Improving Prediction of Favourable Outcome After 6 Months in Patients with Severe Traumatic Brain Injury Using Physiological Cerebral Parameters in a Multivariable Logistic Regression Model

• DOI: 10.1007/s12028-020-00930-6
• Publication Date: 2020-02-13T19:03:50Z
• Authors: Frank C. Bennis; Bibi Teeuwen; Frederick A. Zeiler; Jan Willem Elting; Joukje van der Naalt; Pietro Bonizzi; Tammo Delhaas; Marcel J. Aries
• Keywords: Stepwise regression

Abstract Background/Objective Current severe traumatic brain injury (TBI) outcome prediction models calculate the chance of unfavourable after 6 months based on parameters measured at admission. We aimed to improve current with addition continuously neuromonitoring data within first 24 h intensive care unit neuromonitoring. Methods Forty-five TBI patients intracranial pressure/cerebral perfusion pressure monitoring from two teaching hospitals covering period May 2012 January 2019 were analysed. Fourteen high-frequency physiological selected over multiple time periods start (0–6 h, 0–12 0–18 0–24 h). Besides systemic and extended Corticosteroid Randomisation Significant Head Injury (CRASH) score, we added estimates (dynamic) cerebral volume, compliance cerebrovascular reactivity indices model. A logistic regression model was trained for each predict months. The using forward feature selection. Each validated by leave-one-out cross-validation. Results CRASH as sole parameter resulted in an area under curve (AUC) 0.76. For period, increased AUC found up 5 additional parameters. highest (0.90) 0–6 that describe mean arterial blood indices. Conclusions can be improved bedside As these factors might modifiable treatment during admission, testing a larger (multicenter) set is warranted.