The Dept of Anaesthesia & Intensive Care, CUHK thanks

for an unrestricted education grant
BASIC instructor/provider course, Hong Kong, July 2nd-4th
Other upcoming courses
Home Feedback Contents

Head injury

Up Trauma basics Basic trauma crit care Traumatic shock Trauma in elderly Abdomen & pelvis Chest injuries Fat embolism Head injury Larynx & trachea Maxillofacial Pregnancy Spinal injury

ICP complication

Severe blunt head injury

Immediate management

Indications for immediate neurosurgical consultation

Indications for CT brain with 2-4 h (in addition to above)

Pathophysiology of secondary brain injury


ICP monitoring

Treatment of raised intracranial pressure


The management of severe blunt head injury revolves around minimizing further neurological injury due to secondary injury

Immediate management

3 main priorities:

  • resuscitation. Avoid hypotension and hypoxia - both significantly increase mortality and morbidity
  • rapid diagnosis of a brain lesion from head injury
  • measures to inhibit secondary brain insults and to prevent rise in ICP


  • Airway
  • Breathing
    • Apnoea or cyanosis in the field or PaO2 < 60 mm Hg associated with poor outcome
    • Intubation and ventilation to prevent aspiration and ensure adequate oxygenation. PEEP of up to 12 cmH2O well tolerated in terms of rise in ICP
      • consider pre-treatment with lignocaine (1-2 mg/kg 4 mins prior to intubation) to blunt rise in ICP in response to laryngoscopy
      • consider defasciculating dose of non-depolarizing muscle relaxant to prevent suxamethonium induced rise in ICP
    • Sedation and neuromuscular block may be necessary to prevent coughing, which may result in prolonged rises in ICP. Use short acting agents.
  • Circulation
    • NB shock is rarely due to isolated head injury except in young children and in patients with medullary injuries or large scalp lacerations
    • Hypotension is an independent predictor of poor outcome and the only one of the five major predictors which can be altered
    • Aim for MAP > 90 mm Hg in an attempt to maintain adequate cerebral perfusion pressure
    • Fluid resuscitation to re-establish normal cerebral blood flow as soon as possible -the most likely cause of hypotension in a trauma patient is hypovolaemia
      • Although it is not currently standard practice there are data supporting use of small volume resuscitation with hypertonic or hyperoncotic solutions. Small volume resuscitation with hypertonic/hyperoncotic solutions is associated with a faster restoration of normal cardiac output in severe haemorrhagic shock than any other form of fluid resuscitation. Use of hypertonic saline instead of isotonic fluid has been demonstrated to reduce ICP in patients with traumatic brain injury and intracranial hypertension. Retrospective subgroup analysis of a RCT of patients with severe traumatic brain injury demonstrated a higher systolic blood pressure and improved survival in patients resuscitated with hypertonic saline and a meta-analysis has shown that patients who receive hypertonic saline/dextran are about twice as likely to survive as patients who receive standard therapy. However an, as yet unpublished, large RCT shows no benefit from use of hypertonic saline.
      • patients resuscitated with albumin have a worse outcome thoan those resuscitated with saline. (Click here to view paper)
  • Note that blood-brain barrier does not function like other membranes: tonicity and not osmolality is important in determining fluid shifts.
  • Aim for mild dehydration
  • Specific treatment aimed at intracranial hypertension only if signs of transtentorial herniation or progressive neurological deterioration not attributable to extracranial causes are present.
    • Hyperventilation
    • Mannitol after adequate volume resuscitation
  • Secondary survey including:
    • GCS: initial score of 3-8 indicative of severe head injury
    • Pupil size
    • Motor function

Rapid diagnosis of brain injury

Indications for immediate consultation with neurosurgeon (usually prior to urgent CT)

  • Coma persisting after resuscitation
  • Deteriorating consciousness or progressive neurological signs
  • Skull fracture with any of:
    • Confusion
    • Fits
    • Neurological symptoms or signs
  • Open injury: depressed compound fracture of skull vault, fracture of base of skull, penetrating injury

Indications for CT within 2-4 h of admission (in addition to above)

  • age>60 yrs
  • failure to reach GCS of 15 within 2 h
  • skull fracture
  • fit
  • headache
  • vomiting
  • focal neurological signs
  • unstable systemic state precluding transfer to neurosurgery
  • uncertain diagnosis
  • tense fontanelle or suture diastasis in a child

Refer all patients with an abnormal CT scan or unsatisfactory clinical progress despite normal scan to neurosurgeon.

CT of subdural haematoma

Subdural haematoma: gross pathology

CT of extradural haematoma

EDH: gross pathology

Pathophysiology of secondary brain damage

Extracranial causes

  • Failure of adequate cerebral perfusion from circulatory shock
  • Failure of cerebral oxygenation from disturbance of pulmonary ventilation from:
    • aspiration
    • instability of chest wall (eg flail chest)
    • neurogenic disturbances of ventilatory drive

Intracranial mechanisms

  • rise in ICP due to mass lesions and cerebral oedema leading to fall in cerebral perfusion pressure and to herniation of brain is a major factor in poor outcome
  • cerebral blood flow of great significance in pathophysiology of secondary brain damage. Almost all patients dying from severe head injury have pathomorphological evidence of cerebral ischaemia. CBF heterogenous with areas of hyperperfusion and hypoperfusion. Hyperperfusion of damaged brain may increase cerebral oedema and thus ICP
  • cerebral autoregulation impaired

Prevention of secondary injury

ICP monitoring and management

Intracranial hypertension is important for two reasons:

  • in patients in whom ICP exceeds central venous pressure, cerebral perfusion pressure is determined by the difference between mean arterial pressure and intracranial pressure

  • intracranial hypertension may lead to coning

Indications for ICP monitoring:

  • severe head injury with abnormal admission CT (ie haematoma, contusion, oedema, or compressed basal cisterns)
  • severe head injury and normal CT but 2 of: age > 40 yrs, unilateral or bilateral motor posturing, systolic BP < 90 mm Hg
  • not routinely indicated in patients with mild/moderate head injury however may be appropriate in certain conscious patients with traumatic mass lesions


  • ventricular catheter attached to external strain gauge is most accurate, low cost method and allows drainage of CSF. ICP transduction via fibreoptic or strain gauge devices placed in ventricular catheters provide similar benefits but at higher cost
  • parenchymal ICP monitoring with fibreoptic or strain gauge catheter tip transduction is similar to ventricular ICP monitoring but has the potential for measurement drift
  • subarachnoid, subdural and epidural


  • include infection, haemorrhage(1.5%), malfunction (~10%), obstruction or malposition
  • rarely produce long term morbidity in patients
  • most studies define infection as positive CSF culture in ventricular and subarachnoid bolt devices however this would better be defined as colonization as there have been no reports in large prospective series of clinically significant intracranial infections associated with ICP monitoring devices

Treatment of raised ICP: (hierachical)

  • evacuate extradural/subdural haematomas +/- intracerebral haematomas, drain hydrocephalus
  • analgesia and sedation, nurse head up (30o), position head and neck to ensure venous drainage is not obstructed, control temperature (and treat cause of pyrexia) and fits to minimize rise in CMRO2
  • ventricular drainage
  • mannitol (0.3-1 g/kg initially followed by 0.25-0.5 g/kg 6 hourly). Alternative: hypertonic saline (2ml/kg of 7.5% saline)
  • hyperventilation to PaCO2 30-35 mm Hg
  • second line therapy eg high dose barbiturate therapy, hyperventilation to PaCO2 <30 mm Hg (monitoring of SjO2, AVdO2 and/or CBF recommended
  • ICP treatment should be initiated at an upper threshold of 20-25 mm Hg. Impact of ICP on outcome appears to lie in its role in determining CPP and as an indicator of mass effect. As CPP can be managed by manipulation of MAP to a great extent risk of herniation is a more important consideration in determining the treatment threshold

Cerebral perfusion pressure

  • aim for a CPP (ie MAP-ICP) of >60 mmHg by maintaining an adequate MAP (use norepinephrine if necessary) and control of ICP (NB the choice of this threshold is based on class III evidence)


  • prophylactic hyperventilation to PaCO2 £ 35 mm Hg during first 24h associated with a worse prognosis probably because it reduces already low CBF
  • hyperventilation may be necessary for brief periods when there is acute neurological deterioration or for longer periods if there is intracranial hypertension refractory to sedation, paralysis, CSF drainage and osmotic diuretics
  • SjO2, AVdO2 and CBF monitoring may help to identify cerebral ischaemia if hyperventilation (with PaCO2<30 mm Hg) necessary


  • Probably has 2 distinct effects in the brain:
    • Immediate plasma expanding effect. Reduces hct and blood viscosity. Increases cerebral blood flow and oxygen delivery causing autoregulatory vasoconstriction. Probable explanation for reduction in ICP within a few mins and why reduction is most marked in those with low CPP
    • Osmotic dehydration of brain. Delayed for 15-30 mins and persists for 90 mins to 6h
  • significant risk of renal failure if mannitol given in large doses, particularly if serum osmolality >320 mOsm. May be increased risk in patient receiving other potentially nephrotoxic drugs, in septic patients or patients with pre-existing renal disease. At levels >350 serious cellular damage may occur.
  • should be discontinued if does not cause a diuresis
  • euvolaemia should be maintained by adequate fluid replacement
  • may cause cardiovascular collapse in hypovolaemic patient. Contra-indicated in the unresuscitated patient
  • in common with other osmotics "opens" BBB with the result that mannitol and other small molecules in circulation may pass into brain. This effect becomes harmful after many doses because mannitol may accumulate in brain causing a reverse osmotic shift. Accumulation most marked when mannitol is in circulation for long periods (eg with continuous infusion) and therefore should be given as repeated boluses (0.25-1 g/kg)
  • frusemide may produce less changes in electrolytes and osmolality than mannitol. Reduction in cerebral oedema may be due to factors other than just diuresis. ? acts synergistically with mannitol. Little data to support use of frusemide in addition to mannitol


  • decrease ICP
  • high dose barbiturate therapy may be considered in haemodynamically stable salvageable patients with intracranial hypertension refractory to maximal medical and surgical ICP lowering therapy.
  • titrate dose to achieve EEG burst suppression
  • >85% of patients with intractable intracranial hypertension requiring barbiturates die

Other treatment

  • prevent hyperglycaemia: exacerbates ischaemic cerebral damage
  • attention to electrolyte status. These patients are prone to electrolyte abnormalities due to osmotic diuresis, cerebral salt losing states, SIADH and diabetes insipidus
    • initial IV fluid: normal saline
  • prospective specific treatment:
    • prevention of secondary brain damage by antagonizing neurotoxic mediators. Requires identification of relevant mediators. So far these include glutamate, kallikrein-kinin system and arachidonic acid. Other possible mediators include platelet activating factor and oxygen derived free radicals
    • steroids do not reduce ICP, do not improve outcome and are not recommended.
      • large RCT currently in progress
    • moderate hypothermia (32-33° ) may reduce ICP but contrary to previous small studies, recent data suggests that it does not improve outcome
  • fits. Recommended that anti-convulsants may be used to prevent early fits in patients at high risk of fits. Factors associated with increased risk:
    • GCS <10
    • cortical contusion
    • depressed skull #
    • subdural haematoma
    • epidural haematoma
    • intracerebral haematoma
    • penetrating head wound
    • seizure within 24h of injury

    Prevention of fits: phenytoin and carbamazepine decrease early (£ 7 days) post-traumatic fits but neither of these drugs nor phenobarbitone decrease the incidence of late post-traumatic fits. Available evidence does not indicate that prevention of early fits improves outcome. Valproate does not prevent post-traumatic fits

  • stress ulcer prophylaxis
  • physiotherapy
  • nutrition
    • metabolic rate ~140% of resting metabolic rate
    • increase largely abolished by neuromuscular blockade or barbiturate coma
    • negative nitrogen balance for approximately first 3 weeks after injury
    • early (<7 days) establishment of full feeding associated with lower mortality
    • same neurological outcome with parenteral or enteral feeding but fewer infectious complications with enteral feeding
    • aim for:
      • early establishment of enteral feeding. May be facilitated by jejunal feeding
      • calories = 140% of resting requirement. In paralysed patients or patients in barbiturate coma give resting requirement
      • 15% of calories as protein
  • look for and treat coagulopathy
    • brain injury predisposes to coagulopathy due to high levels of thromboplastin in brain
    • DIC may develop within 1 h
    • in patients with mechanical heart valves who are receiving warfarin probably safe to stop warfarin, reverse its effects and then re-start 1-2 weeks later
  • therapeutic hypothermia
    • currently available data do not support routine use of hypothermia for severe traumatic brain injury


  • Main determinants:
    • age
      • patients >60 yrs have a worse prognosis.~75% mortality in severely brain injured.
    • GCS on admission, especially motor function
      • 20% of patients with an initial GCS=3 survive and 8-10% will be moderately disabled or better
    • pupillary response
      • bilaterally unresponsive pupils associated with poor outcome (90% mortality in one study, ~75% dead, vegetative or severely disabled in two smaller studies)
      • interaction between pupillary response and pathology. Unresponsive pupils (prior to surgery) due to extradural haematoma: 56%; subdural haematoma: 88%
    • hypotension (systolic BP <90)
      • hypotension consistently one of the strongest predictors of poor outcome and the only one of the five major predictors that is amenable to treatment
      • evidence that correcting hypotension improves outcome
      • iatrogenic hypotension also associated with worse outcome. As a result great care should be taken to avoid intra-operative hypotension or procedures should be delayed until patient is more stable
    • CT appearance (initial CT). Features associated with worse outcome
      • abnormal CT associated with better prognosis
      • acute subdural haematoma worse than diffuse injury which is worse than extradural
      • higher abbreviated injury score CT component
      • absent or compressed basal cisterns
      • traumatic subarachnoid haemorrhage
      • midline shift (probably less important than other parameters)
  • presence of hypoxia or ischaemia
  • ICP > 20 mmHg
  • Coma persisting > 6 hrs following head injury associated with a 40% 6 month mortality
  • Overall mortality in severe head injury: approx. 30-50%
  • Recovery may continue for up to 18-24 months after head injury although the most significant gains are made in the first 6 months

Further reading

Brain Trauma Foundation & American Association of Neurosurgeons. Management and prognosis of severe traumatic brain injury. Published 2000

Update to BTF guidelines (2003)

Guidelines for children

Vincent JL, Berré J. Primer on medical management of severe brain injury. Critical Care Medicine, 2005; 33:1392-99

İCharles Gomersall, April, 2014 unless otherwise stated. The author, editor and The Chinese University of Hong Kong take no responsibility for any adverse event resulting from the use of this webpage.
Copyright policy    Contributors