Randomized Controlled Trial of Intravenous Fluid In Severely Injured Paediatric Trauma Patients: Comparison of Normal Saline Versus Ringer's Lactate

Background: Trauma is the major killer of children and adolescents. The care of paediatric trauma patients is guided by the Advanced Trauma Life Support (ATLS) guidelines. Fluid resuscitation is a critical aspect of trauma resuscitation to avoid hypoperfusion and metabolic acidosis, and to maintain adequate oxygen delivery to tissues. ATLS recommends the use of either intravenous Normal Saline (NS) or Ringer's Lactate (RL) in trauma fluid resuscitation. There are currently no recommendations or any studies to suggest which of these two fluids are more appropriate for pediatric trauma patients and their ionic compositions do differ. As such, both solutions have potential benefits and potential complications. There are three primary issues that relate to choice of solution. Sodium concentration and risk of iatrogenic hyponatremia: NS has higher sodium content relative to RL (154 vs. 130 mmol/L, respectively; normal blood levels are 135-145 mmol/L). Maintenance of a normal or slightly elevated sodium level in paediatric trauma patients is often imperative, given that the vast majority of these patients have significant traumatic brain injury. Hyponatremia can worsen cerebral edema and increase intracranial pressure. In a previous study conducted by us at Children's Hospital, LHSC we reported that paediatric trauma patients in our centre received NS. Despite NS used as the resuscitation fluid, pediatric trauma patients all showed a trend towards low-normal levels of sodium in their blood. Given the lower content of sodium in RL, we question whether NS is superior for maintenance of blood sodium. Chloride concentration and risk of iatrogenic hyperchloremic metabolic acidosis: NS contains significantly more chloride than RL to maintain electro-neutrality (154 vs. 109 mmol, respectively; normal blood levels are 98-108 mmol/L). In contrast, the RL solution replaces 28 mmol/L of chloride with equimolar lactate. As the elevated chloride content in NS is postulated to instigate the undesired consequences of driving blood bicarbonate lower and producing a hyperchloremic metabolic acidosis,6, 7 we question whether NS is inferior to RL for maintenance of blood pH. Exacerbation of trauma-induced inflammation: Trauma results in a systemic inflammatory response (Fraser Lab, unpublished results). Exacerbation of the inflammatory state with non-optimal resuscitation practices might worsen overall outcome. Published data on non-trauma patients suggests that NS administration is pro-inflammatory relative to RL,9 but this possibility has not been studied in paediatric patients, or in any trauma patients. The inflammatory cascade can be assessed in blood by measuring the levels of global inflammatory markers [erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)] and specific pro-inflammatory mediators elevated by multisystem trauma [interleukin-6 (IL-6), interleukin-8 (IL-8), granulocyte colony stimulating factor (G-CSF) and monocyte chemotactic protein-1 (MCP-1)]. Given the higher inflammatory potential that has been associated with NS, we question whether NS is inferior to RL for minimizing inflammatory cascades. Hypothesis: We hypothesize that severely injured paediatric trauma patients resuscitated with NS will have optimal blood sodium levels compared to patients resuscitated with RL, but at the expense of hyperchloremic metabolic acidosis and exaggerated inflammation. Methodology: This study will be a prospective, single-blinded, RCT of severely injured paediatric trauma patients admitted to the Children's Hospital, London Health Sciences Centre. While the clinical staff treating the patient will be aware of treatment group, the patient, as well as the laboratory technicians determining the serum changes will be blinded to the fluid administered. Randomization: Patients will be immediately randomized to either receive NS or RL as their resuscitation and maintenance fluid for the first 24 hours. Randomization, via selection of sealed envelope based on a computer-generated list, will be done by the trauma resuscitation room intravenous access nurse without delay on arrival. Intravenous access and solution administration occurs as a priority within minutes of arrival to the trauma room. Clinical Data and Blood Analyses: All patients included in this study will be entered into the LHSC Trauma Database, comprised of over 400 data elements, by a single, trained Trauma Data Analyst. The database regularly undergoes quality monitoring to ensure the data is complete and is of highest quality. Descriptive analyses and epidemiologic profiles on demographic, clinical and injury data will be undertaken, as we have published previously. A baseline trauma blood panel will be drawn upon presentation to our resuscitation room (Time "0"), and then repeated at 24 hours. Blood measurements for analyses will include serum sodium, chloride, bicarbonate, base excess, osmolality, blood gas, ESR and CRP. An extra blood vial will be taken and stored for batch analyses of pro-inflammatory cytokines by multiplex assay (IL- 6, IL-8, G-CSF and MCP-1). Other interventions will be carried out as per standard of care at the discretion of the Trauma most responsible physician. Any medications or any additional fluids the patients receive will be recorded. The total amount of intravenous fluid received in the first 24 hours will be measured for all patients. Data Analysis: Descriptive analyses and epidemiologic profiles on demographic, clinical and injury data will be undertaken, as we have published previously. Treatment groups will be compared using Mann Whitney U test for the outcome measures at 24 hours, as they are continuous. This will include the primary outcome variable (serum sodium) and the secondary outcome variables (serum chloride, bicarbonate, pH and inflammatory markers: ESR, C-RP, IL-6, Il-8, G-CSF, MCP-1).