Warfarin use, mortality, bleeding and stroke in haemodialysis patients with atrial fibrillation

Background Oral anticoagulation therapy (OAT) is the choice treatment for thromboembolism prevention in atrial fibrillation (AF), although data about OAT use in haemodialysis (HD) patients with AF are contradictory. Methods The effect of OAT on the risk of mortality, stroke and bleeding was prospectively evaluated in a population of HD patients with AF. All the patients of 10 HD Italian centres alive on 31 October 2010 with documented AF episode(s) were recruited and followed-up for 2 years. OAT and antiplatelet intake, age, dialytic age, comorbidities and percentage time in the target international normalized ratio (INR) range (target therapeutic range; TTR) were considered as predictors of hazard of death, thromboembolic and bleeding events. Results At recruitment, 134 patients out of 290 were taking OAT. During the follow-up, 115 patients died (4 strokes, 3 haemorrhagic and 1 thromboembolic). Antiplatelet therapy, but not OAT, was associated with increased mortality (HR 1.71, CI 1.10-2.64, P = 0.02). The estimated survival of patients always taking OAT tended to be higher than that of patients who stopped taking (68.6 versus 49.6%, P = 0.07). OAT was not correlated to a significant decreased risk of thromboembolic events (HR 0.12, CI 0.00-3.59, P = 0.20), while it was associated with an increased risk of bleeding (HR 3.96, CI 1.15-13.68, P = 0.03). Higher TTR was associated with a reduced bleeding risk (HR 0.09, CI 0.01-0.76, P = 0.03), while previous haemorrhagic events were associated with higher haemorrhagic risk (HR 2.17, CI 1.09-4.35, P = 0.03). Conclusions In our population of HD patients with AF, the mortality is very high. OAT is not associated with increased mortality, while antiplatelet drugs are. OAT seems, on the contrary, associated with a better survival; however, it does not decrease the incidence of ischaemic stroke, whereas it increases the incidence of bleeding. Bleeding risk is lower in subjects in whom the INR is kept within the therapeutic range.