Résumé
An analysis of 3794 craniocerebral missile injuries managed, monitored and treated by the Sher-i-Kashmir Institute of Medical Sciences [SKIMS], India, over a period of 20 years from September 1988 - December 2008 was analyzed. It revealed an overall mortality of 88% [3327 out of 3794] and a good survival of 5% [176 out of 3794]. Most of the deaths occurred within 30 minutes of the patient's arrival to the hospital and only 694 patients lived beyond one hour of arrival. All patients with an admission Glasgow Coma Scale [GCS] of 3 and the majority of those with 4 died. No deaths occurred in the group of patients with GCS score 9-15. Poor and delayed mode of transportation accounted for up to 90% of deaths. Predictors of poor outcome were low admission GCS score, fixed and dilated pupils, poor and delayed mode of transportation, haemodynamic instability, abnormal breathing at admission, coagulopathy and disseminated intravascular coagulation, CT visualisation of subarachnoid haemorrhage, ventricular haemorrhage, midline shift, bihemispheric and multilobar injuries and scalp wounds at occipital, temporal and frontal areas. However, retained missile and bone fragments were less harmful than retained wooden [pulped mulberry stem] and cardboard wads and pneumocephalus in causing infections, cortical atrophy and seizures in the long run