Comparative study of computed tomographic scoring systems and predictors of early mortality in severe traumatic brain injury

Abstract

Traumatic Brain Injury (TBI) is a significant cause of disability, mortality and socio-economic loss in developing countries. Prognostication and early prediction of outcome will help in judicious allocation of resources. The three CT scoring systems, commonly employed in predicting outcome are the Marshall, Rotterdam and the Helsinki CT Scoring system. This is a prospective observational study to compare the above scoring systems and their role in predicting the outcome after TBI. A total of 157 patients were included in this study. Primary objective was early outcome assessed by Glasgow Outcome Scale at discharge and 3 months. On comparing the three CT scoring systems, the Helsinki CT score provided a better positive predictive value of 87.5% for mortality in comparison with Marshall (79.3%) and Rotterdam Score (82.5%). Summarising the results for ROC analysis of the three CT scoring systems, the area under curve was 0.742 for Marshall score, 0.751 for the Rotterdam score and 0.815 for the Helsinki score, which upholds the superiority of the Helsinki CT score in predicting the mortality compared to other two. The linear relationship of higher the score with added risk of mortality was seen with Rotterdam score and Helsinki score. Individual CT findings of positive midline shift, obliteration of suprasellar cisterns, presence of subarachnoid haemorrhage, intraventricular haemorrhage and large mass lesions were found to be independent predictors of mortality. The Helsinki Scoring system is a better prognostic model with significantly improved outcome prediction accuracy, although the Marshall and Rotterdam scoring systems have good predictability for assessing mortality.