Management of traumatic intracranial haematoma.

Abstract
Deciding which head-injured patients should be transferred to a neurosurgical unit can be difficult. Traditional criteria emphasise the development of deteriorating responsiveness but lead to delayed diagnosis and to avoidable mortality and morbidity. To discover if a more liberal admission policy improved results a study was conducted analysing data collected prospectively from 683 patients who had a traumatic intracranial haematoma evacuated in the Glasgow neurosurgical unit between 1974 and 1980. In the first four years, before the change in policy, mortality was 38% but decreased to 29% afterwards. This reflected a reduction in the proportion of patients who talked after injury but who deteriorated into coma before operation--that is, 31% before the change in policy, 16% afterwards. If the potential benefits of CT scanning in the management of head injuries are to be realised patients must be scanned sooner than in the past. This will usually mean that more patients should go to a neurosurgical unit and that simple criteria for transfer should be established.