BEST:TRIP: A Trial of Intracranial-Pressure Monitoring in Traumatic Brain Injury

Chestnut. NEJM 1012;367(26):2471-2481. doi:10.1056/NEJMoa1207363

Clinical Question

• In patients with severe traumatic brain injury does monitoring intracranial pressure improve mortality and improve neurological function?

Design

• Randomised controlled trial
• Computer generated randomisation
• Randomisation stratified according to study site, severity of injury and age
• Parallel-group
• Blinding of examiners assessing neuro-functional outcome
• Sample size calculation
  • 324 patients would give 80% power to detect an increase of 10% of patients with a good outcome or moderate disability according to the Extended Glasgow Outcome Scale (GOS-E)

Setting

• 6 sites in Bolivia and Ecuador
  • All had ICUs staffed with intensivists, 24 hour CT and neurosurgery coverage
• September 2008 – October 2011

Population

• Inclusion:
  • Age ≥13
  • On admission or within 48 hours of injury: GCS 3-8 (with a score on the motor component of 1-5 if the patient was intubated)
• Exclusion:
  • GCS 3 and bilateral fixed and dilated pupils
  • Injury thought to be unsurvivable

• Concerning intervention and control groups
  • Median time to arrival at 1st hospital was 1.0hr for direct admissions and 2.7hrs for transfers
  • Age (median) 29 vs. 29
  • Transferred from another hospital 62% vs. 61%
  • Marshall classification on initial CT head: Diffuse injury III/IV: 51% vs. 48%
  • Evacuated mass lesion 31% vs. 35%
• 324 patients randomised

Intervention

• Intracranial pressure monitoring
  • Intra-parenchymal monitor placed as soon as possible
  • Treated to maintain ICP <20mmHg according to standard guidelines
  • Monitored for median 3.6 days

Control

• Treatment based on clinical finding and CT scan results
  • Clinical/imaging signs of intracranial hypertension treated first with protocol specific doses of hyperosmolar therapy, optional mild hyperventilation + optional ventricular drainage
  • If continuing oedema, guideline based care with consideration of rescue treatments e.g. high dose barbiturates, decompressive craniectomy

Outcome

• Primary outcome: Composite of 21 components including survival time, duration of impaired consciousness and neuropsychological status, at 6 months – no significant difference
  • Comparing intervention to control groups
    • Median 56 vs. 53, p=0.49, (Proportional Odds Ratio 1.09, 95% C. I. 0.74-1.58)
• Secondary outcomes: comparing intervention to control groups
  • No significant difference in mortality or GOS-E Scale
    • 14 day mortality: 21% vs. 30% (Hazard ratio 1.36, 95% C.I. 0.87-2.11, P=0.18)
    • 6 month mortality 39% vs. 40% (Hazard ratio 1.1, 95% C.I. 0.77-1.57, P=0.6)
    • Favourable outcome on GOS-E Scale at 6 months 44% vs. 39%
  • Significant increase in Decubitus ulcers in ICP monitoring group
    • 12% vs. 5%, P=0.03

Authors’ Conclusions

• Management guided by ICP monitoring is not superior to management guided by neurological examination and serial CT imaging in patients with severe traumatic brain injury

Strengths

• Randomised
• Multi-centre
• Blinding of assessors of primary outcome

Weaknesses

• Composite primary outcome with mortality equally weighted with 20 other variables such as motor dexterity
• Not powered for mortality
• Differences in treatment strategies between intervention and control groups e.g. in intervention group drainage of CSF was 1st option for raised ICP, where as this was not an option in control group
• Study performed in Bolivia and Ecuador. There are significant differences between the health care provided, particularly pre-hospital resuscitation, between Bolivia/Ecuador and Western European countries. Therefore these results may not be valid in other settings.