Association of Thromboelastography with Progression of Hemorrhagic Injury in Children with Traumatic Brain Injury.

ABSTRACT

INTRODUCTION: Progression of hemorrhagic injury (PHI) in children with traumatic brain injury portends poor outcomes. The association between thromboelastography (TEG), functional coagulation assays, and PHI is not well characterized in children. METHODS: This was a retrospective cohort study of children presenting with PHI at a pediatric level I academic trauma center from 2015 to 2020. Inclusion criteria were as follows age less than 18 years, intracranial hemorrhage on admission head computed tomography scan, and admission rapid TEG assay and conventional coagulation tests. PHI was defined by the following radiographic criteria any expansion of or new intracranial hemorrhage on subsequent head computed tomography scan. Rapid TEG values included Activated Clotting Time (ACT), alpha angle, maximum amplitude, and lysis at 30 min. Wilcoxon rank-sum test was used to assess baseline differences between groups with PHI and without PHI, including laboratory assays. Univariate analysis was performed to examine the association between variables of interest and PHI. Patients were dichotomized on the basis of this cut point to generate a "low ACT" group and a "high ACT" group. These variables were included in a multivariable logistic regression model to determine independent association with traumatic brain injury progression. RESULTS: In total, 219 patients met criteria for analysis. In this cohort, the median (interquartile range [IQR]) age = 6 (2-12) years, median (IQR) Injury Severity Score = 21 (11-27), 68% were boys, and 69% sustained blunt injury. The rate of PHI was 25% (54). Median (IQR) time to PHI was 1 (0-4) days. Children with PHI had a higher Injury Severity Score (p < 0.001), lower Glasgow Coma Scale (p < 0.001), greater incidence of shock (p = 0.04), and lower admission hemoglobin (p = 0.02) compared with those without PHI. Children with PHI had a higher International Normalized Ratio (INR) and longer TEG-ACT; other TEG values (alpha angle, maximum amplitude, and lysis at 30 min) were not associated with PHI. In the logistic regression model accounting for other covariates associated with PHI, elevated ACT remained an independent predictor of progression (odds ratio = 2.25, 95% confidence interval 1.09-4.66; p = 0.03; area under the receiver operating characteristic curve = 0.76). After adjusting for confounders, INR fell out of the model and was not an independent predictor of progression (odds ratio = 1.32, 95% confidence interval 0.60-2.93; p = 0.49). CONCLUSIONS: Although INR was elevated in children with PHI and has been associated with poor clinical outcomes, only admission TEG-ACT was independently associated with PHI. Further study is warranted to determine whether TEG-ACT reflects an actionable therapeutic target.