outcomes include total time to meeting ADA-based definition of DKA resolution (BG<200 mg/dL and 2 of following: bicarbonate ≥15 mEq/L, venous pH >7.3, and AG≤12 mEq/L) from start of insulin infusion, ICU length of stay, proportion of patients that received appropriate fluid management per protocol, and correlation between beta-hydroxybutyrate (BHB) concentration and DKA  resolution. Secondary safety outcomes are incidence of hypokalemia, hypoglycemia (blood glucose <70 mg/dL  and <40 mg/dL), and rebound DKA. Descriptive statistics were used to analyze data collected for study outcomes. 


Results: A total of 100 patients were included in the study analysis (50 patients in 2018 pre-initiation group and 50 patients in 2023 post-initiation group). Baseline characteristics were comparable between groups, except for a lower percentage of Type 1 diabetic patients in post-initiation compared to pre-initiation group (40% vs 66%, respectively). The primary outcome of total time on insulin infusion was reduced post-initiation of the revised DKA protocol to approximately 18 hours compared to 23 hours pre-initiation of the protocol. Time to meeting ADA definition of DKA resolution from start of insulin infusion was 9.7 hours in post-initiation group and 10.5 hours in pre-initiation group. Total ICU length of stay was 1.8 days for post-initiation group and 2.2 days for pre-initiation group. The proportion of patients that received appropriate fluid management per protocol was 62% in post-initiation group compared to 90% in pre-initiation group. Mean BHB at discontinuation of insulin drip was 0.34 mmol/L in post-initiation group. For safety outcomes, the incidence of hypokalemia was higher in post-initiation vs pre-initiation group (52% vs 36%). Incidence of hypoglycemia (blood glucose <70 mg/dL) was lower in the post-initiation group (52% vs 56%), including the incidence of severe hypoglycemia (blood glucose <40 mg/dL) (6% vs 14%). Incidence of severe hyperglycemia and rebound DKA were comparable between both groups. 


Conclusion: Patients treated using the revised DKA protocol required less total time on insulin infusion and had a reduced ICU length of stay. Opportunities to further improve the protocol include identifying strategies to minimize the incidence of hypoglycemia and hypokalemia.