We read with great interest the article from Qamar and Bhatt (Anticoagulation therapy: Balancing the risks of stroke and bleeding in CKD. Nat. Rev. Nephrol. 11, 200–202; 2015)1 that discussed the observational study by Bonde et al. on anticoagulation of patients with atrial fibrillation and chronic kidney disease (CKD) (Net clinical benefit of antithrombotic therapy in patients with atrial fibrillation and chronic kidney disease: a nationwide observational cohort study. J. Am. Coll. Cardiol. 64, 2471–2482; 2014).2 The original study and the commentary both conclude that anticoagulation with warfarin is safe and reduces the risk of stroke in patients with CKD, and even in dialysis patients. This latter point contrasts with three large registry studies which reported that patients with atrial fibrillation administered warfarin and undergoing haemodialysis have an increased risk of stroke.3,4,5 Furthermore, a Canadian study reported that warfarin use is not beneficial in reducing stroke risk but is associated with a higher risk of bleeding in patients with atrial fibrillation and undergoing dialysis.6

We therefore question what causes the observed differences between these studies (Table 1)? A central limitation of the study by Bonde et al.,2 and a previous study by the same research group,7 is that the cohort of patients who were undergoing renal replacement therapy were exposed to different treatment modalities—some were on haemodialysis, some on peritoneal dialysis, and a considerable subgroup had received a kidney transplant. This cohort, therefore, combined patients with varying degrees of CKD and heparin exposure. We previously raised this concern with the prior study—that this mixture of patient situations renders interpretation of the effects of warfarin in patient subgroups difficult.8

Table 1 Administration of warfarin and risk of stroke in patients with atrial fibrillation and CKD

Uraemia leads to coagulopathy, thrombocytopathy9 and vitamin K deficiency,10 and patients undergoing haemodialysis administered warfarin exhibit an increased risk of bleeding, cardiovascular calcifications, and calciphylaxis. In weighing the risks of vitamin K antagonism in patients undergoing dialysis against the poorly documented benefits, we argue against the use of warfarin in patients with atrial fibrillation undergoing haemodialysis.11