Background: Although anticoagulation with warfarin has been associated with increased risk of adverse outcomes after trauma, the effects of the new oral agents (NOA) such as dabigatran, apixaban, rivaroxaban are not yet well characterized.
Methods: A retrospective review of a level 1 trauma center database identified all patients aged ≥ 50 admitted after trauma during a 24 month period starting September 2013. Demographics, including preadmission anticoagulation agents, injuries, hospital course and outcomes were abstracted from the electronic medical record.
Result: Over the 24-month period, 3,392 patients were admitted; 112 (3.3%) were anticoagulated with NOA and 373 (11.0%) with warfarin with a trend toward increasing utilization of the new agents compared with warfarin over that period. Although comparable in age, injury severity scores, and mechanism of injury, patients anticoagulated with warfarin had both a higher overall mortality (10.9%) compared with the NOA (6.25%) and the non-anticoagulated control (5.5%) groups (p < 0.001) as well as a higher trauma-related mortality (9.0%) versus NOA (2.8%) and control (3.7%) groups (p < 0.001). Patients on warfarin or NOA were admitted to intensive care unit or step down unit more frequently than control patients. (45.0% and 41.9% vs. 35.7% respectively; p < 0.001). The incidence of traumatic brain injury was similar among the three groups. Although it did not reach statistical significance, trauma-specific mortality in the traumatic brain injury subset was higher in the warfarin group (19.3%) than the NOA (16.7%) or control (10.9%) groups (p = 0.08). In a multivariable logistic regression, warfarin (odds ratio, 2.215; 95% confidence interval, 1.365-3.596; p = 0.001), but not the NOA (odds ratio, 0.871; 95% confidence interval, 0.258-2.939; p = 0.823), was an independent predictor for mortality.
Conclusions: Although the experience with the new oral anticoagulation agents is still limited, patients on these agents appear to have lower mortality after traumatic injury than patients on warfarin.
Level of evidence: Epidemiologic study, level III.