L. E. Folkerson1, D. Sloan1, B. A. Cotton1, J. B. Holcomb1, J. S. Tomasek1, C. E. Wade1 1University Of Texas Health Science Center At Houston,Houston, TX, USA
Introduction: Traumatic brain injury (TBI) and acute coagulopathy of trauma have been the focus of much research as the combination leads to major disability and death. Progressive hemorrhagic injury (PHI) is associated with increased operative interventions and poor outcomes. Identifying which subset of patients will experience PHI based on initial head CT and laboratory coagulation data has proven difficult. We hypothesize that a subtype of TBI and coagulation status would be predictors of PHI.
Methods: This was a retrospective analysis from a single institution of adult patients who presented with the highest level of trauma activation between October 2010- May 2013, (n=1645). Patients (n=617) were identified who underwent at least 2 head CT scans within 24 hours of presentation. Patients with polytrauma (AIS ≥ 3 in all areas other than the head) and those on pre-hospital anticoagulants were excluded, leaving 279 patients in the study group with isolated TBI. Rapid thrombelastography (rTEG) was obtained on Emergency Department (ED) arrival and coagulopathy was defined as an ACT ≥128, MA ≤55, LY-30 ≥3.0 or platelet count ≤150 x 103/µL. Subtypes of TBI were categorized into the following groups: subdural hemorrhage (SDH), subarachnoid hemorrhage (SAH), intraparenchymal contusion/hemorrhage (IPC/H), epidural hematoma (EDH) and combined (CD). PHI was defined as an increase in size of the intracranial hemorrhage from initial head CT as determined by a radiologist. Operative intervention and patient outcomes were recorded. Multivariable logistic regression was used to assess the effect of subtype and coagulation status on PHI. Data are reported as median (IQR).
Results: Of the 279 patients evaluated, 157 patients (56%) had PHI on repeat head CT and 122 (44%) were stable. There was no significant difference in admission GCS, systolic blood pressure or base deficit between groups; all p>0.3. Patients with PHI were older, 44 (27-58) vs 35 (25-52); had a greater incidence of platelet count ≤150 x 103/µL (13% vs 7%); greater incidence of IPC/H (34% vs 11%); a higher ISS, 21 (16-26) vs 17 (14-25); less hospital free days, 14 (0-23) vs 24 (14-27); higher mortality, (17% vs 4%); and higher likelihood of operative intervention (45% vs 34%, OR 1.8, 95% CI 1.0-3.0); all p <0.001. There were no differences in TEG parameters between groups or the incidence of coagulopathy (PHI 51% vs stable 42%; OR 0.6, 95% CI 0.33-0.92). When controlling for age, GCS, and coagulopathy, patients with an IPC/H were more likely to experience PHI than patients with other subtypes of TBI (OR 4.3 p <0.0001, 95% CI 2.2-8.4).
Conclusion: This retrospective analysis demonstrates that patients with IPC/H on initial head CT are more likely to experience PHI. TEG parameters and coagulopathy were not as strong predictors as IPC/H and age for PHI. Therefore, the presence of IPC/H in older patients should raise concern about the probability of PHI.