To investigate the clinical features of infancy traumatic brain injury (TBI) and the prognostic value of the Trauma Infant Neurologic Score (TINS), infants <2 years of age with TBI who were admitted from 2000 to 2007 were retrospectively studied. Fifty-six patients with a mean age of 13.3 +/- 6.5 months (range = 2-24) were identified. The clinical diagnoses, in terms of the severest injury, included scalp hematomas (n = 2), skull bone fractures (n = 3), epidural hematomas (n = 21), subdural hematomas (n = 14), cerebral contusion and laceration (n = 4), intracerebral hematomas (n = 7), traumatic subarachnoid hemorrhage (n = 2), diffuse axonal injury (n = 2) and diffuse brain swelling (n = 1). The most common clinical presentations were vomiting (66.1%), paleness (55.4%), irritability (37.3%), pupillary abnormalities (35.7%) and altered consciousness (32.1%). The mechanism of injury included falls (n = 41), vehicle accident (n = 9), abuse (n = 4) and unknown (n = 2). The TINS score ranged from 1 to 10 with a mean of 3.6 (SD = 2.4) in the whole patient cohort. The Children's Coma Scores (CCS) on admission were 13-15 (n = 31), 9-12 (n = 7) and 3-8 (n = 18). Thirty-nine of the infants were operated on and the other 17 infants were treated nonsurgically. Forty-eight patients (86%) were followed up for a period of 1-8 years (mean = 4.4) after discharge. In the followed-up patient cohort, the mean TINS score at admission was 3.8 +/- 2.5. The total clinical outcome, according to the Glasgow Outcome Scale (GOS), was: 37 (77.1%) good recovery, 4 (8.3%) moderately disabled, 1 (2.1%) vegetative and 6 (12.5%) dead. For those who were operated on the outcome was: 25 (78.1%) good recovery, 4 (12.5%) moderately disabled and 3 (9.4%) dead, and for those who were not operated on: 12 (75.0%) good recovery, 1 (6.3%) vegetative and 3 (25.0%) dead. At two years of follow-up, the GOS included 34 (73.9%) good recovery, 3 (6.5%) moderately disabled, 2 (4.3%) severely disabled, 1 (2.2%) vegetative and 6 (13.0%) dead. Statistical tests revealed that the TINS scores were highly associated with the GOS. Higher TINS scores resulted in worse clinical outcome. The CCS scores were also to some degree associated with the GOS score. However, the CCS score on admission was not as discriminating as TINS, predicting only the best and worst outcome in our series. Our study showed that the clinical features of TBI in infants were different from those seen in adults regarding both the distribution of the pathology type and the clinical presenting symptoms. We found that the TINS scoring system is useful for predicting prognosis and outcome in infancy TBI and suggest that it could be routinely used in the infantile population. Copyright (C) 2010 S. Karger AG, Basel