Following trauma, early surgical evacuation of extradural and subdural haemorrhage is well established.

However, although intracerebral haemorrhage (ICH) is a common finding in patients with traumatic brain injury, the role for surgical management is less clear (review, Cochrane Review).

A series of studies have examined the role for surgery in spontaneous ICH have culminated in this RCT of primary surgical vs conservative management of spontaneous intracerebral haemorrhage, which was authored by our neurosurgical colleague Helen Fernandes (editorial, correspondence). The trial suggested that there was little evidence of benefit for surgery however it generated considerable debate. Moreover, the management of spontaneous ICH may not necessarily be applicable to patients with traumatic brain injury.

This week we discuss this important study. It is an international randomised study of 170 patients comparing early surgical evacuation (<12hrs) of intracerebral haematoma vs. conservative management and delayed evacuation if judged to be required. Patients with co-existing extradural or subdural haematoma requiring surgery were excluded as were those where the estimated blood volume was below 10ml and those with >2 focal haemorrhages. The primary outcome was a favourable Glasgow Outcome Score (GOS) at six months. Secondary outcomes included mortality and other quality of life assessments.

Unfortunately, the study was stopped early due to poor recruitment. Although underpowered, the study did show some important and interesting findings with both outcome and mortality benefits (the outcome benefit was not significant, however, at 0.17) . However, we have seen similar results before in TBI research and find it difficult to put results such as these into the context of our own practice – decompressive craniectomy being a good example. Moreover, a comparison of the GOS from the two groups shows that the proportion of patients achieving a good recovery is similar between the two groups but the proportion of those with both moderate and severe disability was greater in the intervention arm – it is hard to know what to do with this.

Overall, despite the small numbers recruited, there was perhaps evidence of benefit but there are significant concerns about its generalisability.

Questions to discuss:

As always we want to establish whether the study was robust enough to inform our practice, was it?

Is this study applicable to our unit given few patients were from the UK and most did not have ICP monitoring?

Does this study question the role of ICP monitoring given that late surgery e.g. for high ICP was associated with poor outcomes?

Is there a subgroup of patients who benefit from early surgery and if so, how can we predict those who will have expanding haematomas?

Great appraisal by Sunderland ICU Medical Education.