Double Blinded Randomized Control Trial of Types of IVF in Children With DKA

Objectives: Intravenous (IV) fluid administration is a fundamental component of diabetic ketoacidosis (DKA) treatment. Normal saline (NS), the most common IV fluid used in DKA management, contains more chloride than human blood. Excessive amounts of chloride have been shown to cause a detrimental metabolic acidosis. Other IV fluids have more physiologic chloride levels, such as lactated ringers (LR). This study will compare the rates of hyperchloremic metabolic acidosis in children treated with NS to those treated with LR to determine the effect on overall length of acidosis and length of stay in the hospital or intensive care unit. Design: Single-center, double blinded, randomized controlled trial. Subjects: Children aged 0 to 18 years who present with diabetic ketoacidosis and require pediatric intensive care unit admission. Patients with evidence of shock, multi-organ failure or clinically significant cerebral edema will be excluded. The projected study population will be 104 patients, 52 in each arm. Interventions: Patients will be enrolled within 1 hour of presentation to the emergency room or pediatric intensive care unit if transferred directly from another facility. They will be randomized to receive intravenous fluids containing 0.9% saline or lactated ringers. All patients will be treated using the institutional DKA protocol with the content of the intravenous fluids being the only difference in treatment between arms. Study intervention lasts until the end of the acute management of DKA. Planned measurements and study outcomes: The primary study outcome will be duration of metabolic acidosis. Resolution of metabolic acidosis will be defined in three ways: 1. Normalization of the ketosis; 2. Normalization of the serum pH; 3. Normalization of the serum bicarbonate level. Secondary outcomes will include length of stay in the pediatric intensive care unit and length of stay in the hospital. All outcomes will be correlated with the overall chloride load given via intravenous fluids during DKA management. Regression modelling will control for any baseline differences between the groups in regards to severity of DKA, and if newly diagnosed or poorly controlled diabetes mellitus.