RESUMO
BACKGROUND: Prediction of traumatic brain injury (TBI) among children is of great importance for accurate clinical decision making. OBJECTIVES: This study aimed to determine the prognostic value of the Rotterdam scoring system in predicting early outcome among children with TBI. METHODS: This study was conducted in 2017 on 506 children with brain injury in Kashan, Iran. A checklist was used to collect demographic and clinical characteristics of patients such as age, sex, mechanism of trauma, Glasgow Coma Scale (GCS) score, need for surgery, and brain injury outcome. Moreover, each participant's computed tomography scan was evaluated and scored using the Rotterdam system. Sensitivity, specificity, positive and negative predictive values, and the best cut-off score were calculated for the Rotterdam system. The relationships of the Rotterdam score with participants' characteristics were examined using the χ2 test, whereas the predictors of brain injury outcome were identified using the logistic regression analysis. RESULTS: Pediatric death rate was 4.3%. Most deaths were among children who were male, aged <4, had developed brain injury owing to traffic accidents, had a GCS score of 3-8, suffered from compressed skull fracture and frontal lobe injury, had cerebral edema, and had a Rotterdam score of 5. The sensitivity and specificity of a Rotterdam score 3 were 86.4% and 97.9%, respectively. The logistic regression analysis indicated that only GCS and Rotterdam scores were significant predictors of brain injury outcome. CONCLUSIONS: At a cut-off score of 3, the Rotterdam system can be used to predict TBI outcome among children with acceptable sensitivity and specificity.
