Warfarin: damned if you do, damned if you don't

Clinical Question

Is it safe to resume warfarin in patients who have had a warfarin-associated intracranial hemorrhage?

Bottom Line

In patients who resume warfarin after an intracranial hemorrhage (ICH) associated with warfarin, 22% will have serious rebleeding. However, in those who don't resume warfarin, 20% will have venous thromboembolic events. This study demonstrates that it is not easy to care for patients who take warfarin. (LOE = 2b-)


Claassen DO, Kazemi N, Zubkov AY, Wijdicks EF, Rabinstein AA. Restarting anticoagulation therapy after warfarin-associated intracerebral hemorrhage. Arch Neurol 2008;65(10):1313-1318.  [PMID:18852344]

Study Design

Cohort (prospective)


Unknown/not stated


Inpatient (any location) with outpatient follow-up


These researchers identified 88 patients taking warfarin who were admitted to the hospital for ICH associated with international normalized ratios above 1.5. They followed up the patients through review of medical records and mailed surveys. Of the 88 patients, 36 either died in the hospital or were transferred to hospice care. In the remaining 52 patients, 23 resumed warfarin. The mean follow-up interval was 4 years for the patients who restarted warfarin and 3 years for those who didn't. The patients who resumed warfarin were younger (71 years vs 75 years; P = .02) and were more likely to have prosthetic heart valves. Among the 23 who resumed warfarin, 3 had another ICH and 2 had major extracranial hemorrhages (overall major rebleed rate = 21.7%). In the 29 patients who did not resume warfarin, 4 were lost to follow-up. In the remaining 25, 3 had a thromboembolic stroke, 1 had a pulmonary embolism, and 1 had a peripheral artery embolism (overall thromboembolic event rate = 20%). Although we can quibble over how to account for the 4 who dropped out, these data show that no matter what you do, bad things will happen. Since the patients in each group were quite different, it is possible that these differences rather than resumption of warfarin (or not) were important determinants of the outcomes. Only a randomized controlled trial can address this, but I doubt anyone will attempt one.