First Author: Taylor Hodle, PharmD, BCCCP – MICU Clinical Pharmacy Specialist, Baystate Medical Center Co-Author: Anna Tran, PharmD – Clinical Pharmacist, Kaiser Permamente Co-Author: Thomas Presti, MD – Pulmonary Intensivist, Baystate Medical Center
Introduction: Diabetic ketoacidosis (DKA) is a hyperglycemic crisis that accounts for annual medical costs of ~2.4 billion USD and over 500,000 hospital days. Standardized DKA protocols have been shown to reduce mortality, shorten length of stay (LOS), and improve overall patient outcomes. Recent studies suggest that the use of a two-bag system in adults with DKA improves time to anion gap closure, lowers incidence of hypoglycemia and hypokalemia, and decreases resource utilization. This quality improvement (QI) initiative aimed to integrate the two-bag system into a modified protocol in an adult medical ICU and assess its impact on time to DKA resolution.
Methods: This single-center, retrospective pre-post study included ICU patients treated for DKA between 12/2023–05/2024 (pre) and 12/2024–05/2025 (post). Patients were excluded if they had combined DKA and HHS or received insulin for non-DKA diagnoses. The modified DKA protocol focused on a fixed insulin dose and nursing driven fluid titration. The primary outcome was time to anion gap closure. Secondary outcomes were ICU and hospital length of stay, and number of fluid changes per patient. Safety outcomes included hypoglycemic and hypokalemic events, and protocol adherence. A voluntary pre and post nursing survey was also given to evaluate ease of bedside administration.
Results: A total of 23 patients met inclusion criteria for the pre group and 9 patients for the post. The median time to gap closure was 10.7 vs 9 hours in the pre vs post group. Median ICU LOS was 3 days for both groups, and hospital LOS was 7 vs 12 days for the pre vs post group. Fluid was changed a median of 2 vs 0 times in the pre vs post group. Hypoglycemic events occurred in 26% vs 11% in the pre vs post group, while hypokalemic events occurred in 30% vs 33% in the pre vs post group. Protocol adherence increased to 54% in the post group from 17% in the pre. Nursing survey revealed the new protocol had clearer instructions, decreased IV fluid delays, and improved nursing usability.
Conclusions: A two-bag modified DKA protocol in a medical ICU slightly improved time to anion gap closure, did not impact LOS, and had less hypoglycemic events.
