Surgical trial in traumatic intracerebral haemorrhage (STITCH): A randomised controlled trial of early surgery compared with Initial conservative treatment

Background: While it is accepted practice to remove extradural (EDH) and subdural haematomas (SDH) following traumatic brain injury, the role of surgery in parenchymal traumatic intracerebral haemorrhage (TICH) is controversial. There is no evidence to support Early Surgery in this condition.

Objectives: There have been a number of trials investigating surgery for spontaneous intracerebral haemorrhage but none for TICH. This study aimed to establish whether or not a policy of Early Surgery for TICH improves outcome compared with a policy of Initial Conservative Treatment. Design: This was an international multicentre pragmatic parallel group trial. Patients were randomised via an independent telephone/web-based randomisation service. 

Setting: Neurosurgical units in 59 hospitals in 20 countries registered to take part in the study. Participants: The study planned to recruit 840 adult patients. Patients had to be within 48 hours of head injury with no more than two intracerebral haematomas greater than 10 ml. They did not have a SDH or EDH that required evacuation or any severe comorbidity that would mean they could not achieve a favourable outcome if they made a complete recovery from their head injury. 

Interventions: Patients were randomised to Early Surgery within 12 hours or to Initial Conservative Treatment with delayed evacuation if it became clinically appropriate. 

Main outcome measures: The Extended Glasgow Outcome Scale (GOSE) was measured at 6 months via a postal questionnaire. The primary outcome was the traditional dichotomised split into favourable outcome (good recovery or moderate disability) and unfavourable outcome (severe disability, vegetative, dead). Secondary outcomes included mortality and an ordinal assessment of Glasgow Outcome Scale and Rankin Scale. 

Results: Patient recruitment began in December 2009 but was halted by the funding body because of low UK recruitment in September 2012. In total, 170 patients were randomised from 31 centres in 13 countries: 83 to Early Surgery and 87 to Initial Conservative Treatment. Six-month outcomes were obtained for 99% of 168 eligible patients (82 Early Surgery and 85 Initial Conservative Treatment patients). Patients in the Early Surgery group were 10.5% more likely to have a favourable outcome (absolute benefit), but this difference did not quite reach statistical significance because of the reduced sample size. Fifty-two (63%) had a favourable outcome with Early Surgery, compared with 45 (53%) with Initial Conservative Treatment [odds ratio 0.65; 95% confidence interval (CI) 0.35 to 1.21; p = 0.17]. Mortality was significantly higher in the Initial Conservative Treatment group (33% vs. 15%; absolute difference 18.3%; 95% CI 5.7% to 30.9%; p = 0.006). The Rankin Scale and GOSE were significantly improved with Early Surgery using a trend analysis (p = 0.047 and p = 0.043 respectively). 

Conclusions: This is the first ever trial of surgery for TICH and indicates that Early Surgery may be a valuable tool in the treatment of TICH, especially if the Glasgow Coma Score is between 9 and 12, as was also found in Surgical Trial In spontaneous intraCerebral Haemorrhage (STICH) and Surgical Trial In spontaneous lobar intraCerebral Haemorrhage (STICH II). Further research is clearly warranted. 

Trial registration: Current Controlled Trials ISRCTN 19321911. 

Funding: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 19, No. 70. See the NIHR Journals Library website for further project information.