One hundred and seventy-two (73%) patients survived during hospitalization. All the survivors were Caucasians. Attempts to contact all survivors were made over the phone, by mail, and by home visits in a period between 1 and 6 years after hospitalization. A total of 46 patients completed find more neuropsychological assessment between March 2007 and April 2010 (see Figure 1). Control subjects matched for gender, age, and education level were recruited during the same period by convenience.

They were companion persons of patients from other outpatient clinics and had no previous history of neurological or psychiatric disorders (Table 1). Variables collected prospectively during hospitalization included gender, age, admission GCS, admission pupils’ examination, admission serum glucose from peripheral

blood and presence of associated trauma. The admission brain computed tomography (CT) scans were analysed according to the Marshall classification (Marshall et al., 1992) and presence of subarachnoid haemorrhage (SAH). The Marshall classification includes Type I injury, normal CT; Type II injury, http://www.selleckchem.com/products/abc294640.html small lesions but visible cisterns and no midline shift; Type III injury, diffuse swelling with non-visible cisterns; Type IV injury, unilateral swelling with midline shift deviation higher than 5 mm; Type V injury, evacuated mass lesion; and type VI injury, non-evacuated lesion with higher than 25 mm volume. Variables collected during cognitive evaluation were hand dominance before TBI and years of education at the time of neuropsychological assessment. The MCE公司 neuropsychological assessments were performed by a neuropsychologist (*names removed for purposes of anonymity*) blinded for all the clinical, radiological, neurosurgical, and laboratory

variables, on average 3 (SD ±1.8) years after hospitalization (range between 1 and 6 years). When necessary, the individual evaluation scores were discussed between the two examiners and a third neuropsychologist (*names removed for purposes of anonymity*). Patients were evaluated individually with each patient in the TBI outcome clinic of our University Hospital. The time needed to complete the assessment was 76 (±12) and 81 (±18) min for controls and patients, respectively (p = .21). We analysed the raw scores of the neuropsychological tests and adjustments for age and education were done including these variables in the regression model (see ‘Statistical analysis’). The test battery applied was chosen based on literature and experience of neuropsychologists.