B54-01 Fluid Administration Patterns in DKA Versus HHS in Adults: A Retrospective Review

Rationale Diabetic Ketoacidosis (DKA) and Hyperglycemic Hyperosmolar State (HHS) are acute, life-threatening complications of diabetes. Although fluid resuscitation is the cornerstone of management, current recommendations are primarily based on expert consensus rather than quantitative evidence. This retrospective study aimed to evaluate the volume of fluids administered within the first 24 hours in patients presenting with DKA and/or HHS whether adequate blood glucose reduction (≥100 mg/dL/hr) was achieved. Methods This Institutional Review Board (IRB)-exempt retrospective study was conducted at Newark Beth Israel Medical Center. Adult patients diagnosed with DKA or HHS between March 2023 and June 2025 were identified using electronic medical records. Of 103 encounters reviewed, 95 met inclusion criteria: DKA (n = 63), HHS (n = 19), and combined DKA/HHS (n = 13). DKA was defined by hyperglycemia, ketosis, and metabolic acidosis, while HHS was defined by severe hyperglycemia and hyperosmolality without significant ketosis, and combined DKA/HHS is both significant ketoacidosis and marked hyperosmolarity. The primary outcome was blood glucose normalization (<200 mg/dL at 24 hours). Length of stay was also assessed. Multivariable logistic regression adjusted for age, weight, diagnostic group, and the difference between expected and administered fluid volumes at 24 hours. Results Patients with HHS were older (mean 58.4 ± 17.5 years) than those with DKA (46.6 ± 15.2) and combined DKA/HHS (51.1 ± 10.4). Mean fluids administered at 24 hours were similar across groups (∼4.2 L). However, expected fluid requirements were substantially higher for HHS (15.7 ± 5.1 L) and combined DKA/HHS (12.9 ± 2.7 L) compared with DKA (7.8 ± 2.2 L), resulting in larger fluid deficits for HHS (11.6 ± 5.4 L) and DKA/HHS (8.7 ± 3.4 L) than for DKA (3.6 ± 2.9 L) (p < 0.001). Median blood glucose decreased significantly from admission to 6 hours and 24 hours (both p < 0.001), though the magnitude of reduction differed across groups (p = 0.003). Length of stay did not differ significantly (p = 0.48), and no variables were independently associated with glucose normalization on multivariable analysis. Conclusions Across all diagnostic groups, fluids administered within 24 hours was markedly lower than the calculated fluid deficit, with the largest discrepancies in HHS and combined DKA/HHS. Despite comparable fluid volumes, these groups showed slower glucose normalization, suggesting under-resuscitation may delay metabolic recovery. These findings highlight the need for evidence-based, severity-specific fluid resuscitation guidelines in hyperglycemic emergencies. This abstract is funded by: None