INR variability and outcomes in patients with mechanical heart valve prosthesis

Background: The quality of treatment with warfarin is mainly assessed by the time in therapeutic range (TTR) in patients with mechanical heart valve prosthesis (MHV). Our aim was to evaluate if International Normalized Ratio (INR) variability predicted a combined endpoint of thromboembolism, major bleeding and death better than TTR. Methods and results: We included 394 patients at one center with MHV during 2008-2011 with adverse events and death followed prospectively. TTR 2.0-4.0 and log-transformed INR variability was calculated for all patients. In order to make comparisons between the measures, the gradient of the risk per one standard deviation (SD) was assessed. INR variability performed equal as TTR 2.0-4.0 per one SD unit adjusted for covariates, hazard ratio (HR) 1.30 (95% CI 1.1-1.5) and 0.71 (95% CI 0.6-0.8) respectively for the combined endpoint, and performed better for mortality HR 1.47 (95% CI 1.1-1.9) and 0.70 (95% CI 0.6-0.8). INR variability was categorized into high and low group and TTR into tertiles. High variability within the low and high TTR, had a HR 2.0 (95% CI 1.7-3.6) and 2.2 (95% CI 1.1-4.1) respectively, of the combined endpoint compared to the low variability/high TTR group. INR values <2.0 greatly increased the rate of thromboembolism whereas the rate of major bleeding increased moderately between INR 3.0 and 4.0 and increased substantially after INR >4.0. Conclusion: The INR variability is an equal predictor as TTR of the combined endpoint of thromboembolism, major bleeding and death, and adds important information on top of TTR in patients with MHV. (C) 2015 Elsevier Ltd. All rights reserved.