Intracranial complications of preinjury anticoagulation in trauma patients with head injury

Background. We have evaluated our recent experience as a Level I trauma center to test the hypothesis that preinjury anticoagulation adversely affects the morbidity and mortality of trauma patients with an intracranial injury. Methods. Records of 380 patients admitted to the trauma service from January 1997 to December 1998 who at the time of admission were taking warfarin, low-molecular-weight heparin, aspirin, nonsteroidal anti-inflammatory drugs, clopidogrel, dipyridamole, pentoxifylline, or naproxen were reviewed. Thirty-seven patients with intracranial injuries were identified and compared with a matched (age, gender, mechanism, and severity of injury) control group of 37 patients with similar head injury but not taking any anticoagulant randomly selected from the trauma registry for that same time period. Results. The control and anticoagulated groups were comparable in terms of age, 75 +/- 8 versus 74 +/- 11 years (p = 0.655); gender, 22 men/15 women versus 21 men/16 women; mechanism of injury, 30 falls/7 motor vehicle crashes versus 30 falls/7 motor vehicle crashes; and length of hospital stay, 11 +/- 14 versus 10 +/- 11 days (p = 0.853). In the anticoagulated group, the mean Injury Severity Score was 17.0 +/- 7.8 and the mean Glasgow Coma Scale score was 11.8 +/- 4.0; these were not significantly different from the control group, which had a mean Injury Severity Score of 19.8 +/- 8.1 (p = 0.143) and a Glasgow Coma Scale score of 12.5 +/- 2.6 (p = 0.378). There were 14 deaths (38%) in the anticoagulation group, versus 3 deaths in the control group (8%) (p = 0.006). In the anticoagulation group, 4 of 12 patients (33%) taking warfarin died, whereas 9 of 19 patients (47%) taking aspirin died (p = 0.285). All deaths were secondary to head injuries; all deaths in the control group and all but one in the anticoagulated group were the result of a fall; 6 of 10 anticoagulated patients who fell on stairs died, and 5 of these were taking aspirin only. Conclusion. These data indicate that the trauma patient with preinjury anticoagulation such as warfarin or even aspirin who has an intracranial injury has a four to fivefold higher risk of death than the nonanticoagulated patient. The efficacy of reversing the anticoagulant effect at the time of hospital admission remains to be evaluated.