# Characterization and Prognostic Factors of Severe Pediatric Traumatic Brain Injury

## TL;DR

This study examines severe traumatic brain injuries in children, identifying factors like GCS score and lactate levels that predict mortality in intensive care.

## Contribution

The study identifies early clinical and laboratory variables that may guide treatment and predict outcomes in severe pediatric TBI.

## Key findings

- Pediatric Index of Mortality 3 (PIM3) was the only variable independently associated with mortality in multivariable analysis.
- Early variables like GCS score, lactate, and fibrinogen showed strong associations with mortality in univariable analysis.
- Arterial hypotension and intracranial hypertension were also linked to higher mortality rates.

## Abstract

Introduction: Trauma is the leading cause of death and disability among children and young adults. In particular, traumatic brain injury (TBI) stands out as the main cause of mortality and morbidity associated with pediatric trauma. It has a high likelihood of neurological, structural, functional, and behavioral sequelae among survivors, remaining a significant public health issue with notable impact on society. This study aimed to characterize patients with the diagnosis of severe TBI in a Pediatric Intensive Care Unit (PICU) over the past 15 years and to explore prognostic factors associated with mortality at PICU discharge.

Materials and methods: This was an exploratory observational study using a retrospective data collection method, conducted in the PICU of a Portuguese tertiary-level pediatric hospital. All children admitted with severe TBI between January 2010 and December 2024 were included. Demographic, clinical, and laboratory data were collected. Mortality at PICU discharge was the primary outcome analysed.

Results: Ninety-seven children with severe TBI were included (1.6% of all admissions). The main trauma mechanisms were pedestrian accidents and motor vehicle collisions. Overall mortality was 21.6%. In exploratory multivariable analysis, Pediatric Index of Mortality 3 (PIM3) was the only variable significantly associated with mortality. In univariable analysis, arterial hypotension and intracranial hypertension were associated with mortality. In parallel, early clinical and laboratory variables demonstrated relevant discriminatory ability in ROC analyses, and data-derived thresholds associated with mortality were identified for initial Glasgow Coma Scale (GCS) (≤4), PIM3 (>64.3), lactate (>4 mmol/L), and fibrinogen (<158 mg/dL).

Discussion and conclusion: This study characterized a significant cohort of children with severe TBI and highlighted the potential prognostic relevance of early clinical and laboratory variables, including GCS, PIM3, lactate, and fibrinogen. While PIM3 remained the only variable independently associated with mortality in exploratory multivariable analysis, GCS, lactate, and fibrinogen demonstrated clinically and biologically relevant associations with mortality in early assessment. Early identification of these factors may guide clinical decisions and therapeutic interventions, as well as have prognostic value, if confirmed in large-scale multicenter studies.