Randomized Controlled Trial Investigating Optimal Treatment for Chronic Subdural Hematoma

Study design The study is set up as an open randomized controlled three-arm trial. Patients will be randomized into three groups.

Study setup and procedures The goal of this study is to determine if one surgical treatment for chronic subdural hematoma is better than the other. Patients with a clear indication for drainage of subdural hematoma (as stated under "Surgical options") will be randomized into three groups. One group will receive twist drill craniostomy followed by drainage during 48 to 96 hours. One group will undergo burr hole drainage (single if possible, double if necessary) with irrigation and drainage during 48 to 96 hours postoperatively. One group will undergo a minicraniotomy with trephine or craniotome, with wide opening of all visible membranes, rigorous irrigation and placement of Jackson-Pratt drain, followed by drainage during 48 to 96 hours.

Patients diagnosed with chronic subdural hematoma on CT scan, will be operated on as soon as possible. The type of operation will be determined by randomization. The duration of surgery as well as presence of brain reexpansion (in case of burr hole or minicraniotomy) will be recorded. Postoperatively, patients will be transferred to the neurosurgery ward with a subdural drainage system in place. Drainage system is removed 48 to 96 hours postoperatively, regardless of the type of operation (48 hours standard; longer period of drainage if drainage less than 50cc/24h evaluated on the second 24 hours postoperatively. Maximal time of continuous drainage is 4 days or 96 hours postoperatively). Drained volume will be recorded. Evaluation will be performed preoperatively, 48h postoperatively, on discharge, after 6 weeks and after 6 months, and will include:

• Age, sex, comorbidities (anticoagulation or coagulopathy, dementia, etc)
• Clinical status (cognitive state (orientation in time, place and person), Speech (aphasia/dysarthria), Motor function (hemiplegia/paresis); walking independently (y/n); independence for ADL (y/n); Markwalders Scale, modified Rankin Scale
• Imaging characteristics (location (frontal/parietal/occipital, uni-bilateral), size (max diameter), Density(Hounsfield units, age of hematoma), membranes (y/n). An approximate total of five CTscans will be performed, totaling (5x2=)10mSv of radiation exposure Length of hospital stay will be recorded. The primary endpoint will be the reoperation rate at 40 days. We hypothesize that subdural hematomas treated by minicraniotomy will require less reoperation (reduced to 5%) than burr hole craniostomy or twist drill craniotomy (15 to 25% according to literature).

Secondary endpoints: Clinical endpoints: Duration of operation, technical difficulties during operation. Duration of hospital stay. Improvement of Clinical status and speed thereof, Improvement of CT-imaging characteristics, complication rate (bleeding, local or systemic infection, epilepsy,…). We expect that subdural hematomas operated on by minicraniotomy and large opening of membranes will lead to quicker improvement of CT-imaging than burr hole treatment or twist drill craniotomy.

Exploratory endpoints To determine which type of intervention can be beneficial to which type of patients. Hypothesis is that in case of homogenous chronic subdural hematoma, twist drill craniostomy would pose the least risks, whereas in multiloculated mixed-density hematomas, minicraniotomy would be the best option.

The subjects Number of subjects A total of 150 patients will be included in the study. Inclusion criteria All consecutive patients with chronic subdural hematoma with indication for surgical intervention: clinical symptoms as a result of the subdural collection, or important mass effect (as indicated by a midline shift of more than 5mm) Exclusion criteria Patients under the age of 18 will be excluded Patients who have undergone previous cranial surgery which would limit surgical options (for example bone flap already in place) Replacement of subjects Dropout, in the case of this trial, would only signify loss to follow-up since patients will be operated on after consenting to participation in the trial. In case of dropout, the available data will be evaluated and included in the analysis when possible. If dropout occurs after the scheduled six weeks postoperative visit, data is still usable for analysis of primary endpoint. Data from earlier on in treatment (for example duration of surgery, hospital stay, etc) will be usable for analysis of secondary endpoints.

Restrictions and prohibitions for the subjects Antiaggregant drugs (Aspirin, Asaflow, Cardioaspirine, Plavix, Clopidogrel, Ticlid,…) will be stopped for at least 5 days (ideally 10days) before surgery when possible and, if necessary (this will be discussed with the treating physician/cardiologist), replaced by short acting low molecular weight heparins given subcutaneously (no administration on the day of surgery). Anticoagulants (Sintrom, Marcoumar, Marevan) will be replaced by short acting low molecular weight heparins given subcutaneously (no administration on the day of surgery).

Study analysis Sample size calculation Power and sample size estimations are based on data from the literature indicating that the reoperation rate at 40 days for burr hole craniostomy or twist drill craniotomy ranges from 25% to 15% (control groups), and our assumption that minicraniotomy would decrease the reoperation rate at 40 days to 5%. In a test for trend in proportions, sample sizes of 49, 49, and 49 are obtained from 3 groups with proportions equal to 0.25, 0.15, and 0.05. The total sample of 147 subjects achieves 81% power to detect a linear trend using a two-sided Z test with continuity correction and a significance level alpha of 0.05.

Randomization Randomization will be implemented using Research Randomizer (Urbaniak, G. C., Plous, S. (2011). Research Randomizer (Version 3.0) [Computer software]) Data for randomization was retrieved on September 25th, 2011, from http://www.randomizer.org. Three treatments will be randomized for 50 reps (totaling 150 patients) Analysis of the samples Statistical analysis will be performed by Patrick Haentjens (Uz Brussel, Laarbeeklaan 101 1090 Jette; tel 024777909) Statistical analysis Data will be presented as number of cases including nominator and denominator values (and percentages) for categorical variables and as mean values and standard deviation (SD) for continuous variables. Differences between treatment arms will be presented as absolute between-group differences with corresponding 95% confidence intervals (95% CI). Differences between treatment arms will be assessed by one-way ANOVA for continuous variables and the test for trend in proportions for categorical variables. All tests will be two-sided, and a P-value of less than 0.05 will be considered to indicate statistical significance.

Quality control and quality assurance Continuous ad-hoc analysis of clinical results and of data acquisition and data management will take place in a weekly meeting with all cooperating neurosurgeons; this clinical analysis is independent of statistical analysis. In case of aberrant results from one procedure or another, results will be evaluated and measures will be taken.