By Michael H. Crawford, MD

Professor of Medicine, Associate Chief for Education, Division of Cardiology, University of California, San Francisco

Dr. Crawford reports no financial relationships relevant to this field of study.

SYNOPSIS: A large outpatient observational study of patients with atrial fibrillation and chronic kidney disease who were anticoagulated revealed that, compared to warfarin, direct oral anticoagulants exhibited less all-cause mortality and major bleeding with at least equivalent efficacy at preventing stroke.

SOURCE: Makani A, et al. Safety and efficacy of direct oral anticoagulants versus warfarin in patients with chronic kidney disease and atrial fibrillation. Am J Cardiol 2020;125:210-214.

Although atrial fibrillation (AF) is common in patients with chronic kidney (CKD), the relative safety and efficacy of warfarin vs. direct oral anticoagulants (DOACs) is unclear. The seminal trials of DOACs in AF excluded patients with advanced renal disease. However, DOACs were approved for use in CKD patients based on small pharmacokinetic studies. To clarify this issue, Makani et al studied patients seen in the University of Pittsburgh clinics between 2010 and 2017 who received a diagnosis of nonvalvular AF and a CHA2DS2-VASc score of ≥ 2 who were treated with anticoagulants. They were stratified into three groups based on their estimated glomerular filtration rate (GFR): > 60, 30-60, or lower than 30 mL/min. The primary endpoints were all-cause mortality, major bleeding, and stroke. Among the 21,733 patients included, 10,794 were on DOACs and 10,939 on warfarin. Baseline characteristics were similar between the two treatment arms. There was no difference in the two arms in the distribution of the three CKD groups.

During a mean follow-up of 3.4 years, the adjusted risk of death for all three CKD groups was lower in the DOAC arm: GFR > 60 mL/min (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.70-0.84; P < 0.001), GFR 30-60 mL/min (HR, 0.74; 95% CI, 0.68-0.81; P < 0.001), and GFR < 30 mL/min (HR, 0.76; 95% CI, 0.63-0.93; P = 0.005). Major bleeding also occurred much less in the DOAC arm: GFR > 60 mL/min (HR, 0.93), GFR 30-60 mL/min (HR, 0.83), and GFR < 30 mL/min (HR, 0.69). Embolic stroke rates were lower in the DOAC arm: GFR > 60 mL/min (HR, 0.86), GFR 30-60 mL/min (HR, 0.87), and GFR < 30 mL/min (HR, 0.60), as was hemorrhagic stroke: GFR > 60 mL/min (HR, 0.58), GFR 30-60 mL/min (HR, 0.41), and GFR lower than 30 mL/min (HR, 0.55). However, only the rates of hemorrhagic stroke in the two groups with GFR > 30 mL/min were statistically significant. Also, patients who suffered a stroke were more likely to die.

The authors concluded that, compared to warfarin, DOACs were equivalent regarding stroke prevention, with lower all-cause mortality rates and fewer major bleeds in nonvalvular AF patients at all levels of CKD.


Current guidelines state that in AF patients with advanced CKD or who are on dialysis, treatment with warfarin or a DOAC might be reasonable (IIb, B-NR). Although such patients were excluded from large trials, in those patients who developed CKD during those studies, there were no safety issues with DOACs. This large, nonrandomized study confirms these observations and the conclusions of pharmacokinetic studies. This investigation reveals that the efficacy is at least no different from warfarin and perhaps better, especially if hemorrhagic stroke is considered. Also, there was a trend for lower embolic stroke rates with DOACs, which was not statistically significant.

There were several limitations to this study. It was a large, retrospective database study with little detail. The authors chose not to evaluate cardiovascular mortality due to the difficulty of adjudicating cause of death in this type of observational outpatient study. We do not know the causes of death, as the study data lacked this granularity. Also, compliance with medications and international normalized ratio levels on warfarin are unknown. In addition, there are no data on antiplatelet therapy because the authors could not determine the duration of therapy. It would have been useful to know the type and doses of the DOACs used. Presumably, Makani et al followed the manufacturer’s recommended doses for patients with various levels of CKD. Finally, although included in the study, a separate analysis of hemodialysis patients would have been of interest. Those with GFR < 30 mL/min made up 7% of the study population. Perhaps there were too few dialysis patients to analyze. With these new data, it seems reasonable to consider DOACs rather than warfarin in AF patients with CKD, especially if their CHA2DS2-VASc score is ≥ 2. This is a situation in which shared decision-making with the patient is important because of the risks involved with therapy and the paucity of data.