Risk factors of vitamin K antagonist overcoagulation

Abstract

Objectives: The aims of this prospective study were to identify, in vitamin K antagonist (VKA)-treated patients, factors associated with INR values: (i) greater than 6.0. and (ii) ranging from 4.0 to 6.0 complicated with bleeding. We also assessed VKA-related morbidity in these patients. Methods: During a 6-month period, 3090 consecutive patients were referred to our Department of Internal Medicine, including 412 VKA-treated patients. At admission, the medical records of VKA-treated patients were reviewed for type, duration and indication of VKA therapy, previous medical history of VKA-related hemorrhage, comorbidities and concomitant medications. Results: Forty of the 412 VKA-treated patients (9.7%) exhibited oral anticoagulant related overcoagulation. VKA overcoagulation was associated with high morbidity, leading to major bleeding in 27.5% of cases; moreover, 12.5% of these patients died, death being mainly due to major bleeding. Under multivariate analysis, significant factors for VKA-related overcoagulation were as follows: previous medical history of VKA therapy-related hemorrhage (P = 0.00001) and INR levels over therapeutic range (P = 0.0006), chronic liver disease (P = 0.03), therapy with amiodarone (P = 0.009); in contrast, statin therapy was found to be a protective factor of VKA overcoagulation (P = 0.008). Conclusion: The knowledge of predictive factors of VKA-related overcoagulation seems of utmost importance to improve patients' management. Our study underlines the fact that the potential of drug interaction should be taken into account when choosing amiodarone for patients receiving VKAs. Interestingly, long-term (>6 month) statin therapy may be a protective factor of VKA overcoagulation. Our findings, therefore, suggest that there may be no need to switch long-term users of VKA and statin to a safer alternative therapy. © The Author 2011. Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved.