Article Text
PDF
Abstract
Introduction It is difficult to predict outcomes in patients with severe traumatic brain injury (sTBI) using current clinical tools. The goal of this study was to develop a predictive model for post-traumatic vasospasm in sTBI patients by combining clinical data and serum inflammatory and neuronal proteins for improved prognostication.
Methods Fifty-three adult civilian patients were prospectively enrolled in the sTBI arm of the Surgical Critical Care Initiative (SC2i). SC2i is a consortium of research institutions working to implement biology-driven critical care to deliver precision medicine to individual patients. Patient clinical and serum inflammatory and neuronal protein data were combined and evaluated using the machine learning methods of LASSO and CART to construct parsimonious models for predicting development of post-traumatic vasospasm. The results underwent 100-fold cross validation to assess their robustness.
Results Thirty-six (67.9%) patients developed vasospasm and 10 (18.9%) died. The mean age was 39.2; 22.6% were women. There were an equal number of white (25) and black (25) patients and three from other races. For vasospasm prediction, CART identified lower IL9, lower presentation pulse rate and higher eotaxin as predictors (AUC = 0.89, cross validated AUC = 0.47). LASSO improved upon the CART high risk model by identifying higher Rotterdam CT score (p=0.02) and lower age (p=0.01) as risk factors for vasospasm development (full data AUC 0.94, sensitivity 0.86, specificity 0.94; cross-validation AUC 0.87, sensitivity 0.79 and specificity 0.93).
Conclusion Inflammatory and glial-specific protein levels following sTBI may have predictive value that exceeds conventional clinical variables for certain outcomes. Eotaxin, IL9, and pulse rate predict development of post-traumatic vasospasm. These results warrant validation in a prospective cohort.
Disclosures R. Rindler: None. H. Robertson: None. S. Eshraghi: None. S. Grey: None. S. Schobel-McHugh: None. N. Boulis: None. J. Grossberg: 1; C; Grant: GRA, EMCF, Neurosurgery Catalyst. 2; C; Cognition.