• A Fabbri, F Servadei, G Marchesini, S C Stein, and A Vandelli.
• Dipartimento dell'Emergenza, Presidio Ospedaliero Morgagni-Pierantoni, Azienda USL Forlì, Via Forlanini 34, I-47100 Forlì, Italy. andrea.fabbri@formulatre.it
• J. Neurol. Neurosurg. Psychiatr. 2008 May 1;79(5):567-73.

BackgroundSubjects with moderate head injury are a particular challenge for the emergency physician. They represent a heterogeneous population of subjects with large variability in injury severity, clinical course and outcome. We aimed to determine the early predictors of outcome of subjects with moderate head injury admitted to an Emergency Department (ED) of a general hospital linked via telemedicine to the Regional Neurosurgical Centre.Patients And MethodsWe reviewed, prospectively, 12,675 subjects attending the ED of a General Hospital between 1999 and 2005 for head injury. A total of 309 cases (2.4%) with an admission Glasgow Coma Scale (GCS) 9-13 were identified as having moderate head injury. The main outcome measure was an unfavourable outcome at 6 months after injury. The predictive value of a model based on main entry variables was evaluated by logistic regression analysis.Findings64.7% of subjects had a computed tomographic scan that was positive for intracranial injury, 16.5% needed a neurosurgical intervention, 14.6% had an unfavourable outcome at 6 months (death, permanent vegetative state, permanent severe disability). Six variables (basal skull fracture, subarachnoid haemorrhage, coagulopathy, subdural haematoma, modified Marshall category and GCS) predicted an unfavourable outcome at 6 months. This combination of variables predicts the 6-month outcome with high sensitivity (95.6%) and specificity (86.0%).InterpretationA group of selected variables proves highly accurate in the prediction of unfavourable outcome at 6 months, when applied to subjects admitted to an ED of a General Hospital with moderate head injury.

52 keywords...

hide…