Key Learning Points

1. 1. Critical care management of traumatic brain injuries aims to avoid the development of secondary insults, which may impact on survival outcomes.

2. 2. Intracranial hypertension is associated with increased mortality in patients with traumatic brain injury.

3. 3. Interventions proposed to control intracranial hypertension should be made after balancing the associated risks and benefits.

4. 4. Avoidance of corticosteroids in traumatic brain injuries is the only Grade I recommendation included in the Brain Trauma Foundation Guidelines.

5. 5. Further research is needed to make clear recommendations regarding the role of decompressive craniectomy on functional outcomes.

For many years, the management of severe TBI has been based on information gained from intracranial pressure (ICP) monitoring. The rationale for its use is based on the three Ps of prognosis, perfusion, and pathology of TBI, with the hope that using ICP to guide therapy would prevent secondary brain injury and ultimately improve neurological outcome.However, one of the fundamental challenges in neurotrauma has been the inability to demonstrate that the fall in ICP achieved by these measures is subsequently translated into an improvement in clinical outcome. For years, patients with severe TBI were routinely hyperventilated, frequently placed in a barbiturate coma, or more recently rendered hypothermic, because these measures consistently reduce intracranial pressure. However, clinical studies have failed to show that lowering intracranial pressure by these techniques provides clinical benefit, and in some instances they may have caused harm. It is this regard that the use of decompressive craniectomy was thought to be promising, and there have now been two large multicentre randomised controlled trials investigating efficacy of the procedure. The results provide good evidence to guide practice but also raise ethical issues regarding the use of a procedure that reduces mortality but increases survival with severe disability.