Intracranial pressure monitoring in severe traumatic brain injuryBMJ 2013; 346 doi: https://doi.org/10.1136/bmj.f1000 (Published 15 February 2013) Cite this as: BMJ 2013;346:f1000
- Peter J Hutchinson, reader in neurosurgery1,
- Angelos G Kolias, National Institute for Health Research academic clinical fellow in neurosurgery1,
- Marek Czosnyka, reader in brain physics1,
- Peter J Kirkpatrick, consultant neurosurgeon1,
- John D Pickard, professor of neurosurgery1,
- David K Menon, professor of anaesthesia2
- 1Division of Neurosurgery, Addenbrooke’s Hospital and University of Cambridge, Cambridge CB2 0QQ, UK
- 2Division of Anaesthesia, Addenbrooke’s Hospital and University of Cambridge, Cambridge, UK
In a trial recently published in the New England Journal of Medicine, Chesnut and colleagues attempted to provide class I evidence on the impact of intracranial pressure (ICP) monitoring on functional and neuropsychological outcomes after traumatic brain injury (TBI).1 The authors concluded that there was no difference in the primary outcome—a composite of 21 equally weighted components—between the group of patients who had ICP monitoring and the group that did not. This is a landmark study; undertaking such a trial has long been considered impossible because most experts considered ICP monitoring the primary basis for managing patients with severe TBI.2 However, the findings require some scrutiny before we can consider a fundamental change in our approach to managing these patients.
Since it was introduced into clinical practice more than 50 years ago, ICP monitoring has gradually become the standard of care in most centres that treat patients with severe TBI in the United Kingdom, North America, and most developed countries.2 The physiological basis of ICP monitoring in TBI is twofold. Firstly, increasing ICP indicates escalating mass effect (from haematomas, contusions, or diffuse brain swelling). If escalating mass effect is left untreated, brain herniation and death will follow. Secondly, ICP has a direct impact on cerebral perfusion pressure (the mean arterial blood pressure minus ICP). It is important to maintain cerebral perfusion pressure to avoid brain ischaemia—one of the major factors that contribute to unfavourable clinical outcome after TBI.2 ICP monitoring is used to guide the use of treatments for severe TBI, such as hyperventilation, osmotherapy, hypothermia, barbiturate coma, and decompressive craniectomy.
Numerous large cohort studies have shown that raised ICP (around 20-25 mm Hg) is independently associated with a higher risk of death after TBI.3 4 5 However, a study published in 2012, a secondary analysis of data on 365 patients with severe TBI from a randomised trial, found no independent association between average ICP and neuropsychological functioning among survivors.6 The only other study to question the usefulness of ICP monitoring was a retrospective cohort comparison study from the Netherlands, which showed that patients who received ICP monitoring were treated in the intensive care unit for longer than those whose ICP was not monitored, and outcomes were no better in the monitored group.7 Nonetheless, because of abundant class II and III evidence, the Brain Trauma Foundation 2007 guidelines included a level II recommendation (moderate degree of clinical certainty) that ICP should be monitored in all salvageable patients with severe TBI.2 In the UK, head injury guidelines from the National Institute for Health and Clinical Excellence state that treatment in a neuroscience centre would benefit all patients with severe TBI, irrespective of the need for neurosurgical intervention.8 Moreover, a large cohort study has shown that management of severe TBI in neuroscience centres is associated with reduced mortality.9
With such widespread acceptance of ICP monitoring as the standard of care for patients with severe TBI, it would be difficult to recruit patients to a trial where one arm did not receive ICP monitoring. Chesnut and colleagues overcame this problem by identifying a group of intensivists in Bolivia and Ecuador who were unsure about its effectiveness and routinely managed their patients with severe TBI without ICP monitoring.1
The trial hypothesis was that a therapeutic protocol based on ICP monitoring would result in reduced mortality and improved neuropsychological and functional recovery compared with a therapeutic protocol based on imaging and clinical examination (control arm). Importantly, both arms received interventions aimed at lowering ICP, and it seems that significantly more patients in the control arm received osmotherapy and hyperventilation.1 Furthermore, only 45% of participants were transported to the first hospital by ambulance. This should not affect the internal validity of the trial because baseline characteristics were similar in the two arms, analysis was intention to treat, and the follow-up was 92% in both arms. However, external validity is certainly limited because the prehospital management of severe TBI is more advanced in the UK, North America, and most developed countries, where most patients with severe TBI are transported to hospital by ambulance.10
Chesnut and colleagues found no significant difference between groups in the primary outcome, which was a composite of 21 equally weighted components. Because 12 of the 21 items are neuropsychological tests, neuropsychological performance is highly influential in the composite endpoint.1 This is of concern if considered in light of existing literature.2 6 A more conventional outcome measure, the extended Glasgow outcome scale, showed a non-significant 5% difference in both mortality and favourable outcome (favouring the ICP arm).1 Moreover, as the authors acknowledge, the risk of a type II error was high: with 324 cases, the study had only 40% power to detect a 10% increase in favourable outcome on the Glasgow outcome scale.
Although the study investigators should be congratulated for recruiting patients to reach the intended target, the results must be interpreted with extreme caution because of the high risk of a type II error. In our opinion, a move away from ICP monitoring in developed countries would be detrimental to the outcomes of patients with severe TBI. We also believe that a “normal” ICP should not be considered only in light of a particular cut-off value, because waveform analysis of the ICP is also important. Ongoing research has shown that ICP waveform analysis can provide information on the state of cerebrovascular reactivity (PRx index) and can be used to estimate optimal cerebral perfusion pressure levels for individual patients.11 12 Finally, with increasing recognition of the heterogeneity of TBI, further integration of multimodality signals (ICP, brain microdialysis, brain tissue oxygenation, electrocorticography) could enable clinicians to deliver individualised treatments to patients with severe TBI.
Cite this as: BMJ 2013;346:f1000
Contributors: PJH, AGK, and DKM wrote the first draft. MC, PJK, and JDP revised it. The final manuscript was approved by all authors.
Competing interests: All authors have completed the ICMJE uniform disclosure at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no financial support was received for the submitted work; PJH, MC, and JDP are directors of Technicam (manufacturer of cranial access device for neuromonitoring); PJH chairs the British Neurotrauma Group (special interest group of the Society of British Neurological Surgeons) and is a vice president of the European Association of Neurosurgical Societies; AGK is a member of the academic committee of the Society of British Neurological Surgeons; MC is coauthor of brain monitoring software ICM+ (www.neurosurg.cam.ac.uk/icmplus) and has a financial interest in a part of the licensing fee through Cambridge Enterprise; DKM is a paid consultant or member of data monitoring committee for Solvay, GlaxoSmithKline, Brainscope, Ornim Medical, Shire Medical, and Neurovive; DKM co-chairs the European Brain Injury Consortium.
Provenance and peer review: Not commissioned; externally peer reviewed.