Predicting long term outcomes following mild traumatic brain injury: Investigating sub-classification systems and individual factors

Abstract

Mild Traumatic Brain Injury (mild TBI) is a significant public health problem, with research indicating up to 95% of TBI cases are mild injuries (Feigin et al. 2012). There is wide variation in recovery following a mild TBI. The majority of people recover well postinjury, whilst others experience on-going mood, cognitive and physical difficulties that impact upon their lives. Indeed, 48% of mild cases experience persistent symptoms 12 months post-injury (Feigin et al. 2012). Identifying people at risk of developing poor outcomes post-mild TBI is a developing area in the research. The primary aim of this study was to investigate existing mild TBI sub-classification systems and determine their utility in predicting 12 month post-injury outcomes. None of the 15 sub-classification systems investigated was able to predict 12 month outcomes. Therefore analyses were run to identify whether individual factors included in the systems in addition to other factors highlighted in the literature might be predictive of outcomes. At 12 months, anxiety was predicted by previous psychological conditions and baseline anxiety, whilst depression was predicted by age and baseline depressed mood; the Neurocognitive Index was predicted by education, baseline physical weakness and anxiety; and composite memory was predicted by concentration difficulty at baseline. Post-concussive symptoms at 12 months were predicted by gender, baseline depression and sleep. There were no significant predictors found for complex attention at 12 months. The findings suggest that existing sub-classification systems in their current form are unable to predict long-term outcomes and there is a need for a new model to be developed to serve as a predictive tool.