A large simple placebo controlled trial, among adults with head injury and impaired consciousness, of the effects of a 48-hour infusion of corticosteroids on death and neurological disability. CRASH was a randomised, controlled, double-blind trial undertaken in 239 hospitals in 49 countries. A total of 10008 adults with head injury and a Glasgow Coma Score (GCS) of 14 or less within 8 hours of injury were randomly allocated 48 hour infusion of corticosteroids (methylprednisolone) or placebo. Primary outcomes were death within 2 weeks of injury or disability at 6 months. Prespecified subgroup analyses were based on injury severity (GCS) at randomisation and on time from injury to randomisation and analysis was by intention to treat. Access to this dataset is available via https://freebird.lshtm.ac.uk/.

The data set contains patient demographics (age, sex), baseline measures (GCS, pupil reactiveness) and time from injury. The early outcome form (completed at 14 days or prior death, transfer or discharge) comprises information on patient status (dead or alive) and symptoms on transfer/discharge. Also included is patient management (days in ICU, neurosurgical operation) and complications (seizure, haematemesis or melaena requiring transfusion, wound infection with pus, pneumonia and major extracranial injury). Information was also collected on first head CT scan and on trial treatment administered (loading dose and duration of maintenance dose). If the patient was transferred to another hospital prior to 14 days an outcome form was completed at the transfer hospital capturing exactly the same information as the early outcome form and completed at day 14 or prior death, transfer or discharge. At 6 months a questionnaire (Glasgow Outcome Score – with 5 outcome levels ranging from Death to Good Recovery) was completed by the patient alone or by a relative or friend alone or by the patient together with a friend or relative. A small number of patients were allocated the extended Glasgow Outcome Score – with 8 outcome levels ranging from Death to Upper Good Recovery). The data set is a comma separated value file containing 10009 rows comprising a header row of data field names (variables) and 10008 patient records. The data set includes 87 data fields (variables) including the unique 6-digit identifier for each patient in the form of a 4 digit (treatment box) number and a 2 digit (treatment pack) number. Excluded are country, hospital code and intervention (i.e. treatment allocation). These may be obtained on direct request to LSHTM CTU.

Corticosteroids, Injury, Methylprednisolone