The review concluded that, in spontaneous intracerebral haematoma patients, surgical treatment was superior to medical treatment when haematoma volume was large (over 40mL) and the Glasgow Coma Score was at least 6 (severe to very severe). The review had some methodological problems and the quality of the included trials was uncertain, so caution is warranted when interpreting the authors’ conclusions.

To define the position of surgery preference in the treatment choice for spontaneous intracerebral haematoma and to compare the efficacy of surgery with medical treatment.

PubMed and EMBASE were searched from inception to 2010. Search terms were not reported. Reference lists of identified articles and relevant cited references were reviewed.

Randomised controlled trials (RCTs) of treatment for spontaneous spontaneous intracerebral haematoma were eligible for inclusion.

The included trials studied medical treatment and surgical treatment in patients with supratentorial haematoma location. The type of surgery performed was primarily craniotomy; the length of surgery ranged from within three hours of admission to within 72 hours of admission (where reported). Two thirds of the trials were single centre; the remainder were multi-centre. The age of patients ranged from 15 to 80 years (mean age 64.2 years), where reported. The reported outcomes included death and dependency.

The authors did not state how many reviewers performed study selection.

The authors did not state that they assessed trial quality.

Data were extracted on death and dependency, and used to calculate odds ratios (ORs) and 95% confidence intervals (CIs). The authors did not state how many reviewers were involved in data extraction.

Meta-analysis was used to calculate pooled odds ratios and 95% confidence intervals. Subgroup analyses were carried out for age, admission time, haematoma volume, surgical time (under 24 hours versus over 24 hours), and Glasgow Coma Scale (GCS) rating.

Eighteen RCTs were included in the review (n=5,511 participants; 4,969 were included in the analysis). The number of medically treated participants included in the analysis was 3,200 compared with 1,769 surgically treated participants. Sample size ranged from 10 to 1,036. The follow-up time ranged from one to 24 months (mainly six months), where reported.

Compared with medical management, surgical treatment did not have a statistically significant effect on death and dependency (OR 0.92, 0.67 to 1.25; 18 trials).

Subgroup analyses indicated that results for morbidity and mortality were significantly better with surgery compared with medical treatment for Glasgow Coma scores of at least 6, early surgery before 24 hours (morbidity only), and haematoma volume that were greater than 40mL. Results were better for morbidity depending on the type of endoscopic approach.

Surgical treatment results were found to be superior to medical treatment in patients with a haematoma volume greater than 40mL and Glasgow Coma Scores of at least 6 (severe to very severe). The included trials were not adequate to analyse the best type of surgery.

Inclusion criteria for the review were not clearly defined. Only two relevant data sources were searched. Publication bias was not assessed and could not be ruled out. There may have been the potential for language bias, as it was not clear whether language restrictions were imposed. It was not stated if attempts were made to reduce reviewer error and bias during the review, so it was not possible to assess the methodological rigour of the review process.

There was no discussion of the quality of the included trials; quality assessment did not appear to have been undertaken. However, many trials had small sample sizes and the follow-up was generally short. A large number of patients also underwent the medical intervention compared with the surgical procedure. Trials were combined using meta-analysis, but no discussion of statistical heterogeneity was provided, although clinically there were differences across the trials.

This review had some methodological problems and the quality of the included trials was uncertain, so caution is warranted when interpreting the authors’ conclusions.

The authors did not state any implications for research or practice.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.