Prognosis as the Basis for Selection to Treatment After Severe Head Injury

After severe head injury intensive care is sometimes given to patients who turn out to be fatally injured or who never regain mental activity. Large scale prognostic work has been performed in many centres in order to administer treatment and give advice on the basis of previous clinical experience. Jennet et al. were able to predict confidently (more than 0.97 probability) death or survival after severe head injury in no more than 61% of their patients when clinical data for up to 3 days was available1. In addition to this type of clinical data, Narayan et al. used CT scan, ICP monitoring (ventricular catheter or a subarachnoid screw) and multi-modality evoked potentials2. They achieved 64% confident predictions (90% confidence level) by a combination of clinical and evoked potentials data from the first day after accident. Their statistical methods were, however, different from those of Jennett2.