Seven-Year Follow up of An Inception Cohort Study of Patients with Atrial Fibrillation Starting Oral Anticoagulation: Aging as An Independent Bleeding Risk

Introduction: Our objectives were to investigate if aging is an independent risk factor for major bleeding during oral anticoagulant treatment (OAT) for atrial fibrillation (AF).

Patients and Methods: An inception cohort study was designed to include patients (pts) referred to our Department to initiate OAT due to AF from July 2000 to July 2007. Inclusion criteria: no prior OAT, AF being the only reason to OAT, follow-up at our Institution, and signed informed consent. Exclusion criteria: concomitant reason for OAT, and pts refusal to participate. Pts were followed until OAT was stopped by the patient’s cardiologist or until July 2007. Patients who underwent a successful electric cardioversion and were followed up only for 4 weeks were not included unless they presented a bleeding episode. Epidemiological and clinical data were documented for all patients. Major bleeding episodes, age, time on OAT, and INR at the bleeding event were recorded. Deaths were registered and classified as “related” or “not related” to OAT. For the analysis, the inception cohort was divided into three groups according to the age of beginning and ending of OAT. Group 1 included patients younger than 77 at the end of OAT. Group 2 was constituted by the patients that initiated therapy before 80 and were followed after that age. Patients starting after 80 years of age were the group 3. Time in therapeutic range and INR variability were calculated. Incidence rate and relative risk (RR) of major hemorrhage (MH) was estimated.

Results: During the study period, 2.971 patients were referred for initiation of OAT, 40% (1.172) had AF, and 671 met the exclusion criteria (138 patients had a prior OAT, 111 were not going to go on control OAT at the Institution, and 422 had a concomitant reason for OAT (mitral estenosis, heart valve prostheses, venous thrombosis, coronary artery by-pass, etc.) Five-hundred and one patients were included in the study, 68 (13.5%) were lost for follow-up after inclusion or had less than 4 weeks after initiation, and 433 had complete follow-up. Patients’ characteristics and MH are shown in Table 1. Time within therapeutic range was similar in all age-groups (65.0%, 62.8%, and 65.4% for group 1, 2 and 3, respectively). There were 28 MH in 23 pts during the study period. In group 3, the rate of MH was 3.39/100 patient-years while in group 1 it was 0.85/100 patient-years. The RR was 3.0 (1.2–7.6 95%CI) in group 3 compared to group 1(p= 0.02). Among patients in group 2, the RR of bleeding increased to 5.8 (2.27–14.95) after the age of 80. The likelihood ratio of MH with increasing age was 14.1 (p= 0.001). INR variability tended to be higher in Group 3 and in bleeding patients regardless their age. Aspirin, number of concomitant medications and CHADS2 score were not associated with the risk of MH. There were 10 deaths, all in pts older than 80. In 9/10 the cause of death was documented and was unrelated to OAT. One patient died after a non-characterized stroke, it was considered hemorrhagic. Rate of death related to bleeding after age of 80 (groups 2 and 3) was 0.54%.

Conclusions: This long-term follow up cohort study shows that aging is an independent risk factor for MH; this observation might have clinical implications for the management of elderly pts.

Table 1: Patients’ characteristics