Skip to content
View commentaries on this research

This is a plain English summary of an original research article

Newer generation anticoagulants appear to reduce the risk of bleeding compared with older anticoagulants like warfarin, when used to prevent clots in people who also have mild to moderate chronic kidney disease. This group of drugs, called direct oral anticoagulants, have been well researched in healthy people with atrial fibrillation or at risk of thromboembolism but this was the first review to look at their safety, in terms of bleeding and risk of bleeding within the brain, in people with kidney disease.

As a group, these newer drugs appear safer than older anticoagulants such as warfarin.  Evidence from the trials comparing individual new anticoagulants to warfarin was  reanalysed to look at differences between the new anticoagulants too. The analysis suggests that apixaban may pose the least risk of bleeding, particularly in mild kidney disease. The findings support guidance that apixaban is a reasonable choice as an anticoagulant for people with kidney disease and may be better than dabigatran, rivaroxaban, edoxaban or traditional anticoagulant treatment with warfarin.

Why was this study needed?

Atrial fibrillation is one of the most common heart conditions, affecting about 1.6% of the population of England and Wales and increasing with age. It carries increased risk of stroke. Nearly three quarters of people diagnosed with atrial fibrillation take anticoagulant medication – most commonly warfarin – to reduce that risk.

New generation anticoagulants (apixaban, dabigatran, rivaroxaban and edoxaban) are known to be effective at preventing blood clots in veins and they help to prevent strokes in people with atrial fibrillation. Most people have less bleeding-related side effects using them than using older vitamin K-blocking anticoagulant drugs like warfarin.

All anticoagulant treatments carry some risk of bleeding. In 2013, the UK’s Medicines and Healthcare products Regulatory Agency warned that having kidney disease may be a reason to avoid using the newer anticoagulants or reduce their dose.

Evidence on the safety of these drugs in people with kidney disease had not been reviewed before. Many people with kidney disease are excluded from this type of study because of concerns about serious side effects. This review addressed the evidence gap in this group.

What did this study do?

This systematic review looked for published and unpublished studies and included nine randomised controlled trials comparing the safety of new drugs with older vitamin K blocking anticoagulants (mostly warfarin) in 54,667 people with mild or moderate chronic kidney disease. Safety elements included risk of major bleeding or stroke due to bleeding in the brain.

Patients taking the drugs were followed up for between three months and nearly three years. Results were grouped by mild or moderate impairment of kidney function, as measured by a creatinine clearance test (see Definitions).

The review was carried out to a high standard, following PRISMA guidelines. The overall risk of bias was low in the included studies so we can be confident in the findings.

What did it find?

  • Overall, the four new anticoagulants were linked to 13% less risk of major bleeding (relative risk [RR] 0.87; 95% confidence interval [CI] 0.81 to 0.93) for participants with mild kidney disease compared with vitamin K blocking anticoagulants. Subanaysis showed this was significant only for apixaban and edoxaban and not for rivaroxaban or dabigatran.
  • Apixaban seemed more effective in reducing major bleeding in participants with moderate kidney disease, lowering risk by 48% (RR 0.52; 95% CI 0.40 to 0.68). Edoxaban reduced bleeding. The others drugs didn’t reduce the risk and when all four drugs were pooled, there was no overall risk reduction.
  • Newer anticoagulants were associated with a decreased risk for stroke due to a bleed compared to Vitamin K blocking anticoagulants both in patients with mild and moderate kidney disease (RR 0.43; 95% CI 0.34 to 0.56; and RR 0.42; 95% CI 0.30 to 0.61 respectively).
  • Indirect comparisons of the risk of major bleeding between different newer anticoagulants showed apixaban was the best performer in people with moderate kidney disease. All newer anticoagulants performed about the same in people with mild kidney disease.

What does current guidance say on this issue?

NICE guidance on chronic kidney disease from 2014 recommends apixaban rather than warfarin if oral anticoagulant treatment is needed for people with moderately impaired kidney function (measured by glomerular filtration rate of 50 ml/min/1.73 m2). The guidance talks about research findings showing apixaban has greater effectiveness at preventing clots in veins, or for reducing stroke risk, in people who also have atrial fibrillation not involving heart valves, compared with warfarin.

Twelve separate NICE technology appraisals on the individual newer anticoagulants mention the need to be cautious if prescribing anticoagulants to people with reduced kidney function, and to consider reducing the standard dose to minimise risk of bleeding.

What are the implications?

The findings of this systematic review support guidance showing apixaban is a reasonable choice as an anticoagulant for people with kidney disease compared to dabigatran, rivaroxaban, edoxaban or traditional anticoagulant treatment such as warfarin.

The authors acknowledge that their findings might be the result of suboptimal usage of warfarin in these trials, so some caution is still required. The comparative safety and effectiveness of the different drugs compared to each other and to ideal warfarin treatment is not yet reliably established as these were built on indirect comparisons in this review. Direct comparisons and long term follow up studies will be needed to confirm the slight advantages of apixaban over other oral anticoagulants.

This study did not examine costs or cost-effectiveness. Newer oral anticoagulants are substantially more expensive than warfarin although this is offset somewhat as there is no need for the regular tests to check blood clotting when warfarin is used.

Citation and Funding

Raccah BH, Perlman A, Danenberg HD, et al. Major bleeding and hemorrhagic stroke with direct oral anticoagulants in patients with renal failure: systematic review and meta-analysis of randomized trials. Chest.2016. [Epub ahead of print].

No external funding was used for this study.



NHS Choices. Chronic kidney disease. Leeds: NHS Choices; 2014.

NICE. Non-vitamin K antagonist oral anticoagulants (NOACs). NICE advice KTT16. London: National Institute for Health and Care Excellence; 2016.

NICE. Chronic kidney disease in adults: assessment and management. CG182. London: National Institute for Health and Care Excellence; 2014.

NICE. Atrial fibrillation: management. CG180. London: National Institute for Health and Care Excellence; 2014.

Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre


  • Share via:
  • Print article


In this review, the degree of kidney function impairment was defined by estimated creatinine clearance ranges used in the included studies:

  • Estimated creatinine clearance 50-80 mL/min, which approximates to mild kidney disease.
  • Estimated creatinine clearance <50mL/min, which approximates to moderate kidney disease.

Normal reference ranges for estimated creatinine clearance vary, but are typically as follows: men (younger than age 40): 107-139 mL/min, women (younger than age 40): 87-107 mL/min. The ranges decrease with age by 6.5ml/min per ten years.

Creatinine clearance estimations use urine and blood tests to estimate the rate at which a breakdown product of muscle is excreted from the body. In people with a normal adult surface area, the results are similar to another, more commonly used measure of kidney function, the estimated glomerular filtration rate.


Back to top