Skip to content

You are reading 1 of 2 free-access articles allowed for 30 days

Choosing the most effective approach to atrial fibrillation


Atrial fibrillation (AF) occurs in approximately 1-to-2 per cent of all adults worldwide and 9 per cent of those aged 80 years and older. AF is associated with increased rates of stroke and other thromboembolic events, heart failure, dementia, hospitalisations and an overall doubling of mortality. Affecting approximately 45,000 people in Ireland, the prevalence is estimated to double in the next 50 years with our ageing population.

Stroke is the third-leading cause of death and the leading cause of adult disability. Overall, AF is associated with a five-fold increase in  stroke risk. AF has been  implicated in one-in-five of all strokes, though the incidence in Ireland, as reported by the North Dublin Stroke Study, is as high as one-in-three. AF-associated strokes are more severe, incur a higher burden of disability and mortality and are, unsurprisingly, more often recurrent. The risk of death from AF-related stroke is doubled and the cost of care is increased 1.5-fold.

In Ireland, public awareness of AF is poor. A population-based survey showed that only 30 per cent of the general population had heard of AF, and almost half of those did not know what AF was. Most were not aware that AF is a risk factor for stroke.

Understanding of AF is poor but it is a difficult concept to explain without effective language. Public awareness is important for compliance with preventative treatment and for any proposed population screening initiatives. A simple analogy, which we have found useful when explaining to patients about the implications of AF, is that of a ‘faulty’ cement mixer.

The fibrillating atrium represents the malfunctioning mixer, not adequately mixing the cement, so that it hardens and forms clumps (clots), which, if leaving the mixer, will be pumped through the pipework, obstructing flow downstream, resulting in a blockage and thus the stroke.

Assessment of stroke and bleeding risk

All patients with AF have some risk of thromboembolism, the most devastating of which is stroke. All scoring systems to assess stroke risk in atrial AF have a poor ‘c’ statistic and are not overly reliable in their power to predict risk accurately.

Table 1: CHA2D2-VASc assessment scales for risk of stroke with AF

The CHA2DS2-VASc scoring system (see Table 1) is the most widely recommended means of assessing stroke risk, with a clear relationship between CHA2DS2-VASc score and stroke rate. In patients with a score ≥2, the risk is considered high and long-term oral anticoagulant (OAC) therapy (with a vitamin K antagonist (VKA) such as warfarin, direct thrombin inhibitor such as dabigatran or an oral Xa inhibitor such as apixaban, edoxaban or rivaroxaban) is indicated. Those with scores <2 are at a lesser (but not zero) risk and careful consideration needs to be given to risk/ benefits of OAC therapy in these individual situations. 

An individual’s risk of stroke may also depend on factors not accounted for by existing risk scores, such as intra-atrial anatomy, and in particular the left atrial appendage size, ejection volume and morphology, and the nature of the atrial endothelium, ie, the individual’s ‘cement mixer’. An individual’s blood (‘cement’) may also vary from person-to-person or even at different times in the same individual where it may be more pro-coagulant (‘thicker’) at times of infection, surgery, trauma and sepsis, etc.

None of this is assessed by available risk-scoring systems to date and guidelines for stroke prevention in AF have undergone a paradigm shift in the last five years, thus reflecting an approach of asking ourselves ‘why am I not anti-coagulating?’ rather than ‘should I anticoagulate?’.

Despite increasing rates of anticoagulation of older patients with AF, rates of intracerebral haemorrhage are considerably lower than in the past, typically between 0.1 and 0.6 per cent in contemporary reports. This probably reflects both careful dose regulation of warfarin by anticoagulation clinics, better control of hypertension and, more importantly, the advent of safer, newer anticoagulants. Intracranial bleeding was reduced in clinical  trials of all newer anticoagulants (no longer ‘novel’ and now referred to as non-VKA oral anticoagulants — NOACS)  a trend borne out in meta-analysis of almost 60,000 clinical trial patients collectively and now also in real-life registries.

Falls, and risk of falls, are commonly cited as a reason not to anticoagulate someone with AF. However, the risk is often overstated, as one study showed a patient may need to fall several hundred times per year for the risk of traumatic intracranial haemorrhage to outweigh the benefit of OAC therapy in stroke prevention. Contemporary registries, such as the Loire-AF, show that fallers do not have higher rates of intracerebral bleeds than non-fallers.

Table 2: HAS-BLED scoring system for bleeding risk on OAC

A commonly-used tool for assessing bleeding risk is the HAS-BLED scale (see Table 2). A score of ≥3 indicates ‘high risk’. However, this in itself should not preclude starting antithrombotic therapy, but rather indicates the need for a proper discussion of risk and benefit with patient, increased caution with the agent and dose, and a regular review of concomitant therapies.

Antithrombotic therapies

Only anticoagulation therapy has been shown to reduce AF-associated deaths, and effective anticoagulation can reduce the risk of stroke by up to 65 per cent. Stroke outcomes are also improved in patients with AF which occur while on therapeutic anticoagulation.


VKA therapies, eg, warfarin, have been used for decades, and warfarin has shown superiority over controls/aspirin in randomised controlled trials, albeit in small study populations by modern standards. Relative risk reduction for ischaemic strokes versus control was 67 per cent. An INR of 2.0-to-3.0 is the likely optimal range for prevention of stroke.

While effective if used appropriately, there are several difficulties with using warfarin. Careful monitoring of INR is essential to ensure adequate time in therapeutic range (TTR). Studies suggest that the TTR can range from 44-77 per cent.

Having patients below the TTR for 60 per cent of the time may completely offset the benefit of warfarin. In addition, there is high inter-individual and intra-individual variation in INR, as the anticoagulant effect can be affected by certain foods, drugs and alcohol. Certain genotypes can also influence dosing requirements. There is also the practical implication of regular blood testing and the cost of this. As a drug, it can be singularly unsuitable for the patient population with the highest prevalence of, and greatest risk from, AF due to age, impaired mobility, polypharmacy and inconvenience. In Ireland, as with many European countries, it has failed as a population strategy to prevent stroke.

However, while not considered ‘gold standard’ anymore by stroke physicians, warfarin still has its place in stroke prevention where AF occurs in the setting of mechanical heart valves and in advanced renal failure, where NOACs are currently unsuitable.

The RE-ALIGN trial compared dabigatran to warfarin in patients who had undergone aortic or mitral valve replacement and found warfarin superior in this setting.


Over recent years, there have been significant therapeutic developments, resulting in the NOACs. NOACs in AF fall into two classes: the direct thrombin inhibitors (eg, dabigatran) and direct factor Xa inhibitors (eg, apixaban, edoxaban and rivaroxaban). All NOACs have shown non-inferiority compared with VKAs, with better safety, and all consistently showing a reduction in intracranial bleeds compared to warfarin. The European Society of Cardiology now recommends NOACs as broadly preferable to VKAs in the vast majority of patients with non-valvular AF.

The following is a summary of the NOACs licensed for the treatment of non-valvular AF in Ireland, in order of their licensing: 

1. Dabigatran 

The RE-LY trial (n=18,113) compared two doses of dabigatran (110mg bd and 150mg bd) with adjusted-dose warfarin. Dabigatran 110mg was non-inferior to warfarin, and dabigatran 150mg was superior in preventing ischaemic stroke. There were fewer major bleeds on the lower dose of dabigatran. Rates of haemorrhagic stroke and intracranial haemmorhage (ICH) were lower with both doses of dabigatran, but gastrointestinal bleeding was significantly increased with the higher dose. A non-significant numerical increase in myocardial infarction (MI) was observed in the original trial, although not seen in subsequent large ‘real-life’ registries.

2. Rivaroxaban 

The ROCKET-AF trial (n=14,264) randomised patients with AF to treatment with either rivaroxaban (20mg or 15mg OD, depending on renal function) or warfarin. Rivaroxaban was non-inferior to warfarin for the primary endpoint of stroke and systemic embolism. There was no difference in the primary safety endpoint, a composite of major and clinically relevant non-major bleeding, but this represented the dichotomy of a significant reduction in fatal and intracranial bleeding with rivaroxaban, but with an increased trend to gastrointestinal bleeds and occult blood loss requiring transfusion over warfarin.

3. Apixaban

The ARISTOTLE trial (n=18,201) compared apixaban (5mg BD or dose-adjusted 2.5mg BD for >2 frailty markers of: age >80, weight <60kg and creatinine >133umol/L) to dose-controlled warfarin. There was a significant reduction in the primary efficacy outcome of all-cause stroke or systemic embolism by 21 per cent with apixaban compared with warfarin, with a 31 per cent reduction in major bleeding and a significant 11 per cent reduction in all-cause mortality (but not cardiovascular mortality). Rates of haemorrhagic stroke and ICH —but not of ischaemic stroke — were significantly lower in patients treated with apixaban than with warfarin. Gastrointestinal bleeding was similar between the treatment arms.

4. Edoxaban  

The ENGAGE AF-TIMI 48 trial published in 2013 (n=21,105) was the largest trial ever undertaken in AF and compared two doses of edoxaban (30mg and 60mg) to standard warfarin. Both once-daily regimens of edoxaban were non-inferior to warfarin, although there were numerically more strokes than warfarin with the 30mg dose. Both doses were associated with significantly lower rates of major and intracranial bleeding and death from cardiovascular causes. Edoxaban 60mg OD has just recently been licensed for stroke prevention in AF in Europe.

For all of the above NOACs, compliance and adherence to treatment is crucial, because of their relatively short half-lives (average 11-to-13 hours compared to >40 hours with warfarin). Patients may be left without any anticoagulation cover if more than one dose is missed.

All of these drugs have a degree of renal excretion, and doses may be altered based on impaired renal function. It is therefore essential that all patients at baseline have an estimated glomerular filtration rate (eGFR) calculated, preferably using the Cockroft and Gault equation. Patients should have regular checks on compliance, side-effects and renal function.

Further information regarding initiation and monitoring of NOACs can be found at:


The evidence for effective stroke prevention with aspirin in AF is weak. In addition, the ACTIVE-W trial, which compared dual anti-platelets (aspirin and clopidogrel) to warfarin, was stopped early as antiplatelets were inferior, even when combined for stroke prevention in AF compared to warfarin.

Antiplatelets therefore have the potential for harm (as data indicate that the risk of major bleeding or ICH with aspirin may not be all that significantly different to OAC therapies, especially in older groups, without benefit in AF.

Use of antiplatelet therapy for stroke prevention in AF should be limited to patients who decline any form of OAC therapy. Both ESC and UK National Institute for Health and Care Excellence (NICE) guidelines no longer recommend aspirin as a treatment for AF.

Left atrial appendage occlusion

The left atrial appendage (LAA) is considered the main site of thrombus in AF patients. Trans-oesophageal echocardiography detects most thrombi located in the LAA. There are several devices available that can be surgically inserted in order to occlude the LAA, with the hypothesis being that in doing so, they reduce the risk of stroke.

Although the concept of LAA closure seems reasonable, the procedure itself carries a risk of stroke and pericardial tamponade. The recent PREVAIL trial compared LAA closure with the WATCHMAN device vs long-term warfarin therapy.

In this trial, LAA occlusion was as good as warfarin for ischaemic stroke prevention after the initial peri-procedure period of seven days. It is a consideration in patients who cannot tolerate an anticoagulant long-term.

Screening and prevention

Although AF could be viewed as the biggest public health problem in stroke, there is not currently a national screening programme in Ireland. AF, as a disease, and fulfils the original Wilson and Jungner World Health Organisation (WHO) screening criteria.

However, there are several studies ongoing looking at the feasibility of screening in high-risk groups using modern lightweight cardiac telemetry devices. A recent AF project in the west of Ireland found pulse checking with supportive ECG to be a practical, feasible and economic way of detecting AF in general practice, although a concern remains that most of the cases detected were already known about.

It could be falsely reassuring and many cases would be missed if this was the only screening, as most AF is paroxysmal.  

References available on request

Key learning points


  • AF confers a five-fold increase in the risk of stroke.
  • AF can be managed in the community, and specialist referral is advised if treatment fails to control symptoms or anticoagulation decision is difficult.
  • Stroke and bleeding risk helps guide discussion with patients but almost all patients with AF should be anticoagulated.
  • Anticoagulation with NOACs is now preferred over VKAs in non-valvular AF based on very strong trial data on safety and efficacy and the historic failure of warfarin as a population-based strategy to prevent stroke.



Leave a Comment

You must be logged in to post a comment.

Scroll To Top