Early dextrose and lower-chloride fluids in pediatric diabetic ketoacidosis: a faster and safer approach? - A multicenter cohort study

This study aimed to compare different intravenous fluid protocols used in the management of pediatric diabetic ketoacidosis (DKA) and to evaluate their effects on treatment duration, biochemical trends, and complications. This retrospective multicenter study included 139 children treated for DKA between June 2021 and August 2023 in five tertiary centers. Treatment protocols reflected the routine DKA management practices of the participating centers. Demographic data, biochemical parameters, and treatment-related data were collected retrospectively, and linear regression analysis was performed to identify independent predictors of treatment duration. Three different maintenance and deficit fluid protocols were identified across participating centers. Following the initial saline bolus, the sodium and dextrose compositions of the maintenance fluids differed among protocols. Protocol 1 consisted of 0.6% NaCl with 3.3% dextrose throughout treatment. Protocol 2 used 5% dextrose–0.9% NaCl for the first 6 h, followed by 0.6% NaCl–5% dextrose. Protocol 3 consisted of 5% dextrose–0.9% NaCl throughout treatment. Among newly diagnosed children with severe acidosis, Protocol 1, which provided the lowest chloride load (~ 100 mEq/L), was associated with a shorter treatment duration and lower peak chloride levels compared with Protocols 2 and 3. In multivariable analysis, lower initial pH, higher peak chloride, and new-onset diabetes independently predicted longer treatment duration. No cases of cerebral edema were observed in any treatment group. A lower-chloride maintenance-plus-deficit fluid regimen was associated with more rapid acidosis resolution and a lower incidence of treatment related hyperchloremia in severe DKA. Notably, dextrose was included in the initial maintenance and deficit fluids in all protocols, irrespective of serum glucose concentrations. Early dextrose administration may have minimized rapid osmotic shifts, potentially contributing to the absence of cerebral edema in all treatment groups. Not applicable.