Atrial fibrillation increases the risk of ischaemic stroke.
Oral anticoagulants dramatically reduce this risk.
The benefits of anticoagulants outweigh the risk of bleeding for most people – except where risk of stroke is very low.
Stroke prevention and bleeding risk can be estimated using the tools below.
Use the CHA2DS2-VASc score to estimate the risk of someone with AF having an ischaemic stroke, TIA or thrombus in 1 year.
NICE recommends anticoagulation for those with a CHA2DS2-VASc score of 2 or more (and consider it for men with a score of 1 or more).
Oral anticoagulants reduce stroke risk by approximately two-thirds.
Data detail
The data below comes from trials of warfarin v placebo.
These figures are representative enough to apply to all oral anticoagulants for practical purposes.
Figures derived from the 2014 NICE guideline1 (and are current for the 2021 guideline).
NICE suggest calculating the baseline risk of stroke as per the mdcalc.com website.
References
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2014 [Internet]. [London]: NICE; 2014. (Clinical guideline [CG180])
2)Leif Friberg, Mårten Rosenqvist, Gregory Y.H. Lip, Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: the Swedish Atrial Fibrillation cohort study, European Heart Journal, Volume 33, Issue 12, June 2012.
| This research provides a very good indication of the treatment effect.
However, it is possible that the true effect is slightly smaller or greater. |
The evidence that anticoagulation reduces the risk of stroke is rated by NICE as HIGH quality1.
NICE reviewed risk scoring systems in 2021 and concluded CHA2DS2-VASc had the best sensitivity/specificity profile1.
However, the quality of evidence for baseline stroke risks calculated by CHA2DS2-VASc is LOW quality, due to observational study design2. There is likely to be some imprecision in these risk estimates.
| This research provides some indication of risk.
There is a high possibility that the true risk is smaller or greater. |
References
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2014 [Internet]. [London]: NICE; 2014. (Clinical guideline [CG180])
2) Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study.
3)Guidelines: 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. European Heart Journal 2012; 33(21): 2719–2747
The population characteristics in the warfarin v placebo trials were (roughly)1,2:
The population characteristics for the validation study3 of CHA2DS2-VASc were:
References
1)Aguilar MI, Hart R. Oral anticoagulants for preventing stroke in patients with non‐valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks. Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD001927
2)Saxena R, Koudstaal PJ. Anticoagulants for preventing stroke in patients with nonrheumatic atrial fibrillation and a history of stroke or transient ischaemic attack. Cochrane Database of Systematic Reviews 2004, Issue 2. Art. No.: CD000185
3) Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study
The 2021 NICE guideline1 recommends:
The evidence for this is LOW quality overall.
| This research provides some indication of the treatment effect.
There is a high possibility that the true effect is smaller or greater. |
Evidence summary
more
Overall, NICE evidence review concluded that DOACs as a group are likely to be superior to warfarin. Even though some evidence suggested benefits of one DOAC over another there was enough uncertainty to acknowledge that all DOACs might be equally safe and effective.
Key points from the NICE evidence review:
References
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2021 [Internet]. [London]: NICE; 2021. (NICE guideline [NG 196])
2) Oral anticoagulants for prevention of stroke in atrial fibrillation: systematic review, network meta-analysis, and cost effectiveness analysis.
The risk of intracranial haemorrhage (ICH) with anticoagulants is small and outweighed by the benefit of prevention of ischaemic strokes.
The exact excess risk is uncertain, but is likely to be a fraction of a percentage per year.
Data detail
ICH rates in patients taking DOACs in clinical trials ranged from1,2
0.2% to 0.6% per year
ICH rates in patients taking warfarin in the same trials were at least twice as high3
These trials lack a no-anticoagulation control group, but for comparison:
ICH rates for people with AF not taking anticoagulants in observational studies2 ranged from:
0.1% to 0.8% per year
References
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2021 [Internet]. [London]: NICE; 2021. (NICE guideline [NG 196])
2)Angelozzi A, Renda G, Mercuri M et al. The Risk of Intracranial Hemorrhage with Anticoagulation in the Elderly – Estimates of Prevalence and Therapeutic Strategies. Amer Coll Cardiol 2015 online (accessed 02/22)
3) Oral anticoagulants for prevention of stroke in atrial fibrillation: systematic review, network meta-analysis, and cost effectiveness analysis
Many patients have a dual indication for antiplatelets and oral anticoagulants.
Current evidence is not clear enough to provide good estimates of benefit or bleeding risk for the variety of situations we may encounter.
For patients with atrial fibrillation and stable ischaemic heart disease (say, an MI more than one year ago):
For patients with more complex needs, (for example after coronary artery stent insertion, who may need dual anti-platelet therapy in addition to an oral anticoagulant), the evidence is complex:
References
1)Hindricks G, Potpara T, Dagres N et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC European Heart Journal 2021 Feb 1; 42(5): 373-498
2)Lemesle G. Aspirin on Top of Anticoagulation in Patients With Concomitant Stable Coronary Artery Disease and Atrial Fibrillation. Circulation 2019; 139: 617-619
3)National Institute for Health and Care Excellence. Acute Coronary Syndromes 2020 [Internet]. [London]: NICE; 2020 (Clinical guideline [CG185])
4)MASTER DAPT Investigators. Abbreviated Antiplatelet Therapy in Patients at High Bleeding Risk With or Without Oral Anticoagulant Therapy After Coronary Stenting: An Open-Label, Randomized, Controlled Trial. Circulation 2021 Oct 12; 144(15): 1196-1211
Bleeding risk is discussed in the section below.
The ORBIT score estimates risk of someone with AF having a significant bleed whilst on anticoagulants.
Significant bleed = bleeding causing a fall in Hb ≥2g/dL, or symptomatic bleeding in a critical area, or requiring transfusion of ≥2 units of blood, or fatal bleeding.
Unfortunately, there are no direct comparison data for bleeding risk in those with AF who do not take anticoagulants.
To help understand bleeding risk without anticoagulants, see here:
More
This table shows some figures for:
| Low risk primary care population | Number of significant bleeds per 100 people per year |
| Men age 70-79 | 0.5 |
| Men age 60-69 | 0.36 |
| Men age 50-59 | 0.23 |
| Women age 70-79 | 0.42 |
| Women age 60-69 | 0.29 |
| Women age 50-59 | 0.21 |
| Placebo groups in AF warfarin trials | ~1 |
Prediction of significant bleeding with anticoagulation1 :
The 2021 NICE guideline evidence review judged the ORBIT score to have the best calibration data.
Estimates of bleeding risk without anticoagulation were derived from:
References
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2021 [Internet]. [London]: NICE; 2021. (NICE guideline [NG 196])
2)Selak V, Kerr A, Poppe K et al. Annual risk of major bleeding among persons without cardiovascular disease not receiving antiplatelet therapy. JAMA 2018; 319(24): 2507–2520
3)Aguilar MI, Hart R. Oral anticoagulants for preventing stroke in patients with non‐valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks. Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD001927.
| This research provides some indication of risk.
There is a high possibility that the true risk is smaller or greater. |
NICE rated the evidence for all bleeding risk tools as LOW quality due to:
Reference
1)National Institute for Health and Care Excellence. Atrial fibrillation: the management of atrial fibrillation 2021 [Internet]. [London]: NICE; 2021. (NICE guideline [NG 196])
People who have an adverse event
People whose adverse event is prevented by treatment
People who were never going to have an adverse event anyway
Graphics and NNTs are rounded to the nearest integer
These three statistical terms offer three different ways of looking at the results of trial data.
Absolute Risk Reduction
This tells you how many people out of 100 who take a treatment have an adverse event prevented.
MoreThe value of the ARR changes with the baseline risk of the person (or population) taking the treatment. The higher the starting risk, the greater the absolute chance of benefit.
You need to think about over what time the trial data show this benefit, as it is usually assumed that more absolute risk reduction is gained over time.
Your patient might be taking the treatment for much longer than the length of a clinical trial (or, if life expectancy is limited, perhaps for less time).
Number Needed to Treat
This tells you how many people need to take the treatment in order for one person to avoid an adverse event.
The lower the number, the more effective the treatment.
MoreThe value of the NNT changes with the baseline risk of the person (or population) taking the treatment. The higher the starting risk, the smaller the NNT.
You need to think about over what time the trial data show benefit, as it is usually assumed that more benefit is gained over time and therefore the NNT will drop over time.
Your patient might be taking the treatment for much longer than the length of a clinical trial (or, if life expectancy is limited, perhaps for less time).
Relative Risk Reduction
This tells you the proportion of adverse events that are avoided if the entire population at risk is treated.
MoreThe value of the RRR is usually constant in people (or populations) at varying degrees of risk.
It is also usually assumed to stay constant over time.
This can be helpful, especially when thinking about population outcomes, but can be misleading for an individual person:
For example, a RRR of 25% in someone with a baseline risk of 40% would give them an ARR of 10% and an NNT of 10.
A RRR of 25% in someone with a baseline risk of 4% would give them an ARR or 1% and an NNT of 100.
This website is designed for use by General Practitioners and other healthcare professionals. The content is not exhaustive and assumes a standard level of GP professional knowledge. The information here is intended to support clinical judgement and shared decision making alongside clinical guidelines and standard practice.
If you are a patient/member of the public, do feel free to look around, but please don’t make any changes to your treatments based on information here. If you find something which seems relevant to you, you could show this website to your healthcare professional to help a discussion.