Hayley Falk, Kathleen T. Bechtold, Matthew E. Peters, Durga Roy, Vani Rao, Mariel Lavieri, Haris Sair, Timothy E. Van Meter, and Frederick Korley.



A multi-variable logistic regression model using clinical data only was developed using HeadSMART data. Six variables: age, positive head CT, history of depression, moderate or severe headache, difficulty concentrating, and photophobia on the day-of-injury were found to form a model predictive of 1 month recovery status in mild TBI (GCS 15). This model has modest performance (AUC 0.79) but provides information useful in determining outcome risk, particularly when these clinical elements are combined with biological and functional biomarkers.