Risk factors for bleeding during anticoagulation of atrial fibrillation in older and younger patients in clinical practice.

BACKGROUND: The prevalence of atrial fibrillation increases with age, affecting approximately 5% of people aged >65 years and almost 10% of people aged >80 years.

OBJECTIVE: The goal of this study was to identify risk factors for bleeding during warfarin treatment of nonvalvular atrial fibrillation (NNVAF) in older patients (those aged >or=75 years) compared with younger patients (those aged <75 years) in clinical practice.

METHODS: All patients with NVAF newly started on warfarin at an anticoagulation clinic in a large university hospital were included in this prospective observational study. Patient details were recorded at their first visit; details of any bleeding events were recorded via telephone interview every 4 to 6 weeks for a minimum of 10 months. Patients were divided into 2 groups (ie, those >or=75 years old and those <75 years old). Logistic regression analysis was used to identify risk factors for bleeding.

RESULTS: A total of 402 patients were included in the study. Group I comprised 203 patients <75 years old (mean [SD] age, 64.33 [8.33] years) and group II comprised 199 patients >or=75 years old (mean [SD] age, 80.44 [3.99] years). Follow-up ranged from 1 to 31 months (mean [SD], 19 [8.11] months). For major bleeding, number of medications was a significant risk factor in older patients (odds ratio [OR], 3.0; 95% CI, 1.2-7.8 [P = 0.02 ]) and range of the international normalized ratio (INR) was a significant risk factor in both groups. For every unit increase in the range of INR, the odds of major bleeding increased by 0.6 (OR, 1.6; 95% CI, 1.2-2.4 [P = 0.03 ]) in younger patients and by 0.4 (OR, 1.4; 95% CI, 1.07-1.99 [P = 0.04 ])in older patients. For minor bleeding, history of hypertension was the only significant risk factor in older patients (OR, 3.3; 95% CI, 1.3-8.1 [P = 0.01 ]), while history of ischemic heart disease was the only risk factor in younger patients (OR, 1.9; 95% CI, 1.1-5.4 [P = 0.04 ]).

CONCLUSIONS: Bleeding pattern was similar in both age groups regarding severity, onset, anatomic site of bleeding, and INR values during the bleeding event. Risk factors for episodes of major bleeding, which are more of a clinical concern, are potentially modifiable. They include quality of anticoagulation control in both groups and number of medications in the older age group.