Atrial fibrillation (AF) is a common condition whose prevalence increases with advancing age . One of the main complications of this arrhythmia is the occurrence of systemic thromboembolism, the first of which the occurrence of cerebrovascular accident (CVA) . The advanced age itself is an independent risk factor for occurrence of stroke in the context of AF , evidenced by the score awarded to the age over 75 years in the risk estimate thromboembolism in the score CHA2DS2-Vasc . Making the elderly a priority target for the prevention of thromboembolism, both because of the increased thromboembolic risk and prevalence of AF in this population. For many years, vitamin K antagonists (VKA) have proven their effectiveness in this context . But the effectiveness and safety of thromboembolism prevention by anti-vitamin K treatment involves maintaining the INR (International Normalized Ratio) in the therapeutic range. Indeed occurred of a thromboembolic or haemorrhagic adverse event directly correlates to having an INR outside the therapeutic range and maintaining the INR in the therapeutic range is subject to many variations inter- and intra-individual. They include, genetic factors , food , drug interactions , as well as disease processes . Or one of the elderly characteristics is the coexistence of multiple pathological processes or comorbidities . The supposed interaction between comorbidities and variability of INR has been little studied: Only a few studies have examined the factors associated with the stability of the INR and unstable . The identification of patients at risk of instability INR yet allow clinicians greater vigilance and better identification of patients for whom the initiation of treatment with vitamin K does not appear the best therapeutic strategy.

Our study seeks to determine whether comorbidity, assessed by the Charlson comorbidity index, are associated with instability of the INR, among more than 80 years for patients treated with AVK atrial fibrillation.

Inclusion Criteria:

• Patients aged 80 years or more;
• Patients present in geriatric structure June 21, 2011 ( short geriatric , SSR , LTC or nursing homes ) ;
• Patients with AVK June 21, 2011 or having received warfarin in the previous 7 days ;
• Patients with AVK under thromboembolism prevention in atrial fibrillation .

Exclusion Criteria:

• - Patients receiving AVK for another reason thromboembolism prevention for atrial fibrillation : treatment of venous thromboembolism , pulmonary embolism, mechanical heart valve , etc

The study focused on 2,164 patients living in a geriatric structure (short stay , SSR , LTC or nursing homes ) with / who had warfarin for prevention of thromboembolism in the context of atrial fibrillation .