Introduction Cerebral gunshot wounds are highly lethal and literature on the clinical scores for mortality prediction is limited.
MATERIALS AND METHODS A retrospective study was undertaken over a 5-year period at the Pietermaritzburg Metropolitan Trauma Service in South Africa. A simplified clinical prediction score was developed based on clinical and/or physiological variables readily available in the resuscitation room.
RESULTS A total of 102 patients were included; 92% (94/102) were male and the mean age was 29 years; 22% (22/102) died during the admission. The presence of visible brain matter (odds ratio 12.4, P = 0.003) and motor score less than 5 (odds ratio 89.6, P < 0.001) allows the prediction success of 92% (sensitivity 73% and specificity 98%). The area under the receiver operating characteristic curve was 94% (95% confidence interval 88-100%, P < 0.001).
CONCLUSIONS The presence of visible brain matter, together with a motor score of less than 5, allows accurate identification of non-survivors of cerebral gunshot wounds. Further study is required to validate this score.