This site is intended for health professionals only

At the heart of general practice since 1960

June 2007: Assessing stroke risk in patients with AF

How should AF be investigated?

Should rhythm or rate be controlled?

When should aspirin or warfarin be used?

How should AF be investigated?

Should rhythm or rate be controlled?

When should aspirin or warfarin be used?

Atrial fibrillation (AF) is the most common chronic cardiac arrhythmia. It is characterised by rapid and chaotic activation of the atria. This results in the failure of coordinated atrial contraction, predisposing the patient to the formation of intra-atrial thrombi and possible embolic stroke. Rapid conduction through the atrioventricular (AV) node results in a rapid ventricular rate that, along with the loss of the contribution of atrial contraction to ventricular filling, can result in cardiac failure.

AF is associated with considerable morbidity and mortality. The mortality rate of patients with AF is approximately double that of patients in normal sinus rhythm.1 The arrhythmia accounts for 10-15% of all ischaemic strokes, and 25% of ischaemic strokes in patients over 80 years of age.2

In 2001-2002, the prevalence of AF in Scotland was 8.7 per 1,000, increasing dramatically with age to 71 per 1000 in people aged over 85 years. The prevalence was higher among men.3

Currently there are an estimated 4.5m people in the European Union with paroxysmal or persistent AF, costing the EU around €13.5 billion per year.1 This burden is likely to increase as the population ages.

The temporal behaviour of AF is described as paroxysmal, persistent or permanent:

• Paroxysmal AF spontaneously reverts to sinus rhythm within seven days, usually within 48 hours.

• Persistent AF lasts longer than seven days or requires electrical or pharmacological intervention to restore sinus rhythm.

• AF is described as permanent when cardioversion has either failed or been deemed inappropriate.

This classification can be useful in guiding treatment, but the basic principles of treatment remain similar for each category.

In recent years the results of large, randomised trials and the emergence of new catheter-based ablation techniques have modified treatment strategies for AF. Management of this condition revolves around the choice of treatments either to control the ventricular rate or to achieve and maintain sinus rhythm – rate versus rhythm control.

These options are not necessarily mutually exclusive. A patient with paroxysmal AF may benefit from rate control during episodes and from rhythm control to reduce the frequency of symptoms. However, in persistent AF the choice is either to accept AF as the long-term rhythm and ensure rate control, or follow a rhythm control strategy to restore and maintain normal rhythm (although in the knowledge that recurrence rates are high, in which case rate control would also be advantageous).

Some drugs may provide both rate and rhythm control (for example ß-blockers, in particular sotalol, or amiodarone).

Recent guidelines covering the management of patients with AF have been published by NICE4 and European and American cardiology bodies.1

The appropriate use of anticoagulant or antiplatelet agents to reduce the risk of stroke is critical to the success of both treatment strategies.


The diagnosis of AF is suggested by an irregular pulse and is confirmed by an electrocardiogram (ECG) showing an absence of P waves and irregular QRS complexes (see figure 1, below).

Patients may present with symptoms secondary to AF. These include:

• Breathlessness

• Palpitations

• Dizziness or syncope

• Chest pain

• Stroke or transient ischaemic attack.

However, many patients with AF are asymptomatic and the diagnosis is often opportunistic. All patients who have an irregular pulse or symptoms suggestive of AF should have an ECG.

In patients with suspected paroxysmal AF, ambulatory ECG monitors may be required to confirm the diagnosis. Twenty four hour ECG monitors are recommended for patients with suspected asymptomatic recurrent AF or with symptomatic episodes separated by less than 24 hours.4 An event recorder may be used for patients with more infrequent symptomatic episodes.

AF is more likely to occur in patients with other common cardiovascular diseases, such as ischaemic heart disease, valvular heart disease, heart failure, diabetes and hypertension.

Other risk factors for developing AF include:

• Congenital heart disease

• Hyperthyroidism

• Alcohol excess

• Sepsis.

However, AF can develop without an overt underlying cause (‘lone AF').

Given the morbidity and mortality associated with AF, and the possibility of easily identifying at-risk groups, should we be screening patients for AF? The SAFE study, a randomised controlled trial involving GP practices in the West Midlands, found that opportunistic, but not systematic, screening for AF on an annual basis was a cost-effective alternative to routine practice or targeted screening of patients.5

The assessment of patients with AF requires a full history and clinical examination in order to identify additional comorbidities and assess stroke risk. Investigation of a patient with AF should include ECG and TFTs as a minimum. Excluding anaemia is relevant with respect to anticoagulant or antiplatelet therapy.

Echocardiography is indicated in selected patients. Exercise testing is not necessary, and may not be advisable until the arrhythmia is controlled. Patients with a history of chest pain should be investigated appropriately, including exercise testing once AF is controlled.

The main questions to be addressed by history and examination are related to coexisting disease, including cardiac disease. A history of coronary artery disease, rheumatic fever, the finding of hypertension or of a cardiac murmur may contribute to the risk assessment and the indications for echocardiography.

The NICE guideline recommends echocardiography in selected patients to guide treatment decisions. These include patients for whom a rhythm control strategy is being considered and patients for whom there is a high clinical suspicion of structural heart disease, such as left ventricular dysfunction, myocardial infarction, valve disease or hypertension. It may also be helpful for some patients in determining their risk of thromboembolism and the need for antithrombotic therapy (see figure 2, attached), although there is no need for echocardiography to be routinely performed on those who clearly fulfil the current recommendations for anticoagulation.4 Many hospitals provide an open access echo service to which appropriate patients can be referred.


Rate control

The aim of a rate control strategy is to minimise symptoms associated with tachycardia and reduce the risk of tachycardia-induced heart failure by reducing conduction through the AV node. Beta-blockers or rate-limiting calcium channel antagonists, such as diltiazem and verapamil, are the first-line choice of drugs recommended by NICE for ventricular rate control.

Digoxin is not recommended as monotherapy, except in elderly or very sedentary patients, as it is not good at controlling ventricular rate during exercise. It can be useful as an additional agent when adequate rate control is not achieved with a first-line drug.

Optimal rate control may vary depending on the patient, but is generally a heart rate of between 60 and 90 beats per minute at rest and <110 beats per minute during exercise.

Rhythm control

The aim of a rhythm control strategy is to minimise the symptoms and complications of AF by restoring and maintaining sinus rhythm. For patients with persistent AF this initially requires the restoration of sinus rhythm, either by electrical cardioversion or drug therapy with flecainide or amiodarone.

Patients with persistent AF (at least one week's duration) may be considered for cardioversion. Patients presenting with recent onset AF may be considered for cardioversion, as long as the duration is less than 48 hours. After 48 hours, patients should be anticoagulated for at least three weeks before cardioversion.

An alternative is to perform a transoesphageal echo to exclude an intra-atrial thrombus, but the patient should still be anticoagulated after cardioversion, for a minimum of four weeks, or indefinitely if they have moderate to high risk of thrombi.

Unfortunately there is a high rate of recurrence of AF following cardioversion, and many patients require additional treatment with long-term anti-arrhythmic drugs to maintain sinus rhythm.

The choice of anti-arrhythmic will depend on the individual patient and is particularly influenced by the presence of coexisting structural heart disease. In the absence of structural heart disease, ß-blockers are recommended, with flecainide, sotalol or amiodarone as alternatives.

For some patients with infrequent, symptomatic but haemodynamically stable paroxysms of AF, a ‘pill in the pocket' strategy may be employed. The patients can self-medicate with single doses of a drug, such as flecainide or propafenone, if it has previously been shown to be safe and effective in hospital. When structural heart disease is present this treatment strategy is not appropriate, and long-term ß-blockers or amiodarone are safer options.6

Catheter ablation of the pulmonary veins and left atrium can be used to interrupt and prevent the spread of abnormal electrical activity responsible for AF, and can be a long-term curative procedure that removes the need for continued anti-arrhythmic drug therapy. However, it is a procedure requiring a highly skilled operator, success rates are variable and procedural risks include stroke and cardiac tamponade.

The rhythm control strategy is hindered primarily by the limited efficacy and the side-effects of many of the anti-arrhythmic drugs, in particular the toxicity of long-term amiodarone therapy and the risk of inducing life-threatening ventricular arrhythmias with drugs such as sotalol or flecainide.


The potential complexities involved in this strategy and the need for cardioversion or specialised invasive treatments mean that patients for whom a rhythm control strategy is appropriate should be referred to a cardiology specialist (see table 1, above).

Rate versus rhythm control

Although there are clear advantages in maintaining sinus rhythm in terms of preventing complications and symptoms, a rate control strategy would appear to be easier to achieve, involves simpler drug choices, has fewer side-effects and requires less monitoring and specialist input.

There has been increased support for a rate control strategy recently in light of several large, randomised controlled trials comparing rate and rhythm control. The AFFIRM study showed that rate control was not inferior to rhythm control, in terms of mortality and morbidity from cardiovascular causes, for the treatment of persistent AF.7

However, there are patient groups for whom a rhythm control strategy is appropriate, preferable and has a high success rate. The NICE guideline recommends a rhythm control strategy is used first for:

• Patients with paroxysmal AF

• Young patients (? 65 years) with persistent AF

• Patients who present for the first time with lone AF

• Patients with AF secondary

to a treated precipitant

• Patients with cardiac failure.

A rate control strategy should be used for:

• All patients with permanent AF

• Patients with persistent AF who

are >65 years old or have coronary artery disease

• Patients for whom anti-arrhythmic drugs are contraindicated, or who are unsuitable for cardioversion.

Patients who remain particularly symptomatic despite adequate rate control should be considered for rhythm control, and patients in whom rhythm control fails should be rate controlled.


The assessment of stroke risk and the administration of appropriate antithrombotic treatment is arguably the most important aspect of AF management in terms of reducing morbidity and mortality. The risk of stroke is not standard for all patients with AF and varies according to the presence of additional risk factors, such as age and cardiovascular disease.

For young patients with no risk factors for stroke other than AF, the annual risk of stroke is less than 1%. However, for those with multiple risk factors the annual risk can exceed 8%.2

Warfarin and aspirin have both been shown to reduce the incidence of stroke in AF patients compared with placebo. Warfarin is more effective when compared with aspirin and reduces the risk of stroke by 35%, although it does confer a small, non-statistically significant increased risk of intracranial haemorrhage.2

Currently, NICE recommends the use of warfarin in patients at high risk of stroke, and aspirin 75-300mg for patients at low risk (see figure 2, attached). For patients at moderate risk of stroke the physician should determine the best treatment on an individual patient basis; this is a difficult area, and should usually involve a discussion with the patient. The NICE guideline states that ‘Owing to lack of sufficient clear-cut evidence, treatment may be decided on an individual basis, and the physician must balance the risks and benefits of warfarin versus aspirin. As stroke risk factors are cumulative, warfarin may, for example, be used in the presence of two or more moderate stroke risk factors. Referral and echocardiography may help in cases of uncertainty.'4 Clearly a patient's risk profile will change with time, and all patients should be reassessed on a regular basis.

Antithrombotic treatment is not limited to patients with permanent or persistent AF. Patients with paroxysmal AF are at a similar risk and should be assessed using the same algorithm.


Atrial fibrillation is an increasingly common problem. Many patients will be diagnosed in primary care, either as a result of opportunistic findings or symptomatic presentations. A large proportion of these patients can be investigated and managed using a rate control strategy without the need for specialist referral. A rhythm control treatment strategy is optimal for some patients and they should be referred to a cardiology specialist.

The assessment of stroke risk and the prompt initiation of appropriate antithrombotic treatment is a crucial component of AF management and one that is often overlooked.

The authors Authors

Dr Gillian Marshall
research fellow

Professor Andrew C Rankin
Professor of Medical Cardiology, British Heart Foundation, Glasgow Cardiovascular Research Centre and Glasgow Royal Infirmary

Figure 2: Antithrombotic treatment for AF

Antithrombotic treatment for AF

Key points Atrial fibrillation Figure 1: A 12-lead electrocardiogram (ECG) of atrial fibrillation. The ventricular rate is about 100 beats per minute and the rhythm is irregular (most clearly seen in the lower tracing, the lead-II rhythm strip). There are no P waves, but low amplitude Figure 1 AF_tab1

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say