Degree of Anticoagulation, but Not Warfarin Use Itself, Predicts Adverse Outcomes After Traumatic Brain Injury in Elderly Trauma Patients

Abstract

The relationship between preinjury warfarin use and outcomes after traumatic brain injury in elderly trauma patients remains controversial. We hypothesized that, among elderly warfarin users, the degree of anticoagulation, rather than warfarin therapy itself, would predict the severity of traumatic brain injury. Retrospective study (2004-2006) of all elderly trauma patients (age >/=65 years) who were evaluated by the trauma service at a Level I trauma center and underwent computed tomography of the head for suspicion of an intracranial injury was performed. Three cohorts were grouped: (1) warfarin users with an admission International Normalized Ratio >/=2 (therapeutic group), (2) warfarin users with an admission International Normalized Ratio <2 (nontherapeutic group), and (3) warfarin nonusers. Main outcome variables were presenting with a Glasgow Coma Scale (GCS) score </=13 points, intracranial hemorrhage (ICH), overall mortality, and mortality after ICH. A total of 225 trauma patients were studied, including 40 warfarin users (17.3%), of whom 22 (55.0%) were in the therapeutic group. Age, gender, and mechanism of injury were similar among groups. Likelihood of Glasgow Coma Scale score </=13 (odds ratio [OR] = 5.13, 95% confidence interval [CI] 1.97-13.39, p = 0.001), ICH (OR = 2.59, 95% CI 0.92-7.32, p = 0.07), overall mortality (OR = 4.48, 95% CI 1.60-12.50, p = 0.004), and mortality after ICH (OR = 3.42, 95% CI 1.09-10.76, p = 0.03) was increased in the therapeutic as compared with the nonuser group. There was no difference in any measured outcome between the nonuser and nontherapeutic groups. Therapeutic anticoagulation with warfarin, rather than warfarin use itself, is associated with adverse outcomes after traumatic brain injury in elderly patients.