We conducted a retrospective study of all patients hospitalized for ICH at a large stroke center from October 2007 to January 2012. Initial CT scans were used to quantify hematoma volumes using the computer-assisted planimetric analysis. Univariate and multivariable analyses determined the influence of warfarin on hemorrhage location. Median regression analysis was performed to estimate the effects of INR on hematoma volumes.
We included 404 consecutive patients with ICH of whom 69 were on warfarin. Patients on warfarin had larger hematoma volumes (median 23.9 mL vs. 14.2 mL; P = 0.046). In patients excessively anticoagulated with warfarin (defined as INR > 3.0), compared with those in the therapeutic range, brainstem ICH was more frequent (24.0 % vs. 6.1 % ; P = 0.005). Patients with INR > 3.0 had increased odds of infratentorial hemorrhage (OR 3.63; 95 % CI 1.52-8.64; P = 0.004) when compared to non-warfarin ICH patients. After adjustment for hematoma location, there was no significant association between INR and hematoma volume.
Patients with warfarin-associated ICH have a predilection for brainstem ICH. After adjustment for ICH location, no relationship between admission INR and hematoma volume was found.