You are here

Stroke prevention in atrial fibrillation



Atrial fibrillation is found in a third of all ischaemic strokes, even more after post-stroke atrial fibrillation monitoring. Data from stroke registries show that both unknown and untreated or under treated atrial fibrillation is responsible for most of these strokes, which are often fatal or debilitating. Most could be prevented if efforts were directed towards detection of atrial fibrillation before stroke occurs, through screening or case finding, and treatment of all patients with atrial fibrillation at increased risk of stroke with well-controlled vitamin K antagonists or non-vitamin K antagonist anticoagulants. The default strategy should be to offer anticoagulant thromboprophylaxis to all patients with atrial fibrillation unless defined as truly low risk by simple validated risk scores, such as CHA2DS2-VASc. Assessment of bleeding risk using the HAS-BLED score should focus attention on reversible bleeding risk factors. Finally, patients need support from physicians and various other sources to start anticoagulant treatment and to ensure adherence to and persistence with treatment in the long term.

This is the first in a Series of three papers about atrial fibrillation


Ischaemic strokes related to atrial fibrillation usually result from cardioembolism of a large cerebral artery, and therefore tend to be larger (figure 1) and more frequently fatal or associated with greater disability than strokes from other causes.1, 2, and 3 However, strokes related to atrial fibrillation are largely preventable, because oral anticoagulants (OACs) are so effective. In meta-analyses,4 and 5 vitamin K antagonists (VKAs; eg, warfarin) reduced stroke or systemic thromboembolism by 64% and all-cause mortality by 26% compared with placebo (five studies) or untreated controls (one study); the use of non-VKA OACs (NOACs) offers additional significant reductions of 19% and 10%, respectively, relative to warfarin.4 and 5


Figure 1

Imaging of large middle cerebral artery cardioembolic stroke in a patient with atrial fibrillation

(A) CT and (B) subsequent MRI from a 39-year-old man with large middle cerebral artery cardioembolic stroke.


Several steps are needed to reduce the stroke burden associated with atrial fibrillation. The first is recognition of the risk of stroke in patients with atrial fibrillation, followed by risk assessment using simple risk scores such as CHA2DS2-VASc, and prescription of appropriate stroke prevention to all who are not at low risk of stroke. Second, a system is needed to recognise the pre-symptomatic phase of atrial fibrillation rather than wait for stroke to be the first clinical manifestation. Finally, measures are needed to achieve optimum treatment, including excellent international normalised ratio (INR) control if VKAs are used, excellent adherence to thromboprophylactic drugs (ie, VKAs or NOACs) as prescribed, and long-term persistence with treatment. In this paper, we provide an overview of all three aspects of stroke prevention in atrial fibrillation, in the hope that greater awareness will result in reduction of the overall ischaemic stroke burden associated with atrial fibrillation.

Atrial fibrillation as a cause of ischaemic stroke

Of all strokes with an established cause, over 85% are ischaemic strokes,6 and the association of atrial fibrillation with ischaemic stroke of cardioembolic origin is well recognised.7 Indeed, findings from recent population-based studies or stroke registries8, 9, 10, 11, and 12 consistently showed a substantial atrial-fibrillation-attributable risk of stroke, especially in the elderly; at least one in three to four patients with an ischaemic stroke, and over 80% of those with ischaemic stroke of cardioembolic type, also had atrial fibrillation (appendix p 5), suggesting an even stronger association of atrial fibrillation with stroke than previously thought. In over 25% of strokes related to atrial fibrillation, the stroke was the first manifestation of previously unknown atrial fibrillation, which in most cases could have been prevented by OAC treatment, had atrial fibrillation been detected before the stroke (appendix p 5).

Sometimes, even extensive post-stroke diagnostic testing does not elucidate the cause of the stroke (ie, large-vessel disease, cardioembolism, or small-vessel disease); such cryptogenic strokes comprise around 25% of all strokes.13 In two randomised trials assessing various post-stroke cardiac rhythm monitoring strategies in patients with cryptogenic stroke,14 and 15 previously unknown atrial fibrillation was eventually detected by prolonged monitoring in 30% of patients in the CRYSTAL-AF trial14 and in 16% of patients with 30-day monitoring in the EMBRACE trial;15 identification of atrial fibrillation after stroke would qualify patients for secondary prevention using OACs instead of the standard non-atrial fibrillation post-stroke treatment, aspirin. In a meta-analysis of 50 studies,16 atrial fibrillation was detected in 24% of patients after stroke by combined short-term and long-term monitoring. The optimum, cost-effective technique and duration of post-stroke monitoring beyond the first 24 h remains uncertain. At present, prolonged post-stroke monitoring is optional.

Replacement of the term cryptogenic stroke with the more explicitly defined embolic stroke of undetermined source (ESUS), a non-lacunar brain infarct without evident proximal arterial stenosis or cardioembolic sources, has been increasingly advocated.17 A substantial proportion of patients with ESUS have paroxysmal atrial fibrillation.17 and 18 The risk of stroke recurrence is high after incident ESUS.17 and 18 Electrocardiographic documentation of atrial fibrillation is mandatory for OAC use in the context of stroke prevention.19 In the CRYSTAL-AF14 and EMBRACE15 post-stroke monitoring trials, at least 75% of atrial fibrillation episodes were asymptomatic, which emphasises the unreliability of symptoms for detection of atrial fibrillation. Two ongoing randomised trials will compare the efficacy and safety of the NOACs dabigatran (RE-SPECT ESUS20) and rivaroxaban (NAVIGATE ESUS; NCT02313909) versus aspirin in unselected patients after ESUS, which might obviate the need for post-stroke monitoring.

Previous management of atrial fibrillation in patients presenting with stroke

Although most strokes related to atrial fibrillation can be prevented using OACs,21 findings from contemporary registry-based and observational real-world reports from various geographical regions have consistently shown that OAC treatment is underused in patients with atrial fibrillation who are at risk of stroke.22 No OAC is used in around a third of eligible patients with atrial fibrillation, and in over 50% of patients who receive warfarin the quality of anticoagulation control remains suboptimum.23

In the Canadian Stroke Registry,24 only 10% of patients with known atrial fibrillation and acute ischaemic stroke were previously well managed on warfarin (an additional 29% were on subtherapeutic warfarin), whereas in secondary prevention (ie, those with a history of stroke) the percentages increased to 18% and 39%, respectively (appendix p 6). In the Adelaide Stroke Incidence Study,10 warfarin had been prescribed before stroke in 27% of patients with previous atrial fibrillation (15% therapeutic and 12% subtherapeutic; appendix pp 3, 4). Findings from a registry of over 94 000 ischaemic strokes from Sweden12 suggested that of all patients with ischaemic stroke, 20% had known but untreated or inadequately treated atrial fibrillation and 9% had previously unknown atrial fibrillation; in these patients stroke could have been prevented by either treatment with an OAC according to current guidelines or by screening for atrial fibrillation (appendix pp 3, 4). In the UK, according to the Sentinel Stroke National Audit Programme (SSNAP), there has been some improvement over previous anticoagulant treatment rates in patients with stroke with known atrial fibrillation over the past 3 years, but antiplatelet drugs, largely aspirin, were still the sole antithrombotic prescribed in 26% of patients in the 12 months ending March, 2016.

Aspirin is still widely misused for primary or even secondary stroke prevention in a quarter to a third of patients with atrial fibrillation who are eligible for OACs,24 and 25 presumably because of misperception of efficacy and safety for stroke prevention in atrial fibrillation, which is likely to contribute to continuing underuse of anticoagulants.26 Aspirin is neither effective nor safe as thromboprophylaxis for atrial fibrillation,26 even possibly increasing stroke risk in elderly patients,27 and has largely been removed from guidelines.19, 28, and 29 The consequence of aspirin misuse is evident in stroke registries (appendix pp 3, 4, 6), with a high proportion of atrial-fibrillation-related strokes occurring in patients treated only with aspirin, despite a CHADS2 or CHA2DS2-VASc score of at least 2 (appendix pp 3, 4). Replacing aspirin with OACs, and prescribing OACs for the 20% of high-risk patients with known atrial fibrillation who receive no OAC treatment (appendix p 6) constitutes a simple solution to reduce the atrial fibrillation stroke burden, provided effective measures to close this evidence–treatment gap are implemented.

Finding unknown atrial fibrillation to prevent stroke

Almost 10% of all ischaemic strokes (representing >25% of strokes related to atrial fibrillation) occur simultaneously with first-detected atrial fibrillation. Measures to screen or case-find unknown asymptomatic atrial fibrillation, and then treat with OACs, should logically have a major effect on reducing stroke burden. The inbuilt assumptions are that unknown asymptomatic atrial fibrillation is common, and that prognosis of unknown asymptomatic atrial fibrillation is similar to that in the pivotal trials,30 which included a small but unknown proportion of patients with incidentally detected atrial fibrillation (eg, during a routine clinic visit). Such pre-symptomatic atrial fibrillation has sometimes been assumed to have a benign prognosis.31 In a study of asymptomatic incidentally detected atrial fibrillation in general practice, prognosis was far from benign, with a doubling of stroke and all-cause mortality compared with age-matched and sex-matched patients without atrial fibrillation.32 Moreover, anticoagulant treatment reduced the stroke rate from 4% to 1% after only 1·5 years compared with no treatment; this rate is almost identical to that in matched controls without atrial fibrillation seen contemporaneously (figure 2).21 A similar adverse prognosis of atrial fibrillation first discovered in the absence of symptoms was noted in Olmsted county (MN, USA).33 and 34 In the EORP-AF registry,35 patients who had never experienced symptoms actually had a worse prognosis than those with symptoms.


Figure 2

Effect of treatment on incidentally detected atrial fibrillation

AF=atrial fibrillation. OAC=oral anticoagulant. Reproduced with permission from Freedman and colleagues.21


Patients with pacemakers or similar implanted devices frequently have brief or even prolonged episodes of asymptomatic atrial fibrillation, and these have been associated with a more than doubling of the stroke risk, and a 5·5 times increase in the risk of subsequent atrial fibrillation.36 and 37 Longer (>18 h) episodes of atrial fibrillation have the highest adverse prognosis.36 Ongoing randomised trials of anticoagulant treatment of device-detected asymptomatic atrial fibrillation (eg, ARTESiA [NCT01938248] and NOAH [NCT02618577]) are investigating the role of two NOACs, apixaban or edoxaban, versus aspirin for the treatment of device-detected subclinical atrial fibrillation. Even excessive atrial ectopics or runs (defined as ≥20 beats) of atrial tachycardia without definitive atrial fibrillation carry a similar prognosis.38 This finding, coupled with the only partial temporal association of stroke with device-detected atrial fibrillation episodes,39, 40, and 41 suggests that as well as being a risk factor for stroke, atrial fibrillation is also a powerful risk marker for an abnormal atrial or systemic substrate, which can lead to stroke.42

Screening or case finding in either the clinic or community will detect atrial fibrillation in 1·4% of patients on a single screen in those aged at least 65 years.43 Using stepped screening with patient-activated handheld electrocardiograph (ECG) devices for 2 weeks in 75–76-year-olds detects atrial fibrillation in 3% of patients.44 Opportunistic case finding using pulse palpation and ECG if irregular is as effective as systematic 12-lead electrocardiography, and is more cost-effective.45 Hence, opportunistic clinic pulse screening forms the basis of guideline recommendations on screening.19 Cheaper, faster, yet accurate devices providing automated atrial fibrillation diagnosis, including handheld single-lead ECGs,44, 46, and 47 blood pressure oscillometry,48 and smartphone photoplethysmography,49 are likely to improve the cost-effectiveness. Indeed, screening for unknown atrial fibrillation is likely to be cost-effective for stroke prevention,46 and 50 and might lead to revisions of recommendations about screening for atrial fibrillation to prevent stroke.30 and 51

Stroke risk factors and risk stratification

Atrial fibrillation increases the risk of stroke and systemic thromboembolism, but the excess risk also depends on the presence of various additional risk factors, which were defined from findings from the non-warfarin placebo or control arms of historical randomised trials done two decades ago52 or from large observational epidemiological studies. There is good evidence of increased risk in patients with previous stroke or systemic embolism, age at least 65 years, recent decompensated heart failure (irrespective of ejection fraction; hence, heart failure with reduced or preserved ejection fraction, eg, hypertrophic cardiomyopathy),53 moderate-to-severe left ventricular dysfunction on cardiac imaging, diabetes mellitus, hypertension, or vascular disease (ie, peripheral artery disease or previous myocardial infarction).54 Female sex is probably a stroke risk modifier, with accentuation of risk in those older than 65 years or with at least one additional stroke risk factor.55 and 56 The age threshold conferring excess stroke risk seems to be even lower (age ≥50 years) in east Asians.57

More recent attention has been directed towards defining the stroke risk associated with a single stroke risk factor, since not all risk factors carry equal weight, and risk in atrial fibrillation varies depending on clinical setting (eg, community based vs hospitalised) and according to ethnic origin58 and availability of appropriate methods to establish event rates.59 The evidence is compelling that even a single stroke risk factor confers an excess risk of thromboembolism and mortality (figure 3),60, 61, and 62 with a positive net clinical benefit for OAC treatment compared with aspirin or no antithrombotic treatment for such patients.60 and 63


Figure 3

Risk of stroke with a single additional risk factor

Data from Fauchier and colleagues.60


The more common stroke risk factors have been incorporated into stroke risk stratification scores (appendix pp 7, 8), designed to help practical decision making in everyday practice. The most comprehensive review and comparison of stroke and bleeding risk scores is provided in the 2014 National Institute for Health and Care Excellence (NICE) guidelines,28 which are based on systematic reviews, evidence appraisal, and cost-effectiveness. Risk scores based on clinical factors have modest predictive value for identifying high-risk patients, and additional refinement of clinical-factor-based scores to improve identification of high-risk patients can be made by the addition of biomarkers (eg, von Willebrand factor, natriuretic peptides, or troponin) and imaging (eg, cerebral or cardiac imaging).64 Even then, c statistics (ie, how well a score predicts an event) suggest only modest discrimination despite addition of several biomarkers, but with additional costs and a major loss of simplicity and practicality for everyday clinical use.

Stroke risk is a continuum, and the artificial categorisation into low or high risk still leads to many patients with high-risk atrial fibrillation being undertreated. The default should be to use OACs to treat all patients with atrial fibrillation unless clearly defined as truly low risk. Hence, clinicians should be most concerned with identifying the very-low-risk patients who do not need thromboprophylaxis. The CHA2DS2-VASc score65 is useful as a simple clinical score for easy initial identification of low-risk patients (CHA2DS2-VASc score 0 in males and 1 in females) who have stroke rates of 1% or lower per year, who do not need antithrombotic treatment.59, 66, and 67

One common misperception is that paroxysmal atrial fibrillation carries a low risk of stroke, whereas in many studies the risk is almost identical to that of persistent or permanent atrial fibrillation, notwithstanding considerations of variable arrhythmia burden and associated risk factors, especially since patients with paroxysmal atrial fibrillation tend to be younger and have fewer risk factors. Guidelines recommend treating paroxysmal atrial fibrillation with OACs using identical rules as apply to persistent or permanent atrial fibrillation;19 and 68 unfortunately, in practice, this is often not done.69

Thromboprophylaxis in patients with atrial fibrillation

Guidelines recommend different approaches to thromboprophylaxis in atrial fibrillation. Some use CHA2DS2-VASc score in a categorical approach; for example, the American Heart Association/American College of Cardiology/Heart Rhythm Society guidelines68 define patients with atrial fibrillation as low, moderate, or high risk, and recommend antithrombotic treatment on that basis. Patients at high risk are those with a CHA2DS2-VASc score of at least 2, for whom OACs are recommended; low risk are those with a CHA2DS2-VASc score of 0, for whom no antithrombotic treatment should be considered. For those with a CHA2DS2-VASc score of 1, OACs, aspirin, or no antithrombotic treatment can be chosen, depending on risk factors and patient values and preferences.

The 2014 NICE guidelines28 and European Society of Cardiology guidelines19 recommend that the initial step should be to identify low-risk patients (CHA2DS2-VASc score 0 in males, 1 in females) who do not need antithrombotic treatment. The next step is to offer effective stroke prevention with OACs (either VKAs with good quality anticoagulation control or NOACs) to those with at least one additional stroke risk factor. Since the default should be to give anticoagulants to all patients with atrial fibrillation unless defined as truly low risk, we recommend this simplified approach (figure 4), because the decision to provide anticoagulation is already made with at least one additional stroke risk factor irrespective of score value or the addition of biomarkers or imaging. Since patients with atrial fibrillation have high rates of hospital admission and risk stratification is not a static one-off process, low-risk patients should be regularly reviewed to establish whether risk has increased. Aspirin alone should not be offered for stroke prevention in atrial fibrillation.26, 27, and 28


Figure 4

Recommended decision pathway for treatment of newly diagnosed non-valvular atrial fibrillation

VKA=vitamin K antagonist. *Also calculate the HAS-BLED score. If HAS-BLED ≥3, address the modifiable bleeding risk factors and plan a closer clinical follow-up.


Bleeding risk assessment should also be part of the clinical decision-making process. In most cases, an elevated bleeding risk score should not be used as a reason to withhold anticoagulation, because stroke risk almost invariably outweighs serious bleeding risk, but patients at high risk of bleeding should be flagged up for more careful review and follow-up, especially in the era of electronic health record alerts, with prompt attention given to common reversible bleeding risk factors within the score.70 These include uncontrolled hypertension; control of a known previous bleeding site, especially gastrointestinal; labile INRs if on a VKA; and excess alcohol or concomitant non-steroidal anti-inflammatory drug use.71 The simplest and best validated score is HAS-BLED72 (appendix pp 9, 10), which reliably predicts bleeding in patients on OACs (whether VKAs or other anticoagulants73), aspirin, or no antithrombotic treatment. HAS-BLED has also been validated in non-atrial fibrillation populations, including venous thromboembolism and patients undergoing bridging therapy. Simpler scores that aim to provide information valid for both VKAs and NOACs are likely to inappropriately categorise many patients who subsequently bleed as low risk or substantially underestimate bleeding risks in VKA-treated patients by ignoring labile INR as a risk criterion.74 and 75 More complex scores using biomarkers offer statistically improved prediction of high-risk patients, but are less simple for everyday clinical use and do not focus on the reversible bleeding risk factors.70

VKAs are effective and safe if well managed with good quality anticoagulation control, as shown by the time in therapeutic range (TTR) of over 70%.76 However, a good TTR can be difficult to obtain in clinical practice, or in some populations such as Asians.77 The control of VKAs is affected by several clinical features, and some common factors have been incorporated into the SAMe-TT2R2 score (appendix pp 9, 10).78 Although prediction of the actual TTR is modest, the score is best used pragmatically to flag patients unlikely to do well on a VKA (score >2), because of labile INRs or poor TTR and the consequent risk of thromboembolism, death, or bleeding.79 and 80 Such patients should be targeted for more regular review and follow-up; educational interventions and counselling, which improve TTR;81 or use of NOACs.82

With availability of four NOACs in addition to VKAs, we can fit a particular drug to a patient's clinical profile,83 and 84 by use of available evidence from large randomised trials and observational cohorts (figure 5 and discussed later). Various clinical considerations when choosing a particular type or dose of NOAC can be summarised by the mnemonic ABCDE: abnormally low weight (dose reduction might be needed); bleeding risk, especially previous or recent gastrointestinal bleeding; creatinine clearance (ie, renal function); drug interactions (eg, P-glycoprotein inhibitors); and elderly age (dose reduction might be needed).


Figure 5

Choice of oral anticoagulant drug to fit the patient profile

A simplified schema to assist physician choice of anticoagulant (VKA or individual NOAC) according to patient characteristic. BID=twice daily. CrCl=creatinine clearance. NOAC=non-vitamin K antagonist oral anticoagulant. TTR=time in therapeutic range. VKA=vitamin K antagonist. *Reduced to 2·5 mg BID with two of three criteria from age ≥80 years, bodyweight ≤60 kg, or serum creatinine concentration ≥133 μmol/L. †110 mg BID for patients with a CrCl 30–49 mL/min (most countries, but not in the USA); in the USA only, 75 mg BID (available in the USA only) for patients with CrCl 15–29 mL/min (and only 150 mg BID dose available in the USA for CrCl >30 mL/min). ‡30 mg with CrCl 15–49 mL/min, P-glycoprotein inhibitors, or weight <60 kg. §110 mg BID dose not available in the USA for atrial fibrillation. ¶Reduced to 15 mg if CrCl 15–49 mL/min. ||Dose to be halved if the patient has any of the following: CrCl 15–49 mL/min, bodyweight ≤60 kg, or concomitant use of P-glycoprotein inhibitors.


From clinical trials to real-world practice

The efficacy of warfarin compared with placebo or aspirin for stroke prevention in patients with non-valvular atrial fibrillation was established almost 30 years ago (table).4 In a meta-analysis86 of eight more recent stroke prevention trials (2005–11), the pooled rate of residual stroke or systemic embolism in the warfarin arms was significantly lower than in earlier trials (1·66% vs 2·09%), probably because of improved management of warfarin treatment (mean TTR 63·6% vs 42–81%, and four of six earlier trials with a TTR <60%), whereas the rates of major bleeding (1·4–3·4%) and intracranial haemorrhage (ICH; 0·61%) were similar.86


Meta-analyses of randomised, controlled trials or observational studies on the efficacy and safety of antithrombotic treatments for the prevention of stroke or systemic embolism in patients with non-valvular atrial fibrillation


Included trials Number of patients Comparison Stroke or SE RR or HR (95%CI) Major bleeding RR or HR (95%CI) Gastrointestinal bleeding RR or HR (95%CI) ICH RR or HR (95%CI) All-cause mortality RR or HR (95%CI)
Meta-analyses of randomised controlled trials of antithrombotic treatments
Hart et al (2007)4 5 primary, 1 secondary prevention trial 2900 Adjusted-dose warfarin vs placebo or no treatment 64% (49 to 74) RR reduction with warfarin Not calculated −66% (−235 to 18)* 6 vs 3 events 26% (3 to 43)
Hart et al (2007)4 12 trials 3647 Adjusted-dose warfarin vs antiplatelet treatment 39% (22 to 52) RR reduction with warfarin Not calculated −70% (−234 to 14)* 20 vs 7 events 9% (−19 to 30)
Hart et al (2007)4 7 trials 3990 Aspirin vs placebo or no treatment 19% (−1 to 35) RR reduction with aspirin Not calculated 2% (−98 to 52)* 8 vs 4 events 14% (−7 to 31)
Ruff et al (2014)5 4 trials 42 411 on a NOAC, 29 272 on warfarin NOACs vs adjusted-dose warfarin Overall RR 0·81 (0·73 to 0·91); ischaemic stroke RR 0·92 (0·83 to 1·02); haemorrhagic stroke RR 0·49 (0·38 to 0·64) RR 0.86 (0·71–1·00) RR 1·25 (1·01 to 1·55) RR 0·48 (0·39–0·59) RR 0·90 (0·85 to 0·95)
Meta-analyses of observational studies of dabigatran vs warfarin
Romanelli et al (2016)85 7 trials 197 348 on warfarin, 197 348 on dabigatran 150 mg, 11 305 on dabigatran 110 mg Dabigatran 150 mg or 110 mg vs dose-adjusted warfarin Dabigatran 150 mg HR 0·92 (0·84 to 1·10); dabigatran 110 mg HR 0·92 (0·72–1·18) Not reported Dabigatran 150 mg HR 1·23 (1·01 to 1·50); dabigatran 110 mg HR 0·91 (0·55 to 1·51) Dabigatran 150 mg HR 0·44 (0·34 to 0·59); dabigatran 110 mg HR 0·49 (0·34–0·72) Not reported

* Major extracranial bleeding.

HR=hazard ratio. ICH=intracranial haemorrhage. NOAC=non-vitamin K antagonist oral anticoagulant. RR=relative risk. SE=systemic embolism.

Owing to many limitations (panel 1),87 VKA treatment outside the trial setting is often suboptimal, and NOACs are increasingly used as a viable alternative. A meta-analysis5 of four landmark NOAC trials in non-valvular atrial fibrillation revealed a significant 19% stroke risk reduction versus VKAs, which was driven by the reduction in haemorrhagic stroke and no real change in ischaemic stroke; comparable safety in terms of major bleeding, with impressive reductions in ICH, at the cost of increased gastrointestinal bleeding; and a 10% reduction in all-cause mortality relative to warfarin (table and appendix pp 11–13). The efficacy and safety of NOACs over warfarin seems to be even greater in east Asians compared with non-Asians.88

Panel 1

Essential features of vitamin K antagonists and non-vitamin K antagonist oral anticoagulants


Vitamin K antagonists

  • Slow onset and offset of action, with some thrombophilia during onset and offset
  • Narrow therapeutic window (target INR 2·0–3·0)
  • Several interactions with food and other drugs, which affect the anticoagulation intensity
  • Variable dose response depending on the individual's genetic background
  • INR-guided dosing necessitates regular INR monitoring and frequent dose adjustments
  • TTR of >65–70% is vital for optimum stroke prevention
  • Used in clinical practice for a long time; not expensive

Non-vitamin K antagonist oral anticoagulants

  • Fast onset and offset of action; onset faster than offset
  • Fixed once or twice daily dosing
  • A few clinically relevant interactions with other drugs; no food interaction
  • Stable, dose-related anticoagulant effect; no need for regular laboratory monitoring of anticoagulation intensity, but renal function assessment is mandatory at baseline and during follow-up, depending on baseline renal function
  • Strict adherence to non-vitamin K antagonist oral anticoagulant treatment crucial for optimum efficacy
  • Relatively new drugs, expensive, but cost-effective, in comparison with vitamin K antagonists

See appendix pp 11–17 for more details. INR=international normalised ratio. TTR=time in therapeutic range.

Randomised trials provide the most objective evidence on a drug treatment, but the results might not be fully applicable to a range of real-world settings, because of the trial-specific inclusion and exclusion criteria. Notwithstanding some limitations, post-marketing observational studies, including prospective international registries and large administrative datasets, provide valuable complementary information on treatment performance in daily clinical practice. A meta-analysis of real-world observational studies on dabigatran versus warfarin for stroke prevention in non-valvular atrial fibrillation (table)85 yielded results broadly consistent with the main RE-LY trial. Key findings from those studies are shown in panel 2 and the appendix (pp 14–17). In real-world studies of rivaroxaban with lower risk patients, compared with the pivotal ROCKET-AF trial, the rates of stroke, major bleeding, and death seemed to be lower,89 but findings from more recent propensity-score-matched real-world studies with rivaroxaban consistently showed similar thromboembolism and bleeding risks to warfarin (appendix pp 14–17). In a large propensity-weighted analysis from a Danish nationwide cohort study,90 no significant differences were found between NOACs and warfarin for ischaemic stroke, but the risks of death, any bleeding, or major bleeding were significantly lower for apixaban and dabigatran than for warfarin. Warfarin and rivaroxaban had comparable annual bleeding rates.

Panel 2

Key points regarding oral anticoagulant treatment for thromboprophylaxis in patients with non-valvular atrial fibrillation


Key points from the landmark NOAC trials in patients with non-valvular atrial fibrillation*

  • NOACs were superior to warfarin (dabigatran 150 mg or apixaban) or similarly effective as warfarin (dabigatran 110 mg, rivaroxaban, or edoxaban in both doses) at reducing stroke or systemic embolism.
  • Stroke reduction was largely driven by the reduction in haemorrhagic stroke (significant for all NOACs vs VKAs) with minimal effect on ischaemic stroke, in which a significant reduction was only reported for dabigatran 150 mg.
  • NOACs were either safer than warfarin (dabigatran 110 mg, apixaban, or edoxaban in both doses) or as safe as warfarin (dabigatran 150 mg or rivaroxaban) with respect to major bleeding.
  • Apixaban was more effective than aspirin in the prevention of stroke or systemic embolism, with comparable safety (regarding major bleeding, haemorrhagic stroke, ICH, and gastrointestinal bleeding), and was better tolerated.

Key points from large real-world administrative dataset analyses of NOAC use22

  • Effectiveness and safety of dabigatran, rivaroxaban, and apixaban in real-world data were broadly consistent with findings from landmark trials; however, the ROCKET-AF study included higher-risk patients with more events than real-world studies.
  • Gastrointestinal bleeding was the most common major bleeding event.
  • ICH or fatal bleeding was rare.
  • Risk of bleeding during OAC initiation is higher with warfarin than with dabigatran.

Key points from contemporary large observational international prospective registries

  • OAC treatment for stroke prevention has increased in the past decade compared with earlier real-world data. Contemporary OAC use ranges from 45% in US general practice and 50% in Japan general practice, to 82% in European cardiologists and 87% in Japanese cardiovascular specialists.
  • Cardiologists and electrophysiologists are more likely to prescribe OACs across all stroke risk strata than are internal medicine specialists or primary care physicians.
  • OAC use is generally higher in Europe and Japan than in the USA.
  • VKAs are still the most commonly used OACs in many regions.
  • OACs are underused in eligible patients and overused in patients at truly low risk of stroke.
  • Physician perception of stroke and bleeding risk often differs from the evidence-based guideline-recommended risk assessment tool estimate.
  • The most common reason for OAC non-use is physician decision (real or perceived high bleeding risk, need or perceived need for concomitant antiplatelet treatment, paroxysmal or asymptomatic atrial fibrillation, only one stroke risk factor), followed by patient refusal.
  • Permanent discontinuation of warfarin is common among patients with atrial fibrillation (20% to >50%), especially in incident users and young patients. Discontinuation is most commonly because of physician preference, patient refusal, or bleeding events.
  • Antiplatelet treatment, mostly aspirin, is used in about a third of patients, frequently as the sole treatment, especially in elderly people. When concomitant with an OAC, only half have vascular disease.

ICH=intracranial haemorrhage. OAC=oral anticoagulant. NOAC=non-vitamin K antagonist oral anticoagulant. VKA=vitamin K antagonist.

Findings from large prospective international observational registries show that many patients with atrial fibrillation who are eligible for OACs because of high risk of stroke are still not treated with OACs, particularly the elderly or those at high risk of bleeding. Findings from registry studies highlight the gaps in daily clinical practice alluded to earlier, and identify the unmet needs regarding evidence-based guidance on optimum strategies for stroke prevention in some subsets of patients with atrial fibrillation.22 Key findings and a detailed summary are shown in panel 2 and the appendix (pp 18–25).

Specific management considerations

Because of the overlap in stroke and bleeding risk factors, high-risk patients with atrial fibrillation are often denied OACs without an absolute contraindication. Elderly people91 and most patients with a history of bleeding (eg, previous gastrointestinal bleeding with a healed culprit lesion) clearly benefit from OAC resumption.92 Patients with atrial fibrillation after intracerebral haemorrhage or those with severe renal disease represent other high-risk groups that were excluded from randomised trials, but findings from observational studies suggest some benefit from OACs.93, 94, and 95

Patients with atrial fibrillation undergoing percutaneous coronary intervention and stenting need a complex management strategy including OACs with single or dual antiplatelet treatment, to balance risk of stroke, recurrent ischaemia, or stent thrombosis against the risk of serious bleeding with combined treatment. Dependent on the patient risk profile and clinical setting (ie, acute coronary syndrome vs stable disease), triple treatment with OACs plus dual antiplatelet treatment, followed by OACs plus a single antiplatelet drug such as clopidogrel, should be used for the shortest period advisable.96 Thereafter, OAC monotherapy (a NOAC or well-managed VKA) should continue.96

Catheter atrial fibrillation ablation is superior to medical treatment at eliminating clinical atrial fibrillation recurrences, but should not be used to avoid OAC treatment; the decision regarding long-term OAC use after ablation should be based on individual stroke risk and not the estimated procedural success.97 Reports on the association of left atrial appendage isolation with appendage thrombus and stroke are conflicting, with results showing either improvement or a neutral effect, and there might even be a downside depending on the procedural details.98 Two ongoing trials are investigating early rhythm control (EAST; NCT01288352) or atrial fibrillation ablation versus antiarrhythmic drug treatment (CABANA; NCT00911508) on long-term risk of stroke and death.

Percutaneous left atrial appendage occlusion using the WATCHMAN, Amplatzer Cardiac Plug, or WaveCrest device or the Lariat endocardial and epicardial ligation technique might be an alternative for patients with atrial fibrillation who are at high risk of both stroke and bleeding or with contraindications to OACs; however, interventional cardiologists need to be trained in the procedure, and patients must receive dual antiplatelet treatment for at least 6 weeks after the procedure.99

Modifiable cardiovascular risk factors (eg, hypertension, obesity, dyslipidaemia, obstructive sleep apnoea, physical inactivity, and smoking) are important contributors to atrial fibrillation substrate progression and increased atrial fibrillation burden. Emerging evidence shows that aggressive risk factor management is likely to improve symptoms and reduce atrial fibrillation recurrence, thus facilitating rhythm control in patients with or without atrial fibrillation catheter ablation.100 and 101 Lifestyle interventions are likely to favourably affect cardiovascular outcomes, but whether these will reduce stroke remains to be established.

Population-centred or patient-centred interventions

Nurse-led clinics are an attractive possibility to improve uptake of stroke prevention strategies. In a randomised trial of 712 patients,102 appropriate OAC prescription increased from a high base of 83% in the usual care group to 99% in the nurse-led clinic. Although cardiovascular death and hospital admissions were both significantly reduced by the intervention, stroke was infrequent, with only 1% of patients having stroke in 22 months of follow-up, and was not significantly different between groups in this well-managed patient cohort. Another approach is to link population screening for unknown atrial fibrillation with screening for known but untreated atrial fibrillation, referring such individuals to a cardiology team to prescribe OACs.44

Marketing of NOACs has probably resulted in a rise in the proportion of eligible patients receiving anticoagulation; after introduction of NOACs in the UK, the proportion of patients with a CHA2DS2-VASc score of at least 2 starting anticoagulants for atrial fibrillation increased from 41% to 65%.103 Patient support groups (eg, Atrial Fibrillation Association and also have a part to play in increasing patient awareness of atrial fibrillation and its attendant stroke risk, and reducing reluctance to start OACs.

A neglected aspect of stroke prevention is ensuring that patients who start OACs continue to take the treatment indefinitely. Unfortunately, persistence with OACs, or rather non-persistence, is a major issue; 21–50% of patients discontinue VKAs by 1 year after inception.103 Findings from many studies have shown lower persistence with warfarin than a single NOAC. In one study,103 persistence with warfarin was significantly lower than with NOACs at 1 year (65% vs 83%; appendix p 2). This difference in persistence of drug treatment is likely to be a major factor in strokes related to atrial fibrillation, because cessation of VKAs more than doubles the stroke risk, with a peak in the first year after cessation, and a high absolute increase for at least 3 years after cessation.104 In view of the shorter half-life with NOACs, poor patient adherence also translates to a higher risk of stroke and mortality despite overall good adherence to these drugs.105 and 106 Therefore, greater efforts are needed to support the patient to increase adherence and continue OACs long term, whether with decision aids, educational measures, or patient counselling.81 and 107

Future directions

Increasing awareness of the role of unrecognised atrial fibrillation should accelerate efforts to detect atrial fibrillation before stroke has occurred and institute effective thromboprophylaxis with OACs. Widespread recognition of the role of undertreatment of atrial fibrillation in causation of ischaemic stroke will be of crucial importance to focus efforts to close the evidence–treatment gap for OACs, and replace aspirin with OACs in the therapeutic armamentarium. Basic and clinical research is needed to better understand the pathological atrial substrate leading to cardioembolism for which atrial fibrillation is likely to be a marker. In the longer term, efforts should be directed at primary prevention of atrial fibrillation, which might need similar lifestyle modifications as advocated for prevention of coronary heart disease.

Search strategy and selection criteria

We searched MEDLINE and PubMed (date of last search May 2, 2016) using the following search terms: “atrial fibrillation”, “warfarin”, “dabigatran”, “rivaroxaban”, “apixaban”, “edoxaban”, “randomized trial”, “real world”, “cohort study”, “registry”, “stroke prevention”, “stroke risk”, and “bleeding risk”, and checked reference lists from relevant articles. No publication time limits were specified, though preference was given to articles from the past 10 years and some highly cited older articles. This search primarily focused on studies to be included in the table, panel 2, and the appendix pp 14–25.


BF drafted, collated, and revised the manuscript. TSP did the literature searches. TSP and GYHL drafted and revised the manuscript.

Declaration of interests

BF has received research grants to undertake investigator-initiated studies from Bristol-Myers Squibb/Pfizer, Bayer, and Boehringer Ingelheim; has been a consultant for Bayer, Bristol-Myers Squibb/Pfizer, Boehringer Ingelheim, Servier, AstraZeneca, and Gilead; and has been a speaker for Bayer, Bristol-Myers Squibb/Pfizer, and AstraZeneca. TSP has been a consultant for Bayer, AstraZeneca, and Pfizer; and a speaker for Bayer, Pfizer, and AstraZeneca. GYHL has been a consultant for Bayer/Janssen, Bristol-Myers Squibb/Pfizer, Biotronik, Medtronic, Boehringer Ingelheim, Microlife, and Daiichi Sankyo; and a speaker for Bayer, Bristol-Myers Squibb/Pfizer, Medtronic, Boehringer Ingelheim, Microlife, Roche, and Daiichi Sankyo.


We thank Alistair Corbett for providing the stroke images in figure 1, and James Leyden and Leif Friberg for providing additional information on antiplatelet treatment for appendix figure 2 (pp 3, 4) from the studies they led.

Supplementary Material

Download file

Supplementary appendix


  • 1 HS Jorgensen, H Nakayama, J Reith, HO Raaschou, TS Olsen. Acute stroke with atrial fibrillation. The Copenhagen Stroke Study. Stroke. 1996;27:1765-1769 Crossref
  • 2 K Kimura, K Minematsu, T Yamaguchi. Atrial fibrillation as a predictive factor for severe stroke and early death in 15 831 patients with acute ischaemic stroke. J Neurol Neurosurg Psychiatry. 2005;76:679-683 Crossref
  • 3 C Marini, F De Santis, S Sacco, et al. Contribution of atrial fibrillation to incidence and outcome of ischemic stroke: results from a population-based study. Stroke. 2005;36:1115-1119 Crossref
  • 4 RG Hart, LA Pearce, MI Aguilar. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146:857-867 Crossref
  • 5 CT Ruff, RP Giugliano, E Braunwald, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383:955-962 Crossref
  • 6 D Mozaffarian, EJ Benjamin, AS Go, et al. Heart disease and stroke statistics-2016 update: a report from the American Heart Association. Circulation. 2016;133:e38-360
  • 7 PA Wolf, RD Abbott, WB Kannel. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983-988 Crossref
  • 8 S Bjorck, B Palaszewski, L Friberg, L Bergfeldt. Atrial fibrillation, stroke risk, and warfarin therapy revisited: a population-based study. Stroke. 2013;44:3103-3108 Crossref
  • 9 GS Yiin, DP Howard, NL Paul, et al. Recent time trends in incidence, outcome and premorbid treatment of atrial fibrillation-related stroke and other embolic vascular events: a population-based study. J Neurol Neurosurg Psychiatry. 2015; 10.1136/jnnp-2015-311947 published online Oct 20.
  • 10 JM Leyden, TJ Kleinig, J Newbury, et al. Adelaide Stroke Incidence Study: declining stroke rates but many preventable cardioembolic strokes. Stroke. 2013;44:1226-1231 Crossref
  • 11 GS Yiin, DP Howard, NL Paul, et al. Age-specific incidence, outcome, cost, and projected future burden of atrial fibrillation-related embolic vascular events: a population-based study. Circulation. 2014;130:1236-1244 Crossref
  • 12 L Friberg, M Rosenqvist, A Lindgren, A Terent, B Norrving, K Asplund. High prevalence of atrial fibrillation among patients with ischemic stroke. Stroke. 2014;45:2599-2605 Crossref
  • 13 L Li, GS Yiin, OC Geraghty, et al. Incidence, outcome, risk factors, and long-term prognosis of cryptogenic transient ischaemic attack and ischaemic stroke: a population-based study. Lancet Neurol. 2015;14:903-913
  • 14 T Sanna, HC Diener, RS Passman, et al. Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med. 2014;370:2478-2486 Crossref
  • 15 DJ Gladstone, M Spring, P Dorian, et al. Atrial fibrillation in patients with cryptogenic stroke. N Engl J Med. 2014;370:2467-2477 Crossref
  • 16 LA Sposato, LE Cipriano, G Saposnik, ER Vargas, PM Riccio, V Hachinski. Diagnosis of atrial fibrillation after stroke and transient ischaemic attack: a systematic review and meta-analysis. Lancet Neurol. 2015;14:377-387 Crossref
  • 17 RG Hart, HC Diener, SB Coutts, et al. Embolic strokes of undetermined source: the case for a new clinical construct. Lancet Neurol. 2014;13:429-438 Crossref
  • 18 G Ntaios, V Papavasileiou, H Milionis, et al. Embolic strokes of undetermined source in the Athens Stroke Registry: an outcome analysis. Stroke. 2015;46:2087-2093 Crossref
  • 19 AJ Camm, GYH Lip, RD Caterina, et al. 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. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J. 2012;33:2719-2747
  • 20 HC Diener, JD Easton, CB Granger, et al. Design of Randomized, double-blind, Evaluation in secondary Stroke Prevention comparing the EfficaCy and safety of the oral Thrombin inhibitor dabigatran etexilate vs. acetylsalicylic acid in patients with Embolic Stroke of Undetermined Source (RE-SPECT ESUS). Int J Stroke. 2015;10:1309-1312
  • 21 B Freedman, C Martinez, A Katholing, S Rietbrock. Residual risk of stroke and death in anticaogulant-treated patients with atrial fibrillation. JAMA Cardiol. 2016;1:366-368
  • 22 GY Lip, SM Al-Khatib, FG Cosio, et al. Contemporary management of atrial fibrillation: what can clinical registries tell us about stroke prevention and current therapeutic approaches?. J Am Heart Assoc. 2014;3:e001179 Crossref
  • 23 AK Kakkar, I Mueller, JP Bassand, et al. Risk profiles and antithrombotic treatment of patients newly diagnosed with atrial fibrillation at risk of stroke: perspectives from the international, observational, prospective GARFIELD registry. PLoS One. 2013;8:e63479 Crossref
  • 24 DJ Gladstone, E Bui, J Fang, et al. Potentially preventable strokes in high-risk patients with atrial fibrillation who are not adequately anticoagulated. Stroke. 2009;40:235-240 Crossref
  • 25 S Sjalander, A Sjalander, PJ Svensson, L Friberg. Atrial fibrillation patients do not benefit from acetylsalicylic acid. Europace. 2014;16:631-638 Crossref
  • 26 S Ben Freedman, BJ Gersh, GY Lip. Misperceptions of aspirin efficacy and safety may perpetuate anticoagulant underutilization in atrial fibrillation. Eur Heart J. 2015;36:653-656 Crossref
  • 27 GY Lip. The role of aspirin for stroke prevention in atrial fibrillation. Nat Rev Cardiol. 2011;8:602-606 Crossref
  • 28 NICE. Atrial fibrillation: management. NICE guidelines [CG180]. (2014) (accessed Aug 10, 2016).
  • 29 S Ogawa, K Aonuma, HF Tse, et al. The APHRS's 2013 statement on antithrombotic therapy of patients with nonvalvular atrial fibrillation. J Arrhythm. 2013;29:190-200 Crossref
  • 30 S Ben Freedman, N Lowres. Asymptomatic atrial fibrillation: the case for screening to prevent stroke. JAMA. 2015;314:1911-1912
  • 31 BJ Potter, J Le Lorier. Taking the pulse of atrial fibrillation. Lancet. 2015;386:113-115 Crossref
  • 32 C Martinez, A Katholing, SB Freedman. Adverse prognosis of incidentally detected ambulatory atrial fibrillation. A cohort study. Thromb Haemost. 2014;112:276-286 Crossref
  • 33 TS Tsang, ME Barnes, PA Pellikka, et al. Silent atrial fibrillation in Olmsted county: a community-based study. Can J Cardiol. 2011;27:S122 Crossref
  • 34 KC Siontis, BJ Gersh, JM Killian, et al. Typical, atypical, and asymptomatic presentations of new-onset atrial fibrillation in the community: characteristics and prognostic implications. Heart Rhythm. 2016;13:1418-1424
  • 35 G Boriani, C Laroche, I Diemberger, et al. Asymptomatic atrial fibrillation: clinical correlates, management, and outcomes in the EORP-AF Pilot General Registry. Am J Med. 2015;128:509 18.e2 Crossref
  • 36 JS Healey, SJ Connolly, MR Gold, et al. Subclinical atrial fibrillation and the risk of stroke. N Engl J Med. 2012;366:120-129 Crossref
  • 37 TV Glotzer, AS Hellkamp, J Zimmerman, et al. Atrial high rate episodes detected by pacemaker diagnostics predict death and stroke: report of the Atrial Diagnostics Ancillary Study of the MOde Selection Trial (MOST). Circulation. 2003;107:1614-1619 Crossref
  • 38 BS Larsen, P Kumarathurai, J Falkenberg, OW Nielsen, A Sajadieh. Excessive atrial ectopy and short atrial runs increase the risk of stroke beyond incident atrial fibrillation. J Am Coll Cardiol. 2015;66:232-241 Crossref
  • 39 M Brambatti, SJ Connolly, MR Gold, et al. Temporal relationship between subclinical atrial fibrillation and embolic events. Circulation. 2014;129:2094-2099 Crossref
  • 40 MP Turakhia, PD Ziegler, SK Schmitt, et al. Atrial fibrillation burden and short-term risk of stroke: case-crossover analysis of continuously recorded heart rhythm from cardiac electronic implanted devices. Circ Arrhythm Electrophysiol. 2015;8:1040-1047
  • 41 DT Martin, MM Bersohn, AL Waldo, et al. Randomized trial of atrial arrhythmia monitoring to guide anticoagulation in patients with implanted defibrillator and cardiac resynchronization devices. Eur Heart J. 2015;36:1660-1668 Crossref
  • 42 H Kamel, PM Okin, MS Elkind, C Iadecola. Atrial fibrillation and mechanisms of stroke: time for a new model. Stroke. 2016;47:895-900
  • 43 N Lowres, L Neubeck, J Redfern, SB Freedman. Screening to identify unknown atrial fibrillation. A systematic review. Thromb Haemost. 2013;110:213-222 Crossref
  • 44 E Svennberg, J Engdahl, F Al-Khalili, L Friberg, V Frykman, M Rosenqvist. Mass screening for untreated atrial fibrillation: the STROKESTOP study. Circulation. 2015;131:2176-2184 Crossref
  • 45 DA Fitzmaurice, FDR Hobbs, S Jowett, et al. Screening versus routine practice in detection of atrial fibrillation in patients aged 65 or over: cluster randomised controlled trial. BMJ. 2007;335:383 Crossref
  • 46 N Lowres, L Neubeck, G Salkeld, et al. Feasibility and cost-effectiveness of stroke prevention through community screening for atrial fibrillation using iPhone ECG in pharmacies. The SEARCH-AF study. Thromb Haemost. 2014;111:1167-1176 Crossref
  • 47 RG Tieleman, Y Plantinga, D Rinkes, et al. Validation and clinical use of a novel diagnostic device for screening of atrial fibrillation. Europace. 2014;16:1291-1295 Crossref
  • 48 WJ Verberk, S Omboni, A Kollias, GS Stergiou. Screening for atrial fibrillation with automated blood pressure measurement: research evidence and practice recommendations. Int J Cardiol. 2015;203:465-473
  • 49 DD McManus, J Lee, O Maitas, et al. A novel application for the detection of an irregular pulse using an iPhone 4S in patients with atrial fibrillation. Heart Rhythm. 2013;10:315-319 Crossref
  • 50 LA Levin, M Husberg, PD Sobocinski, et al. A cost-effectiveness analysis of screening for silent atrial fibrillation after ischaemic stroke. Europace. 2015;17:207-214 Crossref
  • 51 P Kirchhof, G Breithardt, J Bax, et al. A roadmap to improve the quality of atrial fibrillation management: proceedings from the fifth Atrial Fibrillation Network/European Heart Rhythm Association consensus conference. Europace. 2016;18:37-50
  • 52 Stroke Risk in Atrial Fibrillation Working Group. Independent predictors of stroke in patients with atrial fibrillation: a systematic review. Neurology. 2007;69:546-554
  • 53 A Banerjee, S Taillandier, JB Olesen, et al. Ejection fraction and outcomes in patients with atrial fibrillation and heart failure: the Loire Valley Atrial Fibrillation Project. Eur J Heart Fail. 2012;14:295-301 Crossref
  • 54 R Pisters, DA Lane, F Marin, AJ Camm, GY Lip. Stroke and thromboembolism in atrial fibrillation. Circ J. 2012;76:2289-2304 Crossref
  • 55 AJ Wagstaff, TF Overvad, GY Lip, DA Lane. Is female sex a risk factor for stroke and thromboembolism in patients with atrial fibrillation? A systematic review and meta-analysis. QJM. 2014;107:955-967 Crossref
  • 56 JP Piccini, DN Simon, BA Steinberg, et al. Differences in clinical and functional outcomes of atrial fibrillation in women and men: two-year results from the ORBIT-AF registry. JAMA Cardiol. 2016;1:282-291
  • 57 TF Chao, KL Wang, CJ Liu, et al. Age threshold for increased stroke risk among patients with atrial fibrillation: a nationwide cohort study from Taiwan. J Am Coll Cardiol. 2015;66:1339-1347
  • 58 GY Lip, PB Nielsen. Should patients with atrial fibrillation and 1 stroke risk factor (CHA2DS2-VASc score 1 in men, 2 in women) be anticoagulated? Yes: even 1 stroke risk factor confers a real risk of stroke. Circulation. 2016;133:1498-1503
  • 59 PB Nielsen, TB Larsen, F Skjoth, TF Overvad, GY Lip. Stroke and thromboembolic event rates in atrial fibrillation according to different guideline treatment thresholds: a nationwide cohort study. Sci Rep. 2016;6:27410
  • 60 L Fauchier, N Clementy, F Ivanes, D Angoulvant, D Babuty, G Lip. Should atrial fibrillation patients with only 1 nongender-related CHA2DS2-VASc risk factor be anticoagulated?. Stroke. 2016;47:1831-1836
  • 61 GY Lip, F Skjoth, LH Rasmussen, TB Larsen. Oral anticoagulation, aspirin, or no therapy in patients with nonvalvular AF with 0 or 1 stroke risk factor based on the CHADS-VASc score. J Am Coll Cardiol. 2015;65:1385-1394 Crossref
  • 62 TF Chao, CJ Liu, KL Wang, et al. Should atrial fibrillation patients with 1 additional risk factor of the CHA2DS2-VASc score (beyond sex) receive oral anticoagulation?. J Am Coll Cardiol. 2015;65:635-642 Crossref
  • 63 GY Lip, F Skjoth, PB Nielsen, TB Larsen. Non-valvular atrial fibrillation patients with none or one additional risk factor of the CHA2DS2-VASc score. A comprehensive net clinical benefit analysis for warfarin, aspirin, or no therapy. Thromb Haemost. 2015;114:826-834
  • 64 GY Lip. Stroke and bleeding risk assessment in atrial fibrillation: when, how, and why?. Eur Heart J. 2013;34:1041-1049 Crossref
  • 65 GY Lip, R Nieuwlaat, R Pisters, DA Lane, HJ Crijns. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the euro heart survey on atrial fibrillation. Chest. 2010;137:263-272 Crossref
  • 66 Q Xiong, S Chen, K Senoo, M Proietti, K Hong, GY Lip. The CHADS2 and CHA2DS2-VASc scores for predicting ischemic stroke among east Asian patients with atrial fibrillation: a systemic review and meta-analysis. Int J Cardiol. 2015;195:237-242 Crossref
  • 67 W-G Zhu, Q-M Xiong, K Hong. Meta-analysis of CHADS2 versus CHA2DS2-VASc for predicting stroke and thromboembolism in atrial fibrillation patients independent of anticoagulation. Tex Heart Inst J. 2015;42:6-15 Crossref
  • 68 CT January, LS Wann, JS Alpert, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130:e199-e267 Crossref
  • 69 JC Hsu, PS Chan, F Tang, TM Maddox, GM Marcus. Differences in anticoagulant therapy prescription in patients with paroxysmal versus persistent atrial fibrillation. Am J Med. 2015;128:654.e1-654.e10 Crossref
  • 70 GY Lip, DA Lane. Bleeding risk assessment in atrial fibrillation: observations on the use and misuse of bleeding risk scores. J Thromb Haemost. 2016; 10.1111/jth.13386 published online June 14.
  • 71 GY Lip, DA Lane. Assessing bleeding risk in atrial fibrillation with the HAS-BLED and ORBIT scores: clinical application requires focus on the reversible bleeding risk factors. Eur Heart J. 2015;36:3265-3267
  • 72 R Pisters, DA Lane, R Nieuwlaat, CB de Vos, HJ Crijns, GY Lip. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138:1093-1100 Crossref
  • 73 SV Wang, JM Franklin, RJ Glynn, S Schneeweiss, W Eddings, JJ Gagne. Prediction of rates of thromboembolic and major bleeding outcomes with dabigatran or warfarin among patients with atrial fibrillation: new initiator cohort study. BMJ. 2016;353:i2607
  • 74 M Proietti, K Senoo, DA Lane, GY Lip. Major bleeding in patients with non-valvular atrial fibrillation: impact of time in therapeutic range on contemporary bleeding risk scores. Sci Rep. 2016;6:24376
  • 75 K Senoo, M Proietti, DA Lane, GY Lip. Evaluation of the HAS-BLED, ATRIA and ORBIT bleeding risk scores in atrial fibrillation patients on warfarin. Am J Med. 2016;129:600-607
  • 76 F Björck, H Renlund, GH Lip, P Wester, PJ Svensson, A Själander. Outcomes in a warfarin-treated population with atrial fibrillation. JAMA Cardiol. 2016;1:172-180
  • 77 CE Chiang, KL Wang, GY Lip. Stroke prevention in atrial fibrillation: an Asian perspective. Thromb Haemost. 2014;111:789-797 Crossref
  • 78 S Apostolakis, RM Sullivan, B Olshansky, GY Lip. Factors affecting quality of anticoagulation control among patients with atrial fibrillation on warfarin: the SAMe-TT(2)R(2) score. Chest. 2013;144:1555-1563 Crossref
  • 79 GY Lip, K Haguenoer, C Saint-Etienne, L Fauchier. Relationship of the SAMe-TT2R2 score to poor-quality anticoagulation, stroke, clinically relevant bleeding, and mortality in patients with atrial fibrillation. Chest. 2014;146:719-726 Crossref
  • 80 M Proietti, DA Lane, GYH Lip. Relation of the SAMe-TT2R2 score to quality of anticoagulation control and thromboembolic events in atrial fibrillation patients: observations from the SPORTIF trials. Int J Cardiol. 2016;216:168-172
  • 81 DE Clarkesmith, HM Pattison, GY Lip, DA Lane. Educational intervention improves anticoagulation control in atrial fibrillation patients: the TREAT randomised trial. PLoS One. 2013;8:e74037 Crossref
  • 82 MA Esteve-Pastor, V Roldan, M Valdes, GY Lip, F Marin. The SAMe-TT2R2 score and decision-making between a vitamin K antagonist or a non-vitamin K antagonist oral anticoagulant in patients with atrial fibrillation. Expert Rev Cardiovasc Ther. 2016;14:177-187
  • 83 H-C Diener, J Aisenberg, J Ansell, et al. Choosing a particular oral anticoagulant and dose for stroke prevention in individual patients with non-valvular atrial fibrillation: part 1. Eur Heart J. 2016; 10.1093/eurheartj/ehw069 published online Feb 4.
  • 84 H-C Diener, J Aisenberg, J Ansell, et al. Choosing a particular oral anticoagulant and dose for stroke prevention in individual patients with non-valvular atrial fibrillation: part 2. Eur Heart J. 2016; 10.1093/eurheartj/ehv643 published online Feb 4.
  • 85 RJ Romanelli, L Nolting, M Dolginsky, E Kym, KB Orrico. Dabigatran versus warfarin for atrial fibrillation in real-world clinical practice: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes. 2016;9:126-134
  • 86 S Agarwal, R Hachamovitch, V Menon. Current trial-associated outcomes with warfarin in prevention of stroke in patients with nonvalvular atrial fibrillation: a meta-analysis. Arch Intern Med. 2012;172:623-631
  • 87 R De Caterina, S Husted, L Wallentin, et al. Vitamin K antagonists in heart disease: current status and perspectives (section III). Position paper of the ESC Working Group on Thrombosis—Task Force on Anticoagulants in Heart Disease. Thromb Haemost. 2013;110:1087-1107 Crossref
  • 88 KL Wang, GY Lip, SJ Lin, CE Chiang. Non-vitamin K antagonist oral anticoagulants for stroke prevention in Asian patients with nonvalvular atrial fibrillation: meta-analysis. Stroke. 2015;46:2555-2561
  • 89 AJ Camm, P Amarenco, S Haas, et al. XANTUS: a real-world, prospective, observational study of patients treated with rivaroxaban for stroke prevention in atrial fibrillation. Eur Heart J. 2016;37:1145-1153
  • 90 TB Larsen, F Skjoth, PB Nielsen, JN Kjaeldgaard, GY Lip. Comparative effectiveness and safety of non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study. BMJ. 2016;353:i3189
  • 91 GY Lip, N Clementy, L Pericart, A Banerjee, L Fauchier. Stroke and major bleeding risk in elderly patients aged ≥75 years with atrial fibrillation: the Loire Valley atrial fibrillation project. Stroke. 2015;46:143-150
  • 92 L Staerk, GY Lip, JB Olesen, et al. Stroke and recurrent haemorrhage associated with antithrombotic treatment after gastrointestinal bleeding in patients with atrial fibrillation: nationwide cohort study. BMJ. 2015;351:h5876
  • 93 PB Nielsen, TB Larsen, F Skjoth, A Gorst-Rasmussen, LH Rasmussen, GY Lip. Restarting anticoagulant treatment after intracranial hemorrhage in patients with atrial fibrillation and the impact on recurrent stroke, mortality, and bleeding: a nationwide cohort study. Circulation. 2015;132:517-525 Crossref
  • 94 L Friberg, L Benson, GY Lip. Balancing stroke and bleeding risks in patients with atrial fibrillation and renal failure: the Swedish Atrial Fibrillation Cohort study. Eur Heart J. 2015;36:297-306 Crossref
  • 95 AN Bonde, GY Lip, AL Kamper, et al. Net clinical benefit of antithrombotic therapy in patients with atrial fibrillation and chronic kidney disease: a nationwide observational cohort study. J Am Coll Cardiol. 2014;64:2471-2482 Crossref
  • 96 GY Lip, S Windecker, K Huber, et al. Management of antithrombotic therapy in atrial fibrillation patients presenting with acute coronary syndrome and/or undergoing percutaneous coronary or valve interventions: a joint consensus document of the European Society of Cardiology Working Group on Thrombosis, European Heart Rhythm Association (EHRA), European Association of Percutaneous Cardiovascular Interventions (EAPCI) and European Association of Acute Cardiac Care (ACCA) endorsed by the Heart Rhythm Society (HRS) and Asia-Pacific Heart Rhythm Society (APHRS). Eur Heart J. 2014;35:3155-3179
  • 97 C Sticherling, F Marin, D Birnie, et al. Antithrombotic management in patients undergoing electrophysiological procedures: a European Heart Rhythm Association (EHRA) position document endorsed by the ESC Working Group Thrombosis, Heart Rhythm Society (HRS), and Asia Pacific Heart Rhythm Society (APHRS). Europace. 2015;17:1197-1214 Crossref
  • 98 A Rillig, RR Tilz, T Lin, et al. Unexpectedly high incidence of stroke and left atrial appendage thrombus formation after electrical isolation of the left atrial appendage for the treatment of atrial tachyarrhythmias. Circ Arrhythm Electrophysiol. 2016;9:e003461
  • 99 B Meier, Y Blaauw, AA Khattab, et al. EHRA/EAPCI expert consensus statement on catheter-based left atrial appendage occlusion. Europace. 2014;16:1397-1416 Crossref
  • 100 C Gallagher, JM Hendriks, R Mahajan, et al. Lifestyle management to prevent and treat atrial fibrillation. Expert Rev Cardiovasc Ther. 2016;14:799-809
  • 101 JD Miller, KN Aronis, J Chrispin, et al. Obesity, exercise, obstructive sleep apnea, and modifiable atherosclerotic cardiovascular disease risk factors in atrial fibrillation. J Am Coll Cardiol. 2015;66:2899-2906
  • 102 JM Hendriks, R de Wit, HJ Crijns, et al. Nurse-led care vs. usual care for patients with atrial fibrillation: results of a randomized trial of integrated chronic care vs. routine clinical care in ambulatory patients with atrial fibrillation. Eur Heart J. 2012;33:2692-2699 Crossref
  • 103 C Martinez, A Katholing, C Wallenhorst, SB Freedman. Therapy persistence in newly diagnosed non-valvular atrial fibrillation treated with warfarin or NOAC. A cohort study. Thromb Haemost. 2016;115:31-39
  • 104 SB Freedman, C Martinez, C Wallenhorst, A Katholing. High burden of potentially avoidable stroke from discontinuation of warfarin therapy in non-valvular atrial fibrillation. Eur Heart J. 2015;36(suppl):862 (abstr P4981).
  • 105 A Gorst-Rasmussen, F Skjoth, TB Larsen, LH Rasmussen, GY Lip, DA Lane. Dabigatran adherence in atrial fibrillation patients during the first year after diagnosis: a nationwide cohort study. J Thromb Haemost. 2015;13:495-504 Crossref
  • 106 S Shore, EP Carey, MP Turakhia, et al. Adherence to dabigatran therapy and longitudinal patient outcomes: insights from the Veterans Health Administration. Am Heart J. 2014;167:810-817 Crossref
  • 107 NICE. Atrial fibrillation: management. (accessed Aug 1, 2016).


a Heart Research Institute, Charles Perkins Centre, University of Sydney, Sydney, NSW, Australia

b Department of Cardiology and Anzac Research Institute, Concord Hospital, Concord, NSW, Australia

c Cardiology Clinic, Clinical Center of Serbia, School of Medicine, University of Belgrade, Belgrade, Serbia

d University of Birmingham Institute of Cardiovascular Science, City Hospital, Birmingham, UK

e Aalborg Thrombosis Research Unit, Department of Clinical Medicine, Aalborg University, Aalborg, Denmark

* Correspondence to: Prof Ben Freedman, Heart Research Institute, Charles Perkins Centre, University of Sydney, Sydney, NSW 2006, Australia

© 2016 Elsevier Ltd. All rights reserved.

No responsibility is assumed by Elsevier, its licensors or associates for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.  Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.

This e-print is distributed with the support of Boehringer Ingelheim.      

Edited for:
Elsevier España, S.L.U.
(A member of Elsevier)
Av. Josep Tarradellas, 20-30
08029 Barcelona
Tel.: 932 000 711
Fax: 932 091 136