18: Effect of Fluid Overload on Severe Traumatic Brain Injury Outcomes: A Secondary Analysis of Adapt

Introduction: Previous studies have shown that fluid overload is associated with adverse outcomes in adults with severe traumatic brain injury (sTBI), but pediatric data is limited. This study aimed to evaluate pediatric patients enrolled in the Approaches and Decisions in Acute Pediatric TBI (ADAPT) trial data to determine if cumulative fluid balance (FB) during the first week of ICU care was associated with 6-month neurologic outcomes and new medical technologies. Methods: We conducted a secondary analysis of the ADAPT trial, a prospective multicenter cohort study of 1,000 children < 18 years with sTBI, defined by post-resuscitation Glasgow Coma Scale (GCS) ≤ 8, with intracranial pressure (ICP) monitoring. Patients who died in hospital were excluded. The primary exposure was averaged 7-day cumulative percent FB (7-day FB), categorized into quartiles. Primary outcome was good neurologic function at 6 months, defined as Glasgow Outcome Scale-Extended Pediatric (GOS-E Peds) score 1-3. Secondary outcome was new medical technologies (e.g., tracheostomy, gastrostomy). Logistic regression modeling adjusted for PRISM III, maximum Abbreviated Injury Scale, initial GCS, and abusive head trauma. Results: Of 645 patients with outcome data, higher FB was associated with younger age, lower body weight, higher PRISM III scores, and abusive head trauma. In univariate analysis, higher FB was associated with worse outcomes and new technology use. In adjusted models, FB quartile 3 had lower odds of good outcome than quartile 2 (OR 0.59, 95% CI 0.36-0.95, p = 0.031), with even lower odds in quartile 4 (OR 0.44, 95% CI 0.27-0.72, p = 0.001); AUC 0.70. The odds of requiring new medical technologies also increased with FB (quartile 3 vs 2: OR 2.05, 95% CI 1.13-3.74, p = 0.019; quartile 4 vs 2: OR 4.42, 95% CI 2.39-8.20, p < 0.001); AUC 0.68. Conclusions: Higher 7-day FB was independently associated with worse 6-month neurologic outcomes and greater odds of medical technologies in pediatric sTBI. Residual confounding, including fluid administered for elevated ICP, patient characteristics, illness severity, and intentional trauma, limit causal inference. Future work needs to determine whether FB is a modifiable target to improve outcomes.