2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS

Guidelines, Atrial fibrillation, Anticoagulation, Vitamin K antagonists, Non-vitamin K antagonist oral anticoagulants, Left atrial appendage occlusion, Rate control, Cardioversion, Rhythm control, Antiarrhythmic drugs, Upstream therapy, Catheter ablation, AF surgery, Valve repair, Pulmonary vein isolation, Left atrial ablation

The Task Force for the management of atrial fibrillation of the European Society of Cardiology (ESC) 

Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC 

Endorsed by the European Stroke Organisation (ESO) 

Table of Contents

• 1 Preamble 10

• 2 Introduction 11

• 3 Epidemiology and impact for patients 12

•  3.1 Incidence and prevalence of atrial fibrillation 12

•  3.2 Morbidity, mortality, and healthcare burden of atrial fibrillation 12

•  3.3 Impact of evidence-based management on outcomes in atrial fibrillation patients 12

•  3.4 Gender 13

• 4 Pathophysiological and genetic aspects that guide management 14

•  4.1 Genetic predisposition 14

•  4.2 Mechanisms leading to atrial fibrillation 14

•   4.2.1 Remodelling of atrial structure and ion channel function 14

•   4.2.2 Electrophysiological mechanisms of atrial fibrillation 16

• 5 Diagnosis and timely detection of atrial fibrillation 17

•  5.1 Overt and silent atrial fibrillation 17

•  5.2 Screening for silent atrial fibrillation 17

•   5.2.1 Screening for atrial fibrillation by electrocardiogram in the community 17

•   5.2.2 Prolonged monitoring for paroxysmal atrial fibrillation 17

•   5.2.3 Patients with pacemakers and implanted devices 17

•   5.2.4 Detection of atrial fibrillation in stroke survivors 18

•  5.3 Electrocardiogram detection of atrial flutter 19

• 6 Classification of atrial fibrillation 19

•  6.1 Atrial fibrillation pattern 19

•  6.2 Atrial fibrillation types reflecting different causes of the arrhythmia 20

•  6.3 Symptom burden in atrial fibrillation 20

• 7 Detection and management of risk factors and concomitant cardiovascular diseases 21

•  7.1 Heart failure 22

•   7.1.1 Patients with atrial fibrillation and heart failure with reduced ejection fraction 23

•   7.1.2 Atrial fibrillation patients with heart failure with preserved ejection fraction 24

•   7.1.3 Atrial fibrillation patients with heart failure with mid-range ejection fraction 24

•   7.1.4 Prevention of atrial fibrillation in heart failure 25

•  7.2 Hypertension 25

•   7.2.1 Treatment of hypertension to prevent incident atrial fibrillation 25

•   7.2.2 Blood pressure control in patients with atrial fibrillation 25

•  7.3 Valvular heart disease 25

•  7.4 Diabetes mellitus 26

•  7.5 Obesity and weight loss 26

•   7.5.1 Obesity as a risk factor 26

•   7.5.2 Weight reduction in obese patients with atrial fibrillation 26

•   7.5.3 Catheter ablation in obese patients 26

•  7.6 Chronic obstructive pulmonary disease, sleep apnoea, and other respiratory diseases 26

•  7.7 Chronic kidney disease 27

• 8 Integrated management of patients with atrial fibrillation 28

•  8.1 Evidence supporting integrated atrial fibrillation care 30

•  8.2 Components of integrated atrial fibrillation care 30

•   8.2.1 Patient involvement 30

•   8.2.2 Multidisciplinary atrial fibrillation teams 31

•   8.2.3 Role of non-specialists 31

•   8.2.4 Technology use to support atrial fibrillation care 31

•  8.3 Diagnostic workup of atrial fibrillation patients 31

•   8.3.1 Recommended evaluation in all atrial fibrillation patients 31

•   8.3.2 Additional investigations in selected patients with atrial fibrillation 32

•  8.4 Structured follow-up 32

•  8.5 Defining goals of atrial fibrillation management 32

• 9 Stroke prevention therapy in atrial fibrillation patients 33

•  9.1 Prediction of stroke and bleeding risk 34

•   9.1.1 Clinical risk scores for stroke and systemic embolism 34

•   9.1.2 Anticoagulation in patients with a CHA₂DS₂-VASc score of 1 in men and 2 in women 35

•   9.1.3 Clinical risk scores for bleeding 35

•  9.2 Stroke prevention 36

•   9.2.1 Vitamin K antagonists 36

•   9.2.2 Non-vitamin K antagonist oral anticoagulants 37

•   9.2.3 Non-vitamin K antagonist oral anticoagulants or vitamin K antagonists 41

•   9.2.4 Oral anticoagulation in atrial fibrillation patients with chronic kidney disease 41

•   9.2.5 Oral anticoagulation in atrial fibrillation patients on dialysis 42

•   9.2.6 Patients with atrial fibrillation requiring kidney transplantation 42

•   9.2.7 Antiplatelet therapy as an alternative to oral anticoagulants 42

•  9.3 Left atrial appendage occlusion and exclusion 43

•   9.3.1 Left atrial appendage occlusion devices 43

•   9.3.2 Surgical left atrial appendage occlusion or exclusion 43

•  9.4 Secondary stroke prevention 44

•   9.4.1 Treatment of acute ischaemic stroke 44

•   9.4.2 Initiation of anticoagulation after transient ischaemic attack or ischaemic stroke 44

•   9.4.3 Initiation of anticoagulation after intracranial haemorrhage 45

•  9.5 Strategies to minimize bleeding on anticoagulant therapy 47

•   9.5.1 Uncontrolled hypertension 47

•   9.5.2 Previous bleeding event 47

•   9.5.3 Labile international normalized ratio and adequate non-vitamin K antagonist oral anticoagulant dosing 47

•   9.5.4 Alcohol abuse 47

•   9.5.5 Falls and dementia 47

•   9.5.6 Genetic testing 47

•   9.5.7 Bridging periods off oral anticoagulation 48

•  9.6 Management of bleeding events in anticoagulated patients with atrial fibrillation 48

•   9.6.1 Management of minor, moderate, and severe bleeding 48

•   9.6.2 Oral anticoagulation in atrial fibrillation patients at risk of or having a bleeding event 49

•  9.7 Combination therapy with oral anticoagulants and antiplatelets 50

•   9.7.1 Antithrombotic therapy after acute coronary syndromes and percutaneous coronary intervention in patients requiring oral anticoagulation 51

• 10 Rate control therapy in atrial fibrillation 53

•  10.1 Acute rate control 53

•  10.2 Long-term pharmacological rate control 54

•   10.2.1 Beta-blockers 54

•   10.2.2 Non-dihydropyridine calcium channel blockers 54

•   10.2.3 Digitalis 54

•   10.2.4 Amiodarone 55

•  10.3 Heart rate targets in atrial fibrillation 55

•  10.4 Atrioventricular node ablation and pacing 56

• 11 Rhythm control therapy in atrial fibrillation 58

•  11.1 Acute restoration of sinus rhythm 58

•   11.1.1 Antiarrhythmic drugs for acute restoration of sinus rhythm (‘pharmacological cardioversion’) 58

•   11.1.2 ‘Pill in the pocket’ cardioversion performed by patients 59

•   11.1.3 Electrical cardioversion 60

•   11.1.4 Anticoagulation in patients undergoing cardioversion 60

•  11.2 Long-term antiarrhythmic drug therapy 60

•   11.2.1 Selection of antiarrhythmic drugs for long-term therapy: Safety first! 61

•   11.2.2 Twelve-lead electrocardiogram as a tool to identify patients at risk of pro-arrhythmia 62

•   11.2.3 New antiarrhythmic drugs 64

•   11.2.4 Antiarrhythmic effects of non-antiarrhythmic drugs 64

•  11.3 Catheter ablation 67

•   11.3.1 Indications 67

•   11.3.2 Techniques and technologies 67

•   11.3.3 Outcome and complications 67

•   11.3.4 Anticoagulation – before, during, and after ablation 68

•   11.3.5 Ablation of atrial fibrillation in heart failure patients 69

•   11.3.6 Follow-up after catheter ablation 69

•  11.4 Atrial fibrillation surgery 69

•   11.4.1 Concomitant atrial fibrillation surgery 69

•   11.4.2 Stand-alone rhythm control surgery 71

•  11.5 Choice of rhythm control following treatment failure 72

•  11.6 The atrial fibrillation Heart Team 72

• 12 Hybrid rhythm control therapy 74

•  12.1 Combining antiarrhythmic drugs and catheter ablation 74

•  12.2 Combining antiarrhythmic drugs and pacemakers 74

• 13 Specific situations 74

•  13.1 Frail and ‘elderly’ patients 74

•  13.2 Inherited cardiomyopathies, channelopathies, and accessory pathways 75

•   13.2.1 Wolff–Parkinson–White syndrome 75

•   13.2.2 Hypertrophic cardiomyopathy 75

•   13.2.3 Channelopathies and arrhythmogenic right ventricular cardiomyopathy 76

•  13.3 Sports and atrial fibrillation 77

•  13.4 Pregnancy 77

•   13.4.1 Rate control 77

•   13.4.2 Rhythm control 78

•   13.4.3 Anticoagulation 78

•  13.5 Post-operative atrial fibrillation 78

•   13.5.1 Prevention of post-operative atrial fibrillation 78

•   13.5.2 Anticoagulation 79

•   13.5.3 Rhythm control therapy in post-operative atrial fibrillation 79

•  13.6 Atrial arrhythmias in grown-up patients with congenital heart disease 80

•   13.6.1 General management of atrial arrhythmias in grown-up patients with congenital heart disease 80

•   13.6.2 Atrial tachyarrhythmias and atrial septal defects 80

•   13.6.3 Atrial tachyarrhythmias after Fontan operation 80

•   13.6.4 Atrial tachyarrhythmias after tetralogy of Fallot correction 80

•  13.7 Management of atrial flutter 81

• 14 Patient involvement, education, and self-management 82

•  14.1 Patient-centred care 82

•  14.2 Integrated patient education 82

•  14.3 Self-management and shared decision-making 82

• 15 Gaps in evidence 83

•  15.1 Major health modifiers causing atrial fibrillation 83

•  15.2 How much atrial fibrillation constitutes a mandate for therapy? 83

•  15.3 Atrial high-rate episodes and need for anticoagulation 83

•  15.4 Stroke risk in specific populations 83

•  15.5 Anticoagulation in patients with severe chronic kidney disease 83

•  15.6 Left atrial appendage occlusion for stroke prevention 83

•  15.7 Anticoagulation in atrial fibrillation patients after a bleeding or stroke event 83

•  15.8 Anticoagulation and optimal timing of non-acute cardioversion 84

•  15.9 Competing causes of stroke or transient ischaemic attack in atrial fibrillation patients 84

•  15.10 Anticoagulation in patients with biological heart valves (including transcatheter aortic valve implantation) and non-rheumatic valve disease 84

•  15.11 Anticoagulation after ‘successful’ catheter ablation 84

•  15.12 Comparison of rate control agents 84

•  15.13 Catheter ablation in persistent and long-standing persistent AF 85

•  15.14 Optimal technique for repeat catheter ablation 85

•  15.15 Combination therapy for maintenance of sinus rhythm 85

•  15.16 Can rhythm control therapy convey a prognostic benefit in atrial fibrillation patients? 85

•  15.17 Thoracoscopic ‘stand-alone’ atrial fibrillation surgery 85

•  15.18 Surgical exclusion of the left atrial appendage 85

•  15.19 Concomitant atrial fibrillation surgery 85

• 16 To do and not to do messages from the Guidelines 86

• 17 A short summary of the management of AF patients 89

• 18 Web Addenda 90

• 19 Appendix 90

• 20 References 91

Abbreviations and acronyms

 
• ABC

  age, biomarkers, clinical history

 
• ACE

  angiotensin-converting enzyme

 
• ACS

  acute coronary syndromes

 
• AF

  atrial fibrillation

 
• AFFIRM

  Atrial Fibrillation Follow-up Investigation of Rhythm Management

 
• AFNET

  German Competence NETwork on Atrial Fibrillation

 
• AngII

  angiotensin II

 
• AHRE

  atrial high rate episodes

 
• APACHE-AF

  Apixaban versus Antiplatelet drugs or no antithrombotic drugs after anticoagulation-associated intraCerebral HaEmorrhage in patients with Atrial Fibrillation

 
• ARB

  angiotensin receptor blocker

 
• ARISTOTLE

  Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation

 
• ARNI

  angiotensin receptor neprilysin inhibition

 
• ARTESiA

  Apixaban for the Reduction of Thrombo-Embolism in Patients With Device-Detected Sub-Clinical Atrial Fibrillation

 
• ATRIA

  AnTicoagulation and Risk factors In Atrial fibrillation

 
• AV

  Atrioventricular

 
• AXAFA

  Anticoagulation using the direct factor Xa inhibitor apixaban during Atrial Fibrillation catheter Ablation: Comparison to vitamin K antagonist therapy

 
• BAFTA

  Birmingham Atrial Fibrillation Treatment of the Aged Study

 
• BMI

  body mass index

 
• b.p.m.

  beats per minute

 
• CABANA

  Catheter Ablation versus Antiarrhythmic Drug Therapy for Atrial Fibrillation Trial

 
• CABG

  coronary artery bypass graft

 
• CAD

  coronary artery disease

 
• CHA₂DS₂-VASc

  Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female)

 
• CHADS₂

  Cardiac failure, Hypertension, Age, Diabetes, Stroke (Doubled)

 
• CI

  confidence interval

 
• CKD

  chronic kidney disease

 
• CPG

  Committee for Practice Guidelines

 
• CrCl

  creatinine clearance

 
• CT

  computed tomography

 
• CV

  cardiovascular

 
• CYP2D6

  cytochrome P450 2D6

 
• CYP3A4

  cytochrome P450 3A4

 
• DIG

  Digitalis Investigation Group

 
• EACTS

  European Association for Cardio-Thoracic Surgery

 
• EAST

  Early treatment of Atrial fibrillation for Stroke prevention Trial

 
• ECG

  electrocardiogram/electrocardiography

 
• EHRA

  European Heart Rhythm Association

 
• ENGAGE AF-TIMI 48

  Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48

 
• EORP

  EURObservational Research Programme

 
• ESC

  European Society of Cardiology

 
• ESO

  European stroke Organisation

 
• FAST

  Atrial Fibrillation Catheter Ablation vs. Surgical Ablation Treatment

 
• FEV1

  forced expiratory volume in 1 s

 
• FFP

  four-factor prothrombin complex concentrates

 
• FXII

  factor XII

 
• GDF-15

  growth differentiation factor 15

 
• GFR

  glomerular filtration rate

 
• GUCH

  grown-up congenital heart disease

 
• HARMONY

  A Study to Evaluate the Effect of Ranolazine and Dronedarone When Given Alone and in Combination in Patients With Paroxysmal Atrial Fibrillation

 
• HAS-BLED

  hypertension, abnormal renal/liver function (1 point each), stroke, bleeding history or predisposition, labile INR, elderly (>65 years), drugs/alcohol concomitantly (1 point each)

 
• HEMORR₂HAGES

  Hepatic or renal disease, ethanol abuse, malignancy history, older age >75, reduced platelet count/function/antiplatelet, rebleeding risk (scores double), hypertension (uncontrolled), anaemia, genetic factors, excessive fall risk, stroke history

 
• HF

  heart failure

 
• HFmrEF

  heart failure with mid-range ejection fraction

 
• HFpEF

  heart failure with preserved ejection fraction

 
• HFrEF

  heart failure with reduced ejection fraction

 
• HR

  hazard ratio

 
• ICD

  implantable cardioverter defibrillator

 
• IHD

  ischaemic heart disease

 
• IL-6

  interleukin 6

 
• INR

  international normalized ratio

 
• i.v.

  intravenous

 
• LA

  left atrium/atrial

 
• LAA

  left atrial appendage

 
• LAAOS

  Left Atrial Appendage Occlusion Study

 
• LV

  left ventricular

 
• LVEF

  left ventricular ejection fraction

 
• LVH

  left ventricular hypertrophy

 
• MANTRA-PAF

  Medical ANtiarrhythmic Treatment or Radiofrequency Ablation in Paroxysmal Atrial Fibrillation

 
• MERLIN

  Metabolic Efficiency With Ranolazine for Less Ischemia in Non ST-Elevation Acute Coronary Syndromes

 
• MRA

  Mineralocorticoid receptor antagonist

 
• MRI

  magnetic resonance imaging

 
• NIHSS

  National Institutes of Health stroke severity scale

 
• NOAC

  non-vitamin K antagonist oral anticoagulant

 
• NOAH

  Non vitamin K antagonist Oral anticoagulants in patients with Atrial High rate episodes (NOAH)

 
• NYHA

  New York Heart Association

 
• OAC

  oral anticoagulation/oral anticoagulant

 
• OR

  odds ratio

 
• ORBIT

  Outcomes Registry for Better Informed Treatment of Atrial Fibrillation

 
• PAFAC

  Prevention of Atrial Fibrillation After Cardioversion trial

 
• PAI-1

  plasminogen activator inhibitor 1

 
• PCI

  percutaneous coronary intervention

 
• PCC

  prothrombin complex concentrates

 
• PICOT

  Population, Intervention, Comparison, Outcome, Time

 
• PREVAIL

  Prospective Randomized Evaluation of the Watchman LAA Closure Device In Patients with AF Versus Long Term Warfarin Therapy trial

 
• PROTECT AF

  Watchman Left Atrial Appendage System for Embolic Protection in Patients With AF trial

 
• PUFA

  polyunsaturated fatty acid

 
• PVI

  pulmonary vein isolation

 
• QoL

  quality of life

 
• RACE

  Rate Control Efficacy in Permanent Atrial Fibrillation

 
• RATE-AF

  Rate Control Therapy Evaluation in Permanent Atrial Fibrillation

 
• RCT

  randomized controlled trial

 
• RE-CIRCUIT

  Randomized Evaluation of Dabigatran Etexilate Compared to warfarIn in pulmonaRy Vein Ablation: Assessment of an Uninterrupted periproCedUral antIcoagulation sTrategy

 
• RE-LY

  Randomized Evaluation of Long-Term Anticoagulation Therapy

 
• RF

  radiofrequency

 
• ROCKET-AF

  Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation

 
• RR

  risk ratio

 
• rtPA

  recombinant tissue plasminogen activator

 
• SAMe-TT₂R₂

  Sex (female), age (<60 years), medical history (two of the following: hypertension, diabetes, mi, pad, congestive heart failure, history of stroke, pulmonary disease, hepatic or renal disease), treatment (interacting medications e.g. amiodarone), tobacco use (within 2 years; scores double), race (non-Caucasian; scores double)

 
• SD

  standard deviation

 
• SPAF

  Stroke Prevention in Atrial Fibrillation

 
• SR

  sinus rhythm

 
• TF

  tissue factor

 
• TIA

  transient ischaemic attack

 
• TIMI

  Thrombolysis in Myocardial Infarction

 
• TOE

  transoesophageal echocardiography

 
• TTR

  time in therapeutic range

 
• UFH

  unfractionated heparin

 
• VKA

  vitamin K antagonist

 
• VT

  Ventricular tachycardia

 
• VVI

  Ventricular pacing, ventricular sensing, inhibited response pacemaker

 
• WOEST

  What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing

 
• WPW

  Wolff-Parkinson-White syndrome

1. Preamble

Guidelines summarize and evaluate all available evidence on a particular issue at the time of the writing process, with the aim of assisting health professionals in selecting the best management strategies for an individual patient with a given condition, taking into account the impact on outcome, as well as the risk–benefit ratio of particular diagnostic or therapeutic means. Guidelines and recommendations should help health professionals to make decisions in their daily practice. However, the final decisions concerning an individual patient must be made by the responsible health professional(s) in consultation with the patient and caregiver as appropriate.

A great number of Guidelines have been issued in recent years by the European Society of Cardiology (ESC) and by the European Association for Cardio-Thoracic Surgery (EACTS), as well as by other societies and organisations. Because of the impact on clinical practice, quality criteria for the development of guidelines have been established in order to make all decisions transparent to the user. The recommendations for formulating and issuing ESC Guidelines can be found on the ESC website (http://www.escardio.org/Guidelines-&-Education/Clinical-Practice-Guidelines/Guidelines-development/Writing-ESC-Guidelines). ESC Guidelines represent the official position of the ESC on a given topic and are regularly updated.

Members of this Task Force were selected by the ESC, including representation from the European Heart Rhythm Association (EHRA), and EACTS as well as by the European Stroke Organisation (ESO) to represent professionals involved with the medical care of patients with this pathology. Selected experts in the field undertook a comprehensive review of the published evidence for management (including diagnosis, treatment, prevention and rehabilitation) of a given condition according to ESC Committee for Practice Guidelines (CPG) policy and approved by the EACTS and ESO. A critical evaluation of diagnostic and therapeutic procedures was performed, including assessment of the risk–benefit ratio. Estimates of expected health outcomes for larger populations were included, where data exist. The level of evidence and the strength of the recommendation of particular management options were weighed and graded according to predefined scales, as outlined in Tables 1 and 2.

Table 1

Classes of recommendations

Table 2

Levels of evidence

The experts of the writing and reviewing panels provided declaration of interest forms for all relationships that might be perceived as real or potential sources of conflicts of interest. These forms were compiled into one file and can be found on the ESC website (http://www.escardio.org/guidelines). Any changes in declarations of interest that arise during the writing period must be notified to the ESC and EACTS and updated. The Task Force received its entire financial support from the ESC and EACTS without any involvement from the healthcare industry.

The ESC CPG supervises and co-ordinates the preparation of new Guidelines produced by task forces, expert groups or consensus panels. The Committee is also responsible for the endorsement process of these Guidelines. The ESC Guidelines undergo extensive review by the CPG and external experts, and in this case by EACTS and ESO-appointed experts. After appropriate revisions the Guidelines are approved by all the experts involved in the Task Force. The finalized document is approved by the CPG, EACTS and ESO for publication in the European Heart Journal, Europace, and in the European Journal of Cardio-Thoracic Surgery. The Guidelines were developed after careful consideration of the scientific and medical knowledge and the evidence available at the time of their dating.

The task of developing ESC and EACTS Guidelines covers not only integration of the most recent research, but also the creation of educational tools and implementation programmes for the recommendations. To implement the guidelines, condensed pocket guideline versions, summary slides, booklets with essential messages, summary cards for non-specialists and an electronic version for digital applications (smartphones, etc.) are produced. These versions are abridged and thus, if needed, one should always refer to the full text version, which is freely available on the ESC website. The National Societies of the ESC are encouraged to endorse, translate and implement all ESC Guidelines. Implementation programmes are needed because it has been shown that the outcome of disease may be favourably influenced by the thorough application of clinical recommendations.

Surveys and registries are needed to verify that real-life daily practice is in keeping with what is recommended in the guidelines, thus completing the loop between clinical research, writing of guidelines, disseminating them and implementing them into clinical practice.

Health professionals are encouraged to take the ESC and EACTS Guidelines fully into account when exercising their clinical judgment, as well as in the determination and the implementation of preventive, diagnostic or therapeutic medical strategies. However, the ESC and EACTS Guidelines do not override in any way whatsoever the individual responsibility of health professionals to make appropriate and accurate decisions in consideration of each patient's health condition and in consultation with that patient and the patient's caregiver where appropriate and/or necessary. It is also the health professional's responsibility to verify the rules and regulations applicable to drugs and devices at the time of prescription.

2. Introduction

Despite good progress in the management of patients with atrial fibrillation (AF), this arrhythmia remains one of the major causes of stroke, heart failure, sudden death, and cardiovascular morbidity in the world. Furthermore, the number of patients with AF is predicted to rise steeply in the coming years. To meet the growing demand for effective care of patients with AF, new information is continually generated and published, and the last few years have seen substantial progress. Therefore, it seems timely to publish this 2^nd edition of the ESC guidelines on AF.

Reflecting the multidisciplinary input into the management of patients with AF, the Task Force includes cardiologists with varying subspecialty expertise, cardiac surgeons, stroke neurologists, and specialist nurses amongst its members. Supplementing the evidence review as outlined in the preamble, this Task Force defined three Population, Intervention, Comparison, Outcome, Time (PICOT) questions on relevant topics for the guidelines. The ESC commissioned external systematic reviews to answer these questions, and these reviews have informed specific recommendations.

Further to adhering to the standards for generating recommendations that are common to all ESC guidelines (see preamble), this Task Force discussed each draft recommendation during web-based conference calls dedicated to specific chapters, followed by consensus modifications and an online vote on each recommendation. Only recommendations that were supported by at least 75% of the Task Force members were included in the guidelines.

We hope that these guidelines will help to deliver good care to all patients with AF based on the current state-of-the-art evidence in 2016.

3. Epidemiology and impact for patients

3.1 Incidence and prevalence of atrial fibrillation

In 2010, the estimated numbers of men and women with AF worldwide were 20.9 million and 12.6 million, respectively, with higher incidence and prevalence rates in developed countries.^1,2 One in four middle-aged adults in Europe and the US will develop AF.^3–5 By 2030, 14–17 million AF patients are anticipated in the European Union, with 120 000–215 000 newly diagnosed patients per year.^2,6,7 Estimates suggest an AF prevalence of approximately 3% in adults aged 20 years or older,^8,9 with greater prevalence in older persons¹ and in patients with conditions such as hypertension, heart failure, coronary artery disease (CAD), valvular heart disease, obesity, diabetes mellitus, or chronic kidney disease (CKD).^7,10–15 The increase in AF prevalence can be attributed both to better detection of silent AF^16–18, alongside increasing age and conditions predisposing to AF.¹⁹

3.2 Morbidity, mortality, and healthcare burden of atrial fibrillation

AF is independently associated with a two-fold increased risk of all-cause mortality in women and a 1.5-fold increase in men^20–22 (Table 3). Death due to stroke can largely be mitigated by anticoagulation, while other cardiovascular deaths, for example due to heart failure and sudden death, remain common even in AF patients treated according to the current evidence base.²³ AF is also associated with increased morbidity, such as heart failure and stroke.^21,24,25 Contemporary studies show that 20–30% of patients with an ischaemic stroke have AF diagnosed before, during, or after the initial event.^17,26,27 White matter lesions in the brain, cognitive impairment,^28–30 decreased quality of life,^31,32 and depressed mood³³ are common in AF patients, and between 10–40% of AF patients are hospitalized each year.^23,34,35

Table 3

Cardiovascular morbidity and mortality associated with atrial fibrillation

AF = atrial fibrillation; LV = left ventricular.

The direct costs of AF already amount to approximately 1% of total healthcare spending in the UK, and between 6.0–26.0 billion US dollars in the US for 2008,^36,37 driven by AF-related complications (e.g. stroke) and treatment costs (e.g. hospitalizations). These costs will increase dramatically unless AF is prevented and treated in a timely and effective manner.

3.3 Impact of evidence-based management on outcomes in atrial fibrillation patients

Figure 1 depicts the major milestones in the management of AF. Despite these advances, substantial morbidity remains. Oral anticoagulation (OAC) with vitamin K antagonists (VKAs) or non-VKA oral anticoagulants (NOACs) markedly reduces stroke and mortality in AF patients.^38,39 Other interventions such as rhythm control and rate control improve AF-related symptoms and may preserve cardiac function, but have not demonstrated a reduction in long-term morbidity or mortality.^40,41

Figure 1

Timeline of findings from landmark trials in atrial fibrillation management, including treatment of concomitant conditions and prevention (green), anticoagulation (blue), rate control therapy (orange), rhythm control therapy (red), and atrial fibrillation surgery (purple).

In contemporary, well-controlled, randomized clinical trials in AF, the average annual stroke rate is about 1.5% and the annualized death rate is around 3% in anticoagulated AF patients.⁴⁰ In real life, the annual mortality can be different (both higher and lower).⁴² A minority of these deaths are related to stroke, while sudden cardiac death and death from progressive heart failure are more frequent, emphasizing the need for interventions beyond anticoagulation.^43,44 Furthermore, AF is also associated with high rates of hospitalization, commonly for AF management, but often also for heart failure, myocardial infarction, and treatment-associated complications.^34,45

3.4 Gender

In both developed and developing countries, the age-adjusted incidence and prevalence of AF are lower in women, while the risk of death in women with AF is similar to or higher than that in men with AF.^1,46,47 Female AF patients who have additional stroke risk factors (particularly older age) are also at greater risk than men of having a stroke,^48,49 even those anticoagulated with warfarin⁵⁰ (see Chapter 9 for details). Women with diagnosed AF can be more symptomatic than men and are typically older with more comorbidities.^51,52 Bleeding risk on anticoagulation is similar in both sexes,^49,50,53 but women appear less likely to receive specialist care and rhythm control therapy,⁵⁴ while the outcomes of catheter ablation or AF surgery are comparable to those in men.^55,56 These observations highlight the need to offer effective diagnostic tools and therapeutic management equally to women and men.

Recommendations relating to gender

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

4. Pathophysiological and genetic aspects that guide management

4.1 Genetic predisposition

AF, especially early-onset AF, has a strong heritable component that is independent of concomitant cardiovascular conditions.^58,59 A few young AF patients suffer from inherited cardiomyopathies or channelopathies mediated by disease-causing mutations. These monogenic diseases also convey a risk for sudden death (see Chapter 6). Up to one-third of AF patients carry common genetic variants that predispose to AF, albeit with a relatively low added risk. At least 14 of these common variants, often single nucleotide polymorphisms, are known to increase the risk of prevalent AF in populations.^60–62 The most important variants are located close to the paired-like homeodomain transcription factor 2 (Pitx2) gene on chromosome 4q25.^63,64 These variants modify the risk of AF up to seven-fold.⁶⁴ Several of the AF risk variants are also associated with cardioembolic or ischaemic stroke, possibly due to silent AF (see chapter 5 and 5.2).^62,65,66 Changes in atrial action potential characteristics,^67–70 atrial remodelling, and modified penetration of rare gene defects⁶¹ have been suggested as potential mechanisms mediating increased AF risk in carriers of common gene variants. Genetic variants could, in the future, become useful for patient selection of rhythm or rate control.^71–74 While genomic analysis may provide an opportunity to improve the diagnosis and management of AF in the future,^75,76 routine genetic testing for common gene variants associated with AF cannot be recommended at present.⁷⁷

4.2 Mechanisms leading to atrial fibrillation

4.2.1 Remodelling of atrial structure and ion channel function

External stressors such as structural heart disease, hypertension, possibly diabetes, but also AF itself induce a slow but progressive process of structural remodelling in the atria (Figure 2). Activation of fibroblasts, enhanced connective tissue deposition, and fibrosis are the hallmarks of this process.^78–80 In addition, atrial fatty infiltration, inflammatory infiltrates, myocyte hypertrophy, necrosis, and amyloidosis are found in AF patients with concomitant conditions predisposing to AF.^81–84 Structural remodelling results in electrical dissociation between muscle bundles and local conduction heterogeneities,⁸⁵ favouring re-entry and perpetuation of the arrhythmia.⁸⁶ In many patients, the structural remodelling process occurs before the onset of AF.⁷⁸ As some of the structural remodelling will be irreversible, early initiation of treatment seems desirable.⁸⁷Table 4 gives an overview of the most relevant pathophysiological alterations in atrial tissue associated with AF, and lists corresponding clinical conditions that can contribute to these changes.

Table 4

Pathophysiological alterations in atrial tissue associated with atrial fibrillation and clinical conditions that could contribute to such alterations

AF = atrial fibrillation; CAD = coronary artery disease.

Figure 2

Major mechanisms causing atrial fibrillation that can be considered when choosing therapy. The various aetiological factors (left) cause a complex array of pathophysiological changes in the atria, including stretch-induced atrial fibrosis, hypocontractility, fatty infiltration, inflammation, vascular remodelling, ischaemia, ion channel dysfunction, and Ca²⁺-instability. These changes enhance both ectopy and conduction disturbances, increasing the propensity of the atria to develop or maintain AF. At the same time, some of these alterations are involved in the occurrence of the hypercoagulable state associated with AF. For example, hypocontractility reduces local endothelial shear stress, which increases PAI-1 expression, and ischaemia-induced inflammation enhances the expression of endothelial adhesion molecules or promotes shedding of endothelial cells, resulting in tissue factor exposure to the blood stream. These changes contribute to the thrombogenic milieu in the atria of AF patients. AF in itself can aggravate many of the mechanisms shown, which may explain the progressive nature of the arrhythmia.

The functional and structural changes in atrial myocardium and stasis of blood, especially in the left atrial appendage (LAA), generate a prothrombotic milieu. Furthermore, even short episodes of AF lead to atrial myocardial damage and the expression of prothrombotic factors on the atrial endothelial surface, alongside activation of platelets and inflammatory cells, and contribute to a generalized prothrombotic state.^88,89 The atrial and systemic activation of the coagulation system can partially explain why short episodes of AF convey a long-term stroke risk.

4.2.2 Electrophysiological mechanisms of atrial fibrillation

AF provokes a shortening of the atrial refractory period and AF cycle length during the first days of the arrhythmia, largely due to downregulation of the Ca²⁺-inward current and upregulation of inward rectifier K⁺ currents.^94,95 Structural heart disease, in contrast, tends to prolong the atrial refractory period, illustrating the heterogeneous nature of mechanisms that cause AF in different patients.⁹⁶ Hyperphosphorylation of various Ca²⁺-handling proteins may contribute to enhanced spontaneous Ca²⁺ release events and triggered activity,^97,98 thus causing ectopy and promoting AF. Although the concept of Ca²⁺-handling instability has been challenged recently,^106,107 it may mediate AF in structurally remodelled atria and explain how altered autonomic tone can generate AF.^80,105

4.2.2.1 Focal initiation and maintenance of atrial fibrillation

The seminal observation by Haissaguerre et al.¹⁰⁸ was that a focal source in the pulmonary veins can trigger AF, and ablation of this source can suppress recurrent AF. The mechanism of focal activity might involve both triggered activity and localized reentry.^109,110 Hierarchic organization of AF with rapidly activated areas driving the arrhythmia has been documented in patients with paroxysmal AF,^111,112 but is less obvious in unselected patients with persistent AF.¹¹³

4.2.2.2 The multiple wavelet hypothesis and rotors as sources of atrial fibrillation

Moe and Abildskov¹¹⁴ proposed that AF can be perpetuated by continuous conduction of several independent wavelets propagating through the atrial musculature in a seemingly chaotic manner. As long as the number of wavefronts does not decline below a critical level, they will be capable of sustaining the arrhythmia. Numerous experimental and clinical observations can be reconciled with the multiple wavelet hypothesis.¹¹⁵ All localized sources of AF (ectopic foci, rotors, or other stable re-entry circuits) cause fibrillatory conduction remote from the source, which is difficult to distinguish from propagation sustaining AF by multiple wavelets, and either of these phenomena may generate ‘rotors’ picked up by intracardiac^116,117 or body surface¹¹⁷ recordings.

5. Diagnosis and timely detection of atrial fibrillation

5.1 Overt and silent atrial fibrillation

The diagnosis of AF requires rhythm documentation using an electrocardiogram (ECG) showing the typical pattern of AF: Absolutely irregular RR intervals and no discernible, distinct P waves. ECG-documented AF was the entry criterion in trials forming the evidence for these guidelines. By accepted convention, an episode lasting at least 30 s is diagnostic. Individuals with AF may be symptomatic or asymptomatic (‘silent AF’). Many AF patients have both symptomatic and asymptomatic episodes of AF.^118–121

Silent, undetected AF is common,^120,122 with severe consequences such as stroke and death.^123–125 Prompt recording of an ECG is an effective and cost-effective method to document chronic forms of AF.¹²⁶ The technology to detect paroxysmal, self-terminating AF episodes is rapidly evolving (see Chapter 6.1 for a definition of AF patterns). There is good evidence that prolonged ECG monitoring enhances the detection of undiagnosed AF, e.g. monitoring for 72 h after a stroke,^27,127 or even longer periods.^18,128 Daily short-term ECG recordings increase AF detection in populations over 75 years of age¹²⁹ (Web Figure 1). Ongoing studies will determine whether such early detection alters management (e.g. initiation of anticoagulation) and improves outcomes.

Once the ECG diagnosis of AF has been established, further ECG monitoring can inform management in the context of: (1) a change in symptoms or new symptoms; (2) suspected progression of AF; (3) monitoring of drug effects on ventricular rate; and (4) monitoring of antiarrhythmic drug effects or catheter ablation for rhythm control.

5.2 Screening for silent atrial fibrillation

5.2.1 Screening for atrial fibrillation by electrocardiogram in the community

Undiagnosed AF is common, especially in older populations and in patients with heart failure.¹³⁰ Opportunistic screening for silent AF seems cost-effective in elderly populations (e.g. > 65 years),¹³¹ and similar effects have been reported using single-lead ECG screening in other at-risk populations.^132,133 Screening of older populations (mean age 64 years) yielded a prevalence of 2.3% for chronic forms of AF in 122,571 participants using either short-term ECG or pulse palpation (followed by ECG in those with an irregular pulse).¹³⁴ Previously undiagnosed AF was found in 1.4% of those aged >65 years, suggesting a number needed to screen of 70. These findings encourage the further evaluation of systematic AF screening programmes in at-risk populations.

5.2.2 Prolonged monitoring for paroxysmal atrial fibrillation

Paroxysmal AF is often missed.¹²⁰ Repeated daily ECG recordings increased the detection of silent, asymptomatic paroxysmal AF in an unselected Swedish population aged >75 years.^120,135 Several patient-operated devices^136,137 and extended continuous ECG monitoring using skin patch recorders¹³⁸ have been validated for the detection of paroxysmal AF (Web Figure 1).¹³⁹ The detection of asymptomatic AF by new technologies, such as smartphone cases with ECG electrodes, smart watches, and blood pressure machines with AF detection algorithms, has not yet been formally evaluated against an established arrhythmia detection method.¹⁴⁰

5.2.3 Patients with pacemakers and implanted devices

Implanted pacemakers or defibrillators with an atrial lead allow continuous monitoring of atrial rhythm. Using this technology, patients with atrial high rate episodes (AHRE) can be identified. Depending on the risk profile of the population studied, such AHRE are detected in 10–15% of pacemaker patients.¹⁴¹ AHRE are associated with an increased risk of overt AF [hazard ratio (HR) 5.56; 95% confidence interval (CI) 3.78–8.17; P < 0.001] and ischaemic stroke or systemic embolism (HR 2.49; 95% CI 1.28–4.85; P = 0.007). The stroke risk in AHRE patients seems lower than the stroke risk in patients with diagnosed AF, and not all AHRE represent AF.¹⁴² Strokes often occur without AHRE detected within 30 days before the event.^143–147 Consequently, it is unclear whether AHRE imply the same therapeutic requirements as overt AF,¹⁴⁸ and the benefit of OAC in patients with AHRE is tested in ongoing clinical trials [e.g. Apixaban for the Reduction of Thrombo-Embolism in Patients With Device-Detected Sub-Clinical Atrial Fibrillation (ARTESiA) (NCT01938248) and Non vitamin K antagonist Oral anticoagulants in patients with Atrial High rate episodes (NOAH – AFNET 6) (NCT02618577)]. At present, pacemakers and implanted devices should be interrogated on a regular basis for AHRE, and patients with AHRE should undergo further assessment of stroke risk factors and for overt AF, including ECG monitoring (Figure 3).¹⁴⁹

Figure 3

Management of AHRE detected by an implanted device.

5.2.4 Detection of atrial fibrillation in stroke survivors

Sequential stratified ECG monitoring detected AF in 24% (95% CI 17–31) of stroke survivors,¹⁵¹ and in 11.5% (95% CI 8.9%–14.3%) in another meta-analysis,¹⁷ with large variations depending on the timing, duration, and method of monitoring. AF detection is not uncommon in unselected stroke patients (6.2%, 95% CI 4.4–8.3),¹²⁸ but is more likely in patients with cryptogenic stroke implanted with loop recorders or who have had ECG monitors for several weeks.^18,128,152 Cryptogenic stroke is defined as a stroke in which the cause could not be identified after extensive investigations.¹⁵³ A broader definition is embolic stroke of undetermined source.¹⁵⁴ Several studies have also found AF in patients in whom another competing cause for stroke has been identified clinically (e.g. hypertension or carotid artery stenosis).^27,127 Hence, prolonged ECG monitoring seems reasonable in all survivors of an ischaemic stroke without an established diagnosis of AF.

Recommendations for screening for atrial fibrillation

AF = atrial fibrillation; AHRE = atrial high rate episodes; ECG = electrocardiogram; ICD = implantable cardioverter defibrillator; TIA = transient ischaemic attack.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

5.3 Electrocardiogram detection of atrial flutter

Right atrial isthmus-dependent flutter has a typical ECG pattern and ventricular rate.¹⁵⁸ The prevalence of atrial flutter is less than one-tenth of the prevalence of AF.¹⁵⁹ Atrial flutter often coexists with or precedes AF.¹⁶⁰ In typical, isthmus-dependent flutter, P waves will often show a ‘saw tooth’ morphology, especially in the inferior leads (II, III, aVF). The ventricular rate can be variable (usual ratio of atrial to ventricular contraction 4:1 to 2:1, in rare cases 1:1) and macro-re-entrant tachycardias may be missed in stable 2:1 conduction. Vagal stimulation or intravenous adenosine can therefore be helpful to unmask atrial flutter. The management of atrial flutter is discussed in chapter 13.7. Left or right atrial macro re-entrant tachycardia is mainly found in patients after catheter ablation for AF, AF surgery, or after open heart surgery.¹⁵⁸

6. Classification of atrial fibrillation

6.1 Atrial fibrillation pattern

In many patients, AF progresses from short, infrequent episodes to longer and more frequent attacks. Over time, many patients will develop sustained forms of AF. In a small proportion of patients, AF will remain paroxysmal over several decades (2–3% of AF patients).¹⁶¹ The distribution of paroxysmal AF recurrences is not random, but clustered.¹⁶² AF may also regress from persistent to paroxysmal AF. Furthermore, asymptomatic recurrences of AF are common in patients with symptomatic AF.¹²⁰

Based on the presentation, duration, and spontaneous termination of AF episodes, five types of AF are traditionally distinguished: first diagnosed, paroxysmal, persistent, long-standing persistent, and permanent AF (Table 5). If patients suffer from both paroxysmal and persistent AF episodes, the more common type should be used for classification. Clinically determined AF patterns do not correspond well to the AF burden measured by long-term ECG monitoring.¹⁶³ Even less is known about the response to therapy in patients with long-standing persistent AF or long-standing paroxysmal AF. Despite these inaccuracies, the distinction between paroxysmal and persistent AF has been used in many trials and therefore still forms the basis of some recommendations.

Table 5

Patterns of atrial fibrillation

AF = atrial fibrillation.

^aThe distinction between paroxysmal and persistent AF is often not made correctly without access to long-term monitoring.¹⁶³ Hence, this classification alone is often insufficient to select specific therapies. If both persistent and paroxysmal episodes are present, the predominant pattern should guide the classification.

There is some evidence suggesting that AF burden may influence stroke risk^44,124,164 and could modify the response to rhythm control therapy.^76,165 The evidence for this is weak. Therefore, AF burden should not be a major factor in deciding on the usefulness of an intervention that is deemed suitable for other reasons.

6.2 Atrial fibrillation types reflecting different causes of the arrhythmia

The risk of developing AF is increased in a variety of physiological and disease states (Figure 2), and the historic term ‘lone AF’ is probably misleading and should be avoided.¹⁶⁶ Although the pattern of AF may be the same, the mechanisms underpinning AF vary substantially between patients¹⁶⁷ (Table 6). This suggests that stratifying AF patients by underlying drivers of AF could inform management, for example, considering cardiac and systemic comorbidity (e.g. diabetes and obesity¹⁶⁸), lifestyle factors (e.g. activity level, smoking, alcohol intake^169,170), markers of cardiac structural remodelling (e.g. fibrosis^171–173 or electrocardiographic parameters of AF complexity¹⁷⁴), or genetic background. Table 6 provides such a taxonomy, informed by expert consensus,^76,120,175 but without much evidence to underpin its clinical use.¹⁷⁶ Systematic research defining the major drivers of AF is clearly needed to better define different types of AF.¹⁷⁶

Table 6

Clinical types of atrial fibrillation^a

AF = atrial fibrillation; LV = left ventricular; LVH = left ventricular hypertrophy. It is recognized that these types of AF will overlap in clinical practice, and that their impact for management needs to be evaluated systematically.

6.3 Symptom burden in atrial fibrillation

Patients with AF have significantly poorer quality of life than healthy controls, experiencing a variety of symptoms including lethargy, palpitations, dyspnoea, chest tightness, sleeping difficulties, and psychosocial distress.^32,177–180 Improved quality of life has been noted with both pharmacological and interventional therapies,^181–185 but there are limited data to compare the benefit of different treatments.^32,186 Assessment of quality of life is further constrained by a lack of cross-validation of the several AF-specific quality of life tools.^187–191 With regard to symptom assessment, EHRA suggested the EHRA symptom scale (Table 7) to describe symptom severity in AF patients.¹⁹² A similar scale (the Canadian Cardiovascular Society Severity of Atrial Fibrillation Scale) is used in Canada.¹⁹³ The EHRA scale has been used and validated.^194–199 A modification was proposed in 2014, subdividing EHRA class 2 into mild (2a) or moderate (2b) impact.¹⁹⁹ As symptoms in class 2b (‘troubling’ symptoms) identified patients with a health utility benefit of rhythm control in that study, this modification may provide a threshold for potential treatment decisions, pending independent validation. While some AF patients had no or minimal symptoms (25–40%), many (15–30%) report severe or disabling symptoms.^194,196 The modified EHRA scale should be used to guide symptom-orientated treatment decisions and for longitudinal patient profiling.

Table 7

Modified European Heart Rhythm Association symptom scale (modified from Wynn et al.¹⁹⁹)

AF = atrial fibrillation; EHRA = European Heart Rhythm Association.

^aEHRA class 2a and 2b can be differentiated by evaluating whether patients are functionally affected by their AF symptoms. AF-related symptoms are most commonly fatigue/tiredness and exertional shortness of breath, or less frequently palpitations and chest pain.^{42,194,200–202}

Recommendation on use of the modified European Heart Rhythm Association symptom scale

AF = atrial fibrillation; EHRA = European Heart Rhythm Association.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

7. Detection and management of risk factors and concomitant cardiovascular diseases

Many cardiovascular diseases and concomitant conditions increase the risk of developing AF (Table 8), recurrent AF, and AF-associated complications. The identification of such conditions, their prevention and treatment is an important leverage to prevent AF and its disease burden. Knowledge of these factors and their management is hence important for optimal management of AF patients.^203,204

Table 8

Cardiovascular and other conditions independently associated with atrial fibrillation

AF = atrial fibrillation; BMI = body mass index; CI = confidence interval; FEV1 = forced expiratory volume in 1 second; HR = hazard ratio; OR = odds ratio; RR = risk ratio.

7.1 Heart failure

Heart failure and AF coincide in many patients.^215–217 They are linked by similar risk factors and share a common pathophysiology.²¹⁸ Heart failure and AF can cause and exacerbate each other through mechanisms such as structural cardiac remodelling, activation of neurohormonal mechanisms, and rate-related impairment of left ventricular (LV) function. Patients with AF and concomitant heart failure, both with preserved ejection fraction [LV ejection fraction (LVEF) ≥50%] and reduced ejection fraction (LVEF <40%),^219,220 suffer from a worse prognosis, including increased mortality.^16,221 The recent ESC Guidelines on heart failure²²² have also introduced a new category of heart failure with mid-range ejection fraction (HFmrEF; LVEF 40–49%), although data on AF patients in this group are limited. Prevention of adverse outcomes and maintenance of a good quality of life are the aims of management in all patients with AF and concomitant heart failure, regardless of LVEF.²²³ The general approach to AF management does not differ between heart failure patients and others, but a few considerations are worthwhile. Of note, the only therapy with proven prognostic value in these patients is anticoagulation, and appropriate OAC should be prescribed in all patients at risk of stroke (see Chapter 9).

7.1.1 Patients with atrial fibrillation and heart failure with reduced ejection fraction

In addition to OAC, standard heart failure therapy should be used in patients with heart failure with reduced ejection fraction (HFrEF), as detailed in the ESC Guidelines.²²² This includes angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), mineralocorticoid antagonists, defibrillators, cardiac resynchronization therapy,²¹⁸ and combined angiotensin receptor neprilysin inhibition (ARNI) in patients able to tolerate an ACE inhibitor or ARB with ongoing symptoms.²²⁴

Rate control of AF is discussed in detail in Chapter 10. In brief, only beta-blockers and digoxin are suitable in HFrEF because of the negative inotropic potential of verapamil and diltiazem. Beta-blockers are usually the first-line option in patients with clinically stable HFrEF, although a meta-analysis using individual patient data from randomized controlled trials (RCTs) found no reduction in mortality from beta-blockers vs. placebo in those with AF at baseline (HR 0.97, 95% CI 0.83–1.14).²³ Digoxin is commonly prescribed in clinical practice, but no head-to-head RCTs in AF patients have been performed. In a meta-analysis of observational studies, digoxin had a neutral effect on mortality in patients with AF and concomitant heart failure (adjusted observational studies HR 0.90, 95% CI 0.70–1.16; propensity-matched observational studies RR 1.08, 95% CI 0.93–1.26).²²⁵ Therefore, initial and combination rate control therapy for AF in HFrEF should take account of individual patient characteristics and symptoms; beta-blocker initiation should be delayed in patients with acute decompensated heart failure, and digoxin can accumulate and provoke adverse effects in patients with kidney dysfunction (see Chapter 10).

Patients with AF and HFrEF who present with severe symptoms may require rhythm control therapy in addition to rate control therapy. For patients who develop HFrEF as a result of rapid AF (tachycardiomyopathy), a rhythm control strategy is preferred, based on several relatively small patient cohorts and trials reporting improved LV function after restoration of sinus rhythm.^{185,226–228} The diagnosis of tachycardiomyopathy can be challenging, and at times requires the restoration of sinus rhythm.²²⁹ Catheter ablation may be a useful method to restore LV function and quality of life in AF patients with HFrEF,^{185,226–228} but further data are needed. Figure 4 summarizes the approach to patients with AF and heart failure.

Figure 4

Initial management of patients presenting acutely with atrial fibrillation and heart failure. (Adapted from Kotecha and Piccini.¹²⁸)

7.1.2 Atrial fibrillation patients with heart failure with preserved ejection fraction

The diagnosis of heart failure with preserved ejection fraction (HFpEF) in patients with AF is problematic because of the difficulty in separating symptoms that are due to HF from those due to AF. Although diagnostic differentiation can be achieved by cardioversion and clinical reassessment but should be reserved for symptomatic improvement as a specific therapy that improves prognosis in HFpEF is currently lacking. Echocardiography can support the detection of HFpEF in patients with symptomatic AF by providing evidence of relevant structural heart disease [e.g. LV hypertrophy (LVH)] and/or measurement of diastolic dysfunction. Reduced early diastolic myocardial velocity e' by tissue Doppler reflects impaired LV relaxation, while the ratio of E/e' is correlated with invasive measurement of LV filling pressures.^230–234 Natriuretic peptide levels are part of the diagnostic assessment of HFpEF,²²² although natriuretic peptide levels are elevated in AF patients and the optimum diagnostic cut-off is still unknown.²³⁵ The management of patients with AF and concomitant HFpEF should focus on the control of fluid balance and concomitant conditions such as hypertension and myocardial ischaemia.

7.1.3 Atrial fibrillation patients with heart failure with mid-range ejection fraction

HFmrEF is a recently defined entity, describing patients with symptoms and signs of heart failure, LVEF 40–49%, elevated levels of natriuretic peptides, and either LV hypertrophy, left atrial (LA) enlargement, or evidence of diastolic dysfunction.²²² However, diagnosis is more difficult in patients with AF, as natriuretic peptides are elevated in AF and LA dilatation is common, regardless of concomitant heart failure. LVEF is also variable and difficult to assess in AF patients because of AF-induced reduction in systolic LV function and variable cardiac cycle length. Further study of this group is required before particular treatment strategies in AF patients with HFmrEF can be recommended.

7.1.4 Prevention of atrial fibrillation in heart failure

Retrospective analyses from large randomized trials have reported a lower incidence of new-onset AF in patients treated with ACE inhibitors/ARBs compared with placebo.^236–238 The reduced incidence of AF with ACE inhibitors/ARBs is less evident in patients with HFpEF²³⁹ and is lost in patients without heart failure.^240–242 Neprilysin inhibition does not seem to add to this effect.²²⁴ Beta-blocker therapy was associated with a 33% reduction in the adjusted odds of incident AF in HFrEF patients pre-treated with ACE inhibitors/ARBs, reinforcing the importance of beta-blocker therapy in HFrEF patients in sinus rhythm.²³ Eplerenone, a mineralocorticoid receptor antagonist, also reduced the risk of new-onset AF in patients with LVEF ≤35%, New York Heart Association (NYHA) Class II, when added to ACE inhibitors/ARBs and beta-blockers.²⁴³

7.2 Hypertension

Hypertension is a stroke risk factor in AF; uncontrolled high blood pressure enhances the risk of stroke and bleeding events and may lead to recurrent AF. Therefore, good blood pressure control should form an integral part of the management of AF patients.²⁴⁷ Inhibition of the renin–angiotensin–aldosterone system can prevent structural remodelling and recurrent AF.^236,244 A recent analysis of the Danish healthcare database with long-term monitoring of the effect of different antihypertensive agents on the occurrence of overt AF suggests a beneficial effect of ACE inhibitors or ARBs.²⁴⁵ Secondary analyses of ACE inhibitors or ARBs in patients with heart failure or LVH show a lower incidence of new-onset AF.^238,246 In patients with established AF, but without LV dysfunction or heart failure, ARBs do not prevent recurrent AF better than placebo.^240,241 ACE inhibitors or ARBs may reduce recurrent AF after cardioversion when co-administered with antiarrhythmic drug therapy compared with an antiarrhythmic drug alone.^248,249 Meta-analyses driven by these studies suggested a lower risk of recurrent AF,^{236–238,250} but at least one controlled trial failed to demonstrate benefit.^240,251

7.3 Valvular heart disease

Valvular heart disease is independently associated with incident AF.²⁵² Approximately 30% of patients with AF have some form of valvular heart disease, often detected only by echocardiogram.^{201,253–255} AF worsens prognosis in patients with severe valvular heart disease,²⁵⁶ including those undergoing surgery or transcatheter interventions for aortic or mitral valve disease.^257–262 Valvular heart disease can be associated with an increased thrombo-embolic risk, which probably also adds to the stroke risk in AF patients.²⁶³ Similar to heart failure, valvular disease and AF interact with and sustain each other through volume and pressure overload, tachycardiomyopathy, and neurohumoral factors.^264–270 When valve dysfunction is severe, AF can be regarded as a marker for progressive disease, thus favouring valve repair or replacement.²⁷¹

Traditionally, patients with AF have been dichotomized into ‘valvular’ and ‘non-valvular’ AF.²⁷² Although slightly different definitions have been used, valvular AF mainly refers to AF patients that have either rheumatic valvular disease (predominantly mitral stenosis) or mechanical heart valves. In fact, while AF implies an incremental risk for thrombo-embolism in patients with mitral valve stenosis,^263,273,274 there is no clear evidence that other valvular diseases, including mitral regurgitation or aortic valve disease, need to be considered when choosing an anticoagulant or indeed to estimate stroke risk in AF.²⁷⁵ Therefore, we have decided to replace the historic term ‘non-valvular’ AF with reference to the specific underlying conditions.

Recommendations for patients with valvular heart disease and atrial fibrillation

AF = atrial fibrillation; LV = left ventricular.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

7.4 Diabetes mellitus

Diabetes and AF frequently coexist because of associations with other risk factors.^277–283 Diabetes is a risk factor for stroke and other complications in AF.²⁸⁴ In patients with AF, a longer duration of diabetes appears to confer a higher risk of thrombo-embolism, albeit without greater risk of OAC-related bleeding.²⁸⁵ Unfortunately, intensive glycaemic control does not affect the rate of new-onset AF,²⁸⁴ while treatment with metformin seems to be associated with a decreased long-term risk of AF in diabetic patients²⁸⁶ and may even be associated with a lower long-term stroke risk.¹³ Diabetic retinopathy, a measure of disease severity, does not increase the risk of ocular bleeding in anticoagulated patients.²⁸⁷

7.5 Obesity and weight loss

7.5.1 Obesity as a risk factor

Obesity increases the risk for AF (Table 8)^288–291 with a progressive increase according to body mass index (BMI).^{288,290–292} Obese patients may have more LV diastolic dysfunction, increased sympathetic activity and inflammation, and increased fatty infiltration of the atria.^293–295 Obesity may also be a risk factor for ischaemic stroke, thrombo-embolism, and death in AF patients.²⁹²

7.5.2 Weight reduction in obese patients with atrial fibrillation

Intensive weight reduction in addition to the management of other cardiovascular risk factors (in the range of 10–15 kg weight loss achieved), led to fewer AF recurrences and symptoms compared with an approach based on general advice in obese patients with AF.^203,204,296 Improved cardiorespiratory fitness can further decrease AF burden in obese patients with AF.²⁹⁷ Although the findings in these studies have to be confirmed, they underpin the positive effect of weight reduction in obese AF patients.

7.5.3 Catheter ablation in obese patients

Obesity may increase the rate of AF recurrence after catheter ablation,^298–301 with obstructive sleep apnoea as an important potential confounder. Obesity has also been linked to a higher radiation dose and complication rate during AF ablation.^302,303 Notably, the symptomatic improvement after catheter ablation of AF in obese patients seems comparable to the improvement in normal-weight patients.²⁹⁸ In view of the potential to reduce AF episodes by weight reduction (see chapter 7.5.2.), AF ablation should be offered to obese patients in conjunction with lifestyle modifications that lead to weight reduction.

Recommendation for obese patients with atrial fibrillation

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendation.

7.6 Chronic obstructive pulmonary disease, sleep apnoea, and other respiratory diseases

AF has been associated with obstructive sleep apnoea.^304,305 Multiple pathophysiological mechanisms can contribute to AF in obstructive sleep apnoea, including autonomic dysfunction, hypoxia, hypercapnia, and inflammation.^96,304–307 Obstructive sleep apnoea exaggerates intrathoracic pressure changes, which in itself and via vagal activation can provoke shortening of the atrial action potential and induce AF. Risk factor reduction and continuous positive airway pressure ventilation can reduce AF recurrence.^308–312 It seems reasonable to consider obstructive sleep apnoea screening in AF patients with risk factors. Obstructive sleep apnoea treatment should be optimized to improve AF treatment results in appropriate patients. Servo-controlled pressure support therapy should not be used in HFrEF patients with predominantly central sleep apnoea (of which 25% had concomitant AF).³¹³

Patients with chronic obstructive pulmonary disease often suffer from atrial tachycardias, which need to be differentiated from AF by ECG. Agents used to relieve bronchospasm, notably theophyllines and beta-adrenergic agonists, may precipitate AF and make control of the ventricular response rate difficult. Non-selective beta-blockers, sotalol, propafenone, and adenosine should be used with caution in patients with significant bronchospasm, while they can safely be used in patients with chronic obstructive pulmonary disease. Beta-1 selective blockers (e.g. bisoprolol, metoprolol, and nebivolol), diltiazem, and verapamil are often tolerated and effective (see Chapter 10).

Recommendations for patients with atrial fibrillation and respiratory diseases

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

7.7 Chronic kidney disease

AF is present in 15–20% of patients with CKD.³¹⁶ The definition of CKD in most AF trials is relatively strict. Although an estimated creatinine clearance (CrCl) rate of <60 mL/min is indicative of CKD, a number of trials in AF patients have used CrCl <50 mL/min to adapt NOAC dosage, usually estimated using the Cockroft–Gault formula. CrCl in AF patients can deteriorate over time.³¹⁷ The management of OAC in patients with CKD is discussed in chapter 9.2.4.

Recommendations for patients with kidney disease and atrial fibrillation

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

8. Integrated management of patients with atrial fibrillation

Most patients initially access the healthcare system through pharmacists, community health workers, or primary care physicians. As AF is often asymptomatic (“silent AF”), these healthcare professionals are important stakeholders to enable the adequate detection of AF and to ensure consistent management. The initial assessment should be performed at the point of first contact with the healthcare system, and is feasible in most healthcare settings (when an ECG is available). We propose to consider five domains in the initial assessment of patients presenting with newly diagnosed AF (Figure 5). These domains are:

1. Haemodynamic instability or limiting, severe symptoms;

2. Presence of precipitating factors (e.g. thyrotoxicosis, sepsis, or postoperative AF) and underlying cardiovascular conditions;

3. Stroke risk and need for anticoagulation;

4. Heart rate and need for rate control;

5. Symptom assessment and decision for rhythm control.

An integrated, structured approach to AF care, as applied successfully to other domains of medicine,^322–324 will facilitate consistent, guideline-adherent AF management for all patients³²⁵ (Figure 6), with the potential to improve outcomes.^42,326,327 Such approaches are consistent with the Innovative Care for Chronic Conditions Framework proposal put forward by the World Health Organization.³²⁸ Review by an AF service, or at least referral to a cardiologist, will usually be required after the initial assessment to fully evaluate the effect of AF on cardiovascular health.³²⁹ There may also be reasons for early or urgent referral (Table 9). Integrated care of all patients with newly diagnosed AF should help to overcome the current shortcomings of AF management, such as underuse of anticoagulation, access to rate and rhythm control therapy, and inconsistent approaches to cardiovascular risk reduction. Integrated AF care requires the cooperation of primary care physicians, cardiologists, cardiac surgeons, AF specialists, stroke specialists, allied health practitioners, and patients, encompassing lifestyle interventions, treatment of underlying cardiovascular diseases, and AF-specific therapy (Figure 7).

Table 9

Clinical signs calling for urgent involvement of a specialized atrial fibrillation service^a

^aAnticoagulation should be initiated early in all suitable patients and will not routinely require specialist input.

Figure 5

Acute and chronic management of atrial fibrillation patients, desired cardiovascular outcomes, and patient benefits. Adapted from the report on the 4th AFNET/EHRA consensus conference.⁷⁶

Figure 6

Achieving optimal management of atrial fibrillation patients.

Figure 7

Fundamentals of integrated care in atrial fibrillation patients.

8.1 Evidence supporting integrated atrial fibrillation care

Several structured approaches to AF care have been developed. Some evidence underpins their use, while more research is needed into the best way of delivering integrated AF care. Integrated AF management in an RCT increased the use of evidence-based care, and reduced by approximately one-third the composite outcome of cardiovascular hospitalization and cardiovascular death over a mean follow-up of 22 months (14.3% vs. 20.8%, HR 0.65; 95% CI 0.45–0.93; P = 0.017) compared with usual care in a large tertiary care centre.³³⁰ Integrated AF management appeared cost-effective in that study.³³¹ However, an Australian RCT showed only a marginal effect on unplanned admissions and death using integrated AF care limited to the initial care period, possibly emphasizing the need for sustained integration of AF care.³³² Two observational studies of integrated AF care found fewer hospitalizations,^333,334 one study showed fewer cases of stroke,³³³ and a further non-randomized study identified a trend for a lower rate of the composite outcome of death, cardiovascular hospitalization, and AF-related emergency visits.³³⁵ More research is needed, and integrated AF care is likely to require different designs in different healthcare settings.

8.2 Components of integrated atrial fibrillation care

8.2.1 Patient involvement

Patients should have a central role in the care process. As treatment of AF requires patients to change their lifestyles and adhere to chronic therapy, at times without an immediately tangible benefit, they need to understand their responsibilities in the care process. Physicians and healthcare professionals are responsible for providing access to evidence-based therapy, but adherence to therapy is ultimately the responsibility of informed and autonomous patients, best described as ‘shared accountability’.³³⁶ Hence, information and the education of patients, and often of their partners and relatives, is indispensable to encourage a self-management role and to empower patients to participate in shared decision-making,^326,328 and to support understanding of the disease and the suggested treatments.³³⁷

8.2.2 Multidisciplinary atrial fibrillation teams

Delegation of tasks from specialists to general physicians and from physicians to allied health professionals is a fundamental concept of integrated care models. A multidisciplinary AF team approach includes an efficient mix of interpersonal and communication skills, education, and expertise in AF management, as well as the use of dedicated technology. This approach underlines the importance of redesigning daily practice in a way that encourages non-specialists and allied professionals to have an important role in educating patients and co-ordinating care, while the specialist remains medically responsible. Cultural and regional differences will determine the composition of AF teams.

8.2.3 Role of non-specialists

Some non-specialist health care professionals, e.g. physicians in primary care have extensive expertise in stroke prevention and initial management of AF patients. Others may seek training to acquire such knowledge. Other components of AF management (e.g. assessment of concomitant cardiovascular conditions, antiarrhythmic drug therapy, or interventional treatment) often require specialist input. Integrated AF care structures should support treatment initiation by non-specialists where appropriate, and provide ready access to specialist knowledge to optimize AF care.

8.2.4 Technology use to support atrial fibrillation care

Technology, such as decision support software, has the potential to enhance the implementation of evidence-based care and improve outcomes, when used to enhance expert advice.³³⁸ Electronic tools can also ensure coherent communication within the AF team. With a view to support the wider use of such technology, this Task Force is providing digital decision tools, in the form of freely accessible smartphone apps, to AF healthcare professionals and to AF patients.

Recommendations for an integrated approach to care

AF = atrial fibrillation

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

8.3 Diagnostic workup of atrial fibrillation patients

AF is often found in patients with other, at times undiagnosed, cardiovascular conditions. Thus, all AF patients will benefit from a comprehensive cardiovascular assessment.³³⁹

8.3.1 Recommended evaluation in all atrial fibrillation patients

A complete medical history should be taken and all patients should undergo clinical evaluation that includes thorough assessment for concomitant conditions, establishing the AF pattern, estimation of stroke risk and AF-related symptoms, and assessment of arrhythmia-related complications such as thrombo-embolism or LV dysfunction. A 12-lead ECG is recommended to establish a suspected diagnosis of AF, to determine rate in AF, and to screen for conduction defects, ischaemia, and signs of structural heart disease. Initial blood tests should evaluate thyroid and kidney function, as well as serum electrolytes and full blood count. Transthoracic echocardiography is recommended in all AF patients to guide treatment decisions. Transthoracic echocardiography should be used to identify structural disease (e.g. valvular disease) and assess LV size and function (systolic and diastolic), atrial size, and right heart function.^339,340 Although biomarkers such as natriuretic peptides are elevated in AF patients, there is insufficient data to suggest that blood-based parameters are independent markers for AF.^341–343

8.3.2 Additional investigations in selected patients with atrial fibrillation

Ambulatory ECG monitoring in AF patients can assess the adequacy of rate control, relate symptoms with AF recurrences, and detect focal induction of bouts of paroxysmal AF. Transoesophageal echocardiography (TOE) is useful to further assess valvular heart disease and to exclude intracardiac thrombi, especially in the LAA, to facilitate early cardioversion or catheter ablation.³⁴⁴ Patients with symptoms or signs of myocardial ischaemia should undergo coronary angiography or stress testing as appropriate. In patients with AF and signs of cerebral ischaemia or stroke, computed tomography (CT) or magnetic resonance imaging (MRI) of the brain is recommended to detect stroke and support decisions regarding acute management and long-term anticoagulation. Delayed-enhancement MRI of the left atrium using gadolinium contrast,^345–347 T1 mapping using cardiac MRI,³⁴⁷ and intracardiac echo³⁴⁸ may help to guide treatment decisions in AF, but require external validation in multicentre studies.

8.4 Structured follow-up

Most AF patients need regular follow-up to ensure continued optimal management. Follow-up may be undertaken in primary care, by specially trained nurses, by cardiologists, or by AF specialists.^325,330 A specialist should co-ordinate care and follow-up. Follow-up should ensure implementation of the management plan, continued engagement of the patient, and therapy adaptation where needed.

Recommendations for diagnostic workup of atrial fibrillation patients

AF = atrial fibrillation; ECG = electrocardiogram.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

8.5 Defining goals of atrial fibrillation management

AF management comprises therapies with prognostic impact (anticoagulation and treatment of cardiovascular conditions) and therapies predominantly providing symptomatic benefit (rate control and rhythm control, Table 10). Therapies with prognostic benefit need careful explanation to patients when their benefits are not directly felt. Rhythm control therapy can be successful if symptoms are controlled, even when AF recurs. Explaining the expected benefits to each patient at the start of AF management will prevent unfounded expectations and has the potential to optimize quality of life.

Table 10

Goal-based follow-up

bpm = beats per minute; mEHRA symptoms scale = modified European Heart Rhythm Association symptoms scale; GP = general practitioner; INR = international normalized ratio; LV = left ventricular; NOAC = non-vitamin K antagonist oral anticoagulant; VKA = vitamin K antagonist.

9. Stroke prevention therapy in atrial fibrillation patients

OAC therapy can prevent the majority of ischaemic strokes in AF patients and can prolong life.^{38,39,42,194,201,329,350–352} It is superior to no treatment or aspirin in patients with different profiles for stroke risk.^353,354 The net clinical benefit is almost universal, with the exception of patients at very low stroke risk, and OAC should therefore be used in most patients with AF (Figure 8). Despite this evidence, underuse or premature termination of OAC therapy is still common. Bleeding events, both severe and nuisance bleeds, a perceived ‘high risk of bleeding’ on anticoagulation, and the efforts required to monitor and dose-adjust VKA therapy are among the most common reasons for withholding or ending OAC.^{352,355–359} However, the considerable stroke risk without OAC often exceeds the bleeding risk on OAC, even in the elderly, in patients with cognitive dysfunction, or in patients with frequent falls or frailty.^360,361 The bleeding risk on aspirin is not different to the bleeding risk on VKA³⁶² or NOAC therapy,^354,363 while VKA and NOACs, but not aspirin, effectively prevent strokes in AF patients.^{38,354,362,363}

Figure 8

Stroke prevention in atrial fibrillation.

9.1 Prediction of stroke and bleeding risk

9.1.1 Clinical risk scores for stroke and systemic embolism

Simple, clinically applicable stroke risk-stratification schemes in AF patients were developed in the late 1990s in small cohort studies, and have later been refined and validated in larger populations.^364–368 The introduction of the CHA₂DS₂-VASc score (Table 11) has simplified the initial decision for OAC in AF patients. Since its first incorporation in the ESC guidelines in 2010,³⁶⁹ it has been widely used.³⁷⁰ We recommend estimating stroke risk in AF patients based on the CHA₂DS₂-VASc score.³⁶⁸ In general, patients without clinical stroke risk factors do not need antithrombotic therapy, while patients with stroke risk factors (i.e. CHA₂DS₂-VASc score of 1 or more for men, and 2 or more for women) are likely to benefit from OAC.

Table 11

Clinical risk factors for stroke, transient ischaemic attack, and systemic embolism in the CHA₂DS₂-VASc score

CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female).

Other, less established risk factors for stroke include unstable international normalized ratio (INR) and low time in therapeutic range (TTR) in patients treated with VKAs; previous bleed or anaemia; alcohol excess and other markers for decreased therapy adherence; CKD; elevated high-sensitivity troponin; and elevated N-terminal pro-B-type natriuretic peptide.

9.1.2 Anticoagulation in patients with a CHA₂DS₂-VASc score of 1 in men and 2 in women

Controlled trials studying OAC in AF patients have been enriched for patients at high risk of stroke,^{38,39,42,194,201,329,351,352} and hence there is strong evidence that patients with a CHA₂DS₂-VASc risk score of 2 or more in men, and 3 or more in women, benefit from OAC. Fortunately, we now have a growing evidence base regarding stroke risk in patients with one clinical risk factor (i.e. a CHA₂DS₂-VASc score of 1 for men, and 2 for women), although this relies largely on observed stroke rates in patients not receiving OAC. In many of these patients, anticoagulation seems to provide a clinical benefit.^371–375 The rates of stroke and thrombo-embolism vary considerably in patients with CHA₂DS₂-VASc scores of 1 or 2 due to differences in outcomes, populations, and anticoagulation status (Web Table 1.)^{371,376,377,1041} We therefore commissioned an analysis of stroke risk in men and women with one additional stroke risk factor to inform these guidelines (Web Table 1, last line). OAC should be considered for men with a CHA₂DS₂-VASc score of 1 and women with a score of 2, balancing the expected stroke reduction, bleeding risk, and patient preference. Importantly, age (65 years and older) conveys a relatively high and continuously increasing stroke risk that also potentiates other risk factors (such as heart failure and sex). Hence, an individualized weighing of risk, as well as patient preferences, should inform the decision to anticoagulate patients with only one CHA₂DS₂-VASc risk factor, apart from female sex. Female sex does not appear to increase stroke risk in the absence of other stroke risk factors (Web Table 1).^378,379

Measurement of cardiac troponin (high-sensitivity troponin T or I) and N-terminal pro-B-type natriuretic peptide may provide additional prognostic information in selected AF patients.^380–382 Biomarker-based risk scores may, in the future, prove helpful to better stratify patients (e.g. those at a truly low risk of stroke).^75,382

9.1.3 Clinical risk scores for bleeding

Several bleeding risk scores have been developed, mainly in patients on VKAs. These include HAS-BLED [hypertension, abnormal renal/liver function (1 point each), stroke, bleeding history or predisposition, labile INR, elderly (>65 years), drugs/alcohol concomitantly (1 point each)], ORBIT (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation), and more recently, the ABC (age, biomarkers, clinical history) bleeding score, which also makes use of selected biomarkers.^383–385 Stroke and bleeding risk factors overlap (compare Tables 11 and 12). For example, older age is one of the most important predictors of both ischaemic stroke and bleeding in AF patients.^386,387 A high bleeding risk score should generally not result in withholding OAC. Rather, bleeding risk factors should be identified and treatable factors corrected (see chapter 8.5). Table 12 provides details of modifiable bleeding risk factors.

Table 12

Modifiable and non-modifiable risk factors for bleeding in anticoagulated patients based on bleeding risk scores

ABC = age, biomarkers, clinical history; ATRIA = AnTicoagulation and Risk factors In Atrial fibrillation; CKD = chronic kidney disease; CrCl = creatinine clearance; HAS-BLED = hypertension, abnormal renal/liver function (1 point each), stroke, bleeding history or predisposition, labile INR, elderly (>65 years), drugs/alcohol concomitantly (1 point each); HEMORR₂HAGES = hepatic or renal disease, ethanol abuse, malignancy, older (age >75), reduced platelet count or function, rebleeding risk (prior bleed; 2 points), hypertension (uncontrolled), anaemia, genetic factors (CYP 2C9 polymorphisms), excessive fall risk (including neuropsychiatric disease), and stroke; INR = international normalized ratio; ORBIT = Outcomes Registry for Better Informed Treatment of Atrial Fibrillation; TTR = time in therapeutic range; VKA = vitamin K antagonist.

^aDerived from the HAS-BLED score.³⁸⁴

^bDerived from the HEMORR₂HAGES score.³⁸³

^cDerived from the ATRIA score.³⁸⁵

^dDerived from the ORBIT score.³⁸⁸

^eDerived from the ABC bleeding score.³⁸⁷

Recommendations for prediction of stroke and bleeding risk

AF = atrial fibrillation; CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke or transient ischaemic attack or systemic embolism (doubled), Vascular disease, Age 65–74, and Sex (female).

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

9.2 Stroke prevention

9.2.1 Vitamin K antagonists

Warfarin and other VKAs were the first anticoagulants used in AF patients. VKA therapy reduces the risk of stroke by two-thirds and mortality by one-quarter compared with control (aspirin or no therapy).³⁸ VKAs have been used in many patients throughout the world with good outcomes,^394–396 and this is reflected in the warfarin arms of the NOAC trials (see chapter 9.2.2.). The use of VKAs is limited by the narrow therapeutic interval, necessitating frequent monitoring and dose adjustments, but VKAs, when delivered with adequate time in therapeutic range (TTR), are effective for stroke prevention in AF patients. Clinical parameters can help to identify patients who are likely to achieve a decent TTR on VKA therapy.³⁹⁷ These have been summarized in the SAMe-TT₂R₂ score. Patients who fare well on this score, when treated with a VKA, have on average a higher TTR than patients who do not fare well on the score.^398,399 VKAs are currently the only treatment with established safety in AF patients with rheumatic mitral valve disease and/or a mechanical heart valve prosthesis.⁴⁰⁰

9.2.2 Non-vitamin K antagonist oral anticoagulants

NOACs, including the direct thrombin inhibitor dabigatran and the factor Xa inhibitors apixaban, edoxaban, and rivaroxaban, are suitable alternatives to VKAs for stroke prevention in AF (Table 13). Their use in clinical practice is increasing rapidly.⁴⁰¹ All NOACs have a predictable effect (onset and offset) without need for regular anticoagulation monitoring. The phase III trials have been conducted with carefully selected doses of the NOACs, including clear rules for dose reduction that should be followed in clinical practice (Table 13).

Table 13

Characteristics of approved non-vitamin K antagonist oral anticoagulants compared

9.2.2.1 Apixaban

In the ARISTOTLE (Apixaban for Reduction in Stroke and Other Thrombo-embolic Events in Atrial Fibrillation) trial,³¹⁹ apixaban 5 mg twice daily reduced stroke or systemic embolism by 21% compared with warfarin, combined with a 31% reduction in major bleeding and an 11% reduction in all-cause mortality (all statistically significant). Rates of haemorrhagic stroke and intracranial haemorrhage, but not of ischaemic stroke, were lower on apixaban. Rates of gastrointestinal bleeding were similar between the two treatment arms.⁴⁰²

Apixaban is the only NOAC that has been compared with aspirin in AF patients; apixaban significantly reduced stroke or systemic embolism by 55% compared with aspirin, with no or only a small difference in rates of major bleeding or intracranial haemorrhage.^354,403

9.2.2.2 Dabigatran

In the RE-LY (Randomized Evaluation of Long-Term Anticoagulation Therapy) study,^318,404 dabigatran 150 mg twice daily reduced stroke and systemic embolism by 35% compared with warfarin without a significant difference in major bleeding events. Dabigatran 110 mg twice daily was non-inferior to warfarin for prevention of stroke and systemic embolism, with 20% fewer major bleeding events. Both dabigatran doses significantly reduced haemorrhagic stroke and intracranial haemorrhage. Dabigatran 150 mg twice daily significantly reduced ischaemic stroke by 24% and vascular mortality by 12%, while gastrointestinal bleeding was significantly increased by 50%. There was a non-significant numerical increase in the rate of myocardial infarction with both dabigatran doses,^318,404 which has not been replicated in large post-authorization analyses.³⁹⁶ These observational data have also replicated the benefit of dabigatran over VKA found in the RE-LY trial in patients who were mainly treated with the higher dabigatran dose (150 mg twice daily).³⁹⁶

9.2.2.3 Edoxaban

In the ENGAGE AF-TIMI 48 (Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48) trial,³²¹ edoxaban 60 mg once daily and edoxaban 30 mg once daily (with dose reductions in certain patients, Table 13), were compared with adjusted-dose warfarin.⁴⁰⁵ Edoxaban 60 mg once daily was non-inferior to warfarin (Table 13). In an on-treatment analysis, edoxaban 60 mg once daily significantly reduced stroke or systemic embolism by 21% and significantly reduced major bleeding events by 20% compared with warfarin, while edoxaban 30 mg once daily was non-inferior to warfarin for prevention of stroke and systemic embolism but significantly reduced major bleeding events by 53%. Cardiovascular death was reduced in patients randomized to edoxaban 60 mg once daily or edoxaban 30 mg once daily compared with warfarin. Only the higher dose regimen has been approved for stroke prevention in AF.

9.2.2.4 Rivaroxaban

In the ROCKET-AF (Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation) trial,³²⁰ patients were randomized to rivaroxaban 20 mg once daily or VKA, with a dose adjustment to 15 mg daily for those with estimated CrCl 30–49 mL/min by the Cockroft–Gault formula (Table 13). Rivaroxaban was non-inferior to warfarin for the prevention of stroke and systemic embolism in the intent-to-treat analysis, while the per-protocol on-treatment analysis achieved statistical superiority with a 21% reduction in stroke or systemic embolism compared with warfarin. Rivaroxaban did not reduce the rates of mortality, ischaemic stroke, or major bleeding events compared to VKA. There was an increase in gastrointestinal bleeding events, but a significant reduction in haemorrhagic stroke and intracranial haemorrhage with rivaroxaban compared with warfarin. Comparable event rates have been reported in post-authorization analyses, which are part of the post-approval risk management process.^406,407

9.2.3 Non-vitamin K antagonist oral anticoagulants or vitamin K antagonists

Both VKAs and NOACs are effective for the prevention of stroke in AF. A meta-analysis³⁹ based on the high-dose treatment groups of the pivotal studies of warfarin vs. NOACs included 42 411 patients receiving a NOAC and 29 272 receiving warfarin. NOACs in these dosages significantly reduced stroke or systemic embolic events by 19% compared with warfarin (RR 0.81; 95% CI 0.73–0.91; P < 0.0001), mainly driven by a reduction in haemorrhagic stroke (RR 0.49; 95% CI 0.38–0.64; P < 0.0001). Mortality was 10% lower in patients randomized to NOAC therapy (RR 0.90; 95% CI 0.85–0.95; P = 0.0003) and intracranial haemorrhage was halved (RR 0.48; 95% CI 0.39–0.59; P < 0.0001), while gastrointestinal bleeding events were more frequent (RR 1.25; 95% CI 1.01–1.55; P = 0.04).³⁹ The stroke reduction with NOACs was consistent in all evaluated subgroups, while there was a suggestion of greater relative reduction in bleeding with NOACs at centres with poor INR control (interaction P = 0.022). Notably, the substantial reduction in intracranial haemorrhage by NOACs compared with warfarin seems unrelated to the quality of INR control.^408,409

9.2.4 Oral anticoagulation in atrial fibrillation patients with chronic kidney disease

CKD is associated with stroke and bleeding in large data sets.^410,411 Anticoagulation can be safely used in AF patients with moderate or moderate-to-severe CKD [glomerular filtration rate (GFR) ≥15 mL/min]: the SPAF (Stroke Prevention in Atrial Fibrillation) III trial randomized 805/1936 participants with stage 3 CKD (estimated GFR <59 mL/min/1.73 m²), and reported good outcomes on warfarin (INR 2–3).⁴¹² This finding is supported by a large Swedish database, in which stroke risk was lower in CKD patients with AF treated with warfarin (adjusted HR 0.76; 95% CI 0.72–0.80),⁴¹³ while bleeding was also slightly increased, especially during therapy initiation.⁴¹⁴ In a meta-analysis of the major NOAC trials, patients with mild or moderate CKD suffered fewer strokes, systemic emboli, or major bleeding events on NOACs than on warfarin.⁴¹⁵ Kidney function should be regularly monitored in AF patients on OACs to allow dose adaptation for those on NOACs (Table 14) and to refine risk estimation.⁴¹⁶

Table 14

Dose adjustment for NOACs as evaluated in the PHASE III trials (adapted from Hart et al.³¹⁶)

CKD = chronic kidney disease; CrCl = creatinine clearance; GFR = glomerular filtration rate; NA = not available.

9.2.5 Oral anticoagulation in atrial fibrillation patients on dialysis

Approximately one in eight dialysis patients suffer from AF, with an incidence rate of 2.7/100 patient-years.⁴¹⁷ AF is associated with increased mortality in patients on dialysis.⁴¹⁷ There are no randomized trials assessing OAC in haemodialysis patients,⁴¹⁸ and no controlled trials of NOACs in patients with severe CKD (CrCl <25–30 mL/min).^318–321 Warfarin use was associated either with a neutral or increased risk of stroke in database analyses of patients on dialysis,^419–421 including a population-based analysis in Canada (adjusted HR for stroke 1.14; 95% CI 0.78–1.67, adjusted HR for bleeding 1.44; 95% CI 1.13–1.85).⁴²² In contrast, data from Denmark suggest a benefit of OAC in patients on renal replacement therapy.⁴²³ Hence, controlled studies of anticoagulants (both VKAs and NOACs) in AF patients on dialysis are needed.⁴²⁴

9.2.6 Patients with atrial fibrillation requiring kidney transplantation

There are no randomized trials assessing OAC in patients after kidney transplantation. The prescription of NOAC therapy should be guided by the estimated GFR of the transplanted kidney. Potential pharmacokinetic interactions of OAC with immunosuppressive agents should be considered.

9.2.7 Antiplatelet therapy as an alternative to oral anticoagulants

The evidence supporting antiplatelet monotherapy for stroke prevention in AF is very limited.^38,428–430 VKA therapy prevents stroke, systemic embolism, myocardial infarction, and vascular death better than single or dual antiplatelet therapy with aspirin and clopidogrel (annual risk of 5.6% for aspirin and clopidogrel vs. 3.9% with VKA therapy).⁴³¹ Even greater benefits were seen in VKA-treated patients with a high TTR.⁴³² Antiplatelet therapy increases bleeding risk, especially dual antiplatelet therapy (2.0% vs. 1.3% with antiplatelet monotherapy; P < 0.001),⁴³³ with bleeding rates that are similar to those on OAC.^{354,362,431,434} Thus, antiplatelet therapy cannot be recommended for stroke prevention in AF patients.

Recommendations for stroke prevention in patients with atrial fibrillation

AF = atrial fibrillation; CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female); INR = international normalized ratio; NOAC = non-vitamin K antagonist oral anticoagulant; OAC = oral anticoagulation; TTR = time in therapeutic range; VKA = vitamin K antagonist.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

9.3 Left atrial appendage occlusion and exclusion

9.3.1 Left atrial appendage occlusion devices

Interventional LAA occlusion,^446–449 and limited experience with percutaneous LAA ligation, has mainly been reported in observational studies and registries. Only one device (Watchman®) has been compared with VKA therapy in randomized trials [PROTECT AF (Watchman Left Atrial Appendage System for Embolic Protection in Patients With AF trial), see Web Table 2; and PREVAIL (Prospective Randomized Evaluation of the Watchman LAA Closure Device In Patients with AF Versus Long Term Warfarin Therapy trial)].^449–451 In these data sets, LAA occlusion was non-inferior to VKA treatment for the prevention of stroke in AF patients with moderate stroke risk, with a possibility of lower bleeding rates in the patients who continued follow-up.^452,453 These data were confirmed in a patient-level meta-analysis of the two trials and their associated registries.⁴⁵³ LAA occlusion may also reduce stroke risk in patients with contraindications to OAC.^454,455 The implantation procedure can cause serious complications,^{446,456–458} with high event rates reported in analyses from insurance databases and systematic reviews, possibly identifying a certain degree of reporting bias.^446,456 A large recent European registry reported a high rate of implantation success (98%), with an acceptable procedure-related complication rate of 4% at 30 days.⁴⁵⁹ Most patients who historically would be considered unsuitable for OAC therapy seem to do relatively well on contemporarily managed OAC.^396,407,460 Adequately powered controlled trials are urgently needed to inform the best use of these devices, including LAA occluders in patients who are truly unsuitable for OAC or in patients who suffer a stroke on OAC, randomized comparisons of LAA occluders with NOACs, and assessment of the minimal antiplatelet therapy acceptable after LAA occlusion.

9.3.2 Surgical left atrial appendage occlusion or exclusion

Surgical LAA occlusion or exclusion concomitant to cardiac surgery has been performed for many decades and with various techniques. Multiple observational studies indicate the feasibility and safety of surgical LAA occlusion/exclusion, but only limited controlled trial data are available.^461–464 Residual LAA flow or incomplete LAA exclusion can increase stroke risk.⁴⁶⁵ In most studies, LAA occlusion/exclusion was performed during other open heart surgery, and more recently in combination with surgical ablation of AF⁴⁶³ or as a stand-alone thoracoscopic procedure. One randomized trial evaluating the role of concomitant AF surgery and LAA occlusion reported in 2015, without a clear benefit of LAA exclusion for stroke prevention in the subgroup undergoing AF surgery.⁴⁶⁶ A large randomized trial is currently underway.⁴⁶⁷

Recommendations for occlusion or exclusion of the left atrial appendage

AF = atrial fibrillation; LAA = left atrial appendage.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

9.4 Secondary stroke prevention

The most important risk factors for stroke in patients with AF are advanced age and previous cardioembolic stroke or TIA,³⁸² emphasizing the need for OAC in these patients. The highest risk of recurrent stroke is in the early phase after a first stroke or TIA.^469,470

9.4.1 Treatment of acute ischaemic stroke

Systemic thrombolysis with recombinant tissue plasminogen activator (rtPA) is an effective and approved medical treatment for acute ischaemic stroke in patients presenting within 4.5 h of symptom onset.⁴⁷¹ Systemic thrombolysis is contraindicated in patients on therapeutic OAC.^472,473 Recombinant tissue plasminogen activator can be given in patients treated with a VKA if the INR is below 1.7,⁴⁷⁴ or in dabigatran-treated patients with a normal activated partial thromboplastin time and last intake of drug >48 h previously (based on expert consensus).⁴⁷² Whether specific NOAC antidotes⁴⁷⁵ could be used followed by systemic thrombolysis needs to be investigated. Thrombectomy can be performed in anticoagulated patients with distal occlusion of the internal carotid artery or middle cerebral artery in a 6 h window.⁴⁷⁶

9.4.2 Initiation of anticoagulation after transient ischaemic attack or ischaemic stroke

Data on the optimal use of anticoagulants (heparin, low-molecular-weight heparin, heparinoid, VKA, NOAC) in the first days after a stroke are scarce. Parenteral anticoagulants seem to be associated with a non-significant reduction in recurrent ischaemic stroke when administered 7–14 days after the acute stroke [odds ratio (OR) 0.68; 95% CI 0.44–1.06), with a significant increase in symptomatic intracranial bleeding (OR 2.89; 95% CI 1.19–7.01), and a similar rate of death or disability at final follow-up.⁴⁷⁷ It seems likely that the bleeding risk on parenteral anticoagulation exceeds the stroke prevention benefit in the first days after a large stroke, whereas patients with a TIA or a small stroke may benefit from early (immediate) initiation or continuation of anticoagulation. Therefore, we propose to initiate anticoagulation in AF patients between 1 and 12 days after an ischaemic stroke, depending on stroke severity (Figure 9).⁴⁷⁸ We suggest repeat brain imaging to determine the optimal initiation of anticoagulation in patients with a large stroke at risk for haemorrhagic transformation. Long-term OAC with a VKA^{363,479–481} or NOAC⁴⁸² conveys benefits in AF patients who survived a stroke. NOACs seem to convey slightly better outcomes, mainly driven by fewer intracranial haemorrhages and haemorrhagic strokes (OR 0.44; 95% CI 0.32–0.62).⁴⁸² Detailed data for edoxaban have not yet been published.³²¹ If a patient suffers a stroke or TIA whilst taking an anticoagulant, switching to another anticoagulant should be considered.

Figure 9

Initiation or continuation of anticoagulation in atrial fibrillation patients after a stroke or transient ischaemic attack. This approach is based on consensus rather than prospective data.

9.4.3 Initiation of anticoagulation after intracranial haemorrhage

No prospective studies have investigated the benefit or risk of the initiation of OAC after intracranial haemorrhage,⁴⁸³ and patients with a history of intracranial bleeding were excluded from the randomized trials comparing NOACs with VKAs. The available evidence indicates that anticoagulation in patients with AF can be reinitiated after 4–8 weeks, especially when the cause of bleeding or the relevant risk factor (e.g. uncontrolled hypertension, see Table 12) has been treated, and that such treatment leads to fewer recurrent (ischaemic) strokes and lower mortality.^460,484 If anticoagulation is resumed, it seems reasonable to consider anticoagulants with a low bleeding risk.³⁹Figure 10 depicts a consensus opinion on the initiation or resumption of OAC after an intracranial haemorrhage. We recommend a multidisciplinary decision with input from stroke physicians/neurologists, cardiologists, neuroradiologists, and neurosurgeons.

Recommendations for secondary stroke prevention

AF = atrial fibrillation; INR = international normalized ratio; LMWH = Low Molecular Weight Heparin; NOAC = non-vitamin K antagonist oral anticoagulant; OAC = oral anticoagulation; TIA = transient ischaemic attack; VKA = vitamin K antagonist.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

Figure 10

Initiation or resumption of anticoagulation in atrial fibrillation patients after an intracranial bleed. This approach is based on consensus opinion and retrospective data. In all patients, evaluation by a multidisciplinary panel is required before treatment (stroke physician/neurologist, cardiologist, neuroradiologist, and neurosurgeon).

9.5 Strategies to minimize bleeding on anticoagulant therapy

In a meta-analysis of 47 studies, the overall incidence of major bleeding with VKAs was 2.1 (range 0.9–3.4) per 100 patient-years in controlled trials and 2.0 (range 0.2–7.6) per 100 patient-years for observational data sets.⁴⁸⁸ Minimizing treatable bleeding risk factors (see Table 12) seems paramount to reduce the bleeding rate on anticoagulants.

9.5.1 Uncontrolled hypertension

Uncontrolled hypertension increases the risk of bleeding on OAC.⁵³ Hence, keeping systolic blood pressure well controlled is of particular relevance in anticoagulated patients with AF. Treatment according to current guidelines is recommended in patients with known hypertension.⁴⁸⁹

9.5.2 Previous bleeding event

History of bleeding events and the presence of anaemia are important parts of the assessment of all patients receiving OAC. The majority of bleeding events are gastrointestinal. Compared with warfarin, the risk of gastrointestinal bleeds was increased for dabigatran 150 mg twice daily,^396,490 rivaroxaban 20 mg once daily,⁴⁹¹ and edoxaban 60 mg once daily.³²¹ The risk of gastrointestinal bleeds was comparable to warfarin on dabigatran 110 mg twice daily⁴⁹⁰ and on apixaban 5 mg twice daily.³¹⁹ Recent observational analyses do not replicate these findings, suggesting a smaller effect.^396,492,493 In patients in whom the source of bleeding has been identified and corrected, OAC can be reinitiated. This also appears true for patients who have had an intracranial haemorrhage, once modifiable bleeding risk factors (e.g. uncontrolled hypertension) have been corrected.^460,484

9.5.3 Labile international normalized ratio and adequate non-vitamin K antagonist oral anticoagulant dosing

TTR on VKA therapy is an important predictor of major haemorrhage.^432,441,494 Therefore, we recommend targeting the INR between 2.0 and 3.0 in patients on VKAs, maintaining a high TTR (e.g. ≥70%⁴⁹⁴), and to consider switching to a NOAC when a high TTR cannot be sustained.⁴⁴⁴ NOAC dosing should follow the dose-reduction criteria evaluated in the clinical trials, considering renal function, age, and weight. Patient information and empowerment, best delivered through integrated AF management, seem paramount to achieve this goal.

9.5.4 Alcohol abuse

Alcohol excess is a risk factor for bleeding in anticoagulated patients,³⁸⁴ mediated by poor adherence, liver disease, variceal bleeding, and risk of major trauma. Severe alcohol abuse and binge drinking habits should be corrected in patients eligible for OAC.

9.5.5 Falls and dementia

Falls and dementia are associated with increased mortality in AF patients,⁴⁹⁵ without evidence that these conditions markedly increase the risk of intracranial haemorrhage.^495,496 Hence, anticoagulation should only be withheld from patients with severe uncontrolled falls (e.g. epilepsy or advanced multisystem atrophy with backwards falls), or in selected patients with dementia where compliance and adherence cannot be ensured by a caregiver.

9.5.6 Genetic testing

In addition to food and drug interactions, multiple genetic variations affect the metabolism of VKAs.⁴⁹⁷ The systematic use of genetic information for adjustment of VKA dosage has been evaluated in several controlled clinical studies.^498–500 Genetic testing has little effect on TTR or bleeding risk on warfarin, and is not recommended for clinical use at present.⁵⁰¹

9.5.7 Bridging periods off oral anticoagulation

Most cardiovascular interventions (e.g. percutaneous coronary intervention or pacemaker implantation) can be performed safely on continued OAC. When interruption of OAC is required, bridging does not seem to be beneficial, except in patients with mechanical heart valves: In a randomized trial of 1884 patients with AF, interruption of anticoagulation was non-inferior to heparin bridging for the outcome of arterial thrombo-embolism (incidence of 0.4% and 0.3%, respectively) and resulted in a lower risk of major bleeding (1.3% and 3.2%, respectively).⁵⁰² OAC interruptions should be minimized to prevent stroke.

9.6 Management of bleeding events in anticoagulated patients with atrial fibrillation

9.6.1 Management of minor, moderate, and severe bleeding

General assessment of an anticoagulated patient with AF experiencing a bleeding event should include the assessment of bleeding site, onset, and severity of the bleeding, the time-point of last intake of OAC and other antithrombotic drugs, and other factors influencing bleeding risk such as CKD, alcohol abuse, and concurrent medications. Laboratory tests should include haemoglobin, haematocrit, platelet count, renal function, and, for VKA patients, prothrombin time, activated partial thromboplastin time, and INR. Coagulation tests do not provide much information in patients on NOACs, except for activated partial thromboplastin time in the case of dabigatran. More specific coagulation tests do exist, including diluted thrombin time (HEMOCLOT) for dabigatran and calibrated quantitative anti-factor Xa assays for factor Xa inhibitors.⁵⁰³ However, these tests are not always readily available and are often unnecessary for bleeding management.⁵⁰⁴

We propose a simple scheme to manage bleeding events in patients on OAC (Figure 11). Minor bleeding events should be treated with supportive measures such as mechanical compression or minor surgery to achieve haemostasis. In patients receiving VKAs, the next dose of VKA can be postponed. NOACs have a short plasma half-life of approximately 12 h, and improved haemostasis is expected within 12–24 h after a delayed or omitted dose. Treatment of moderate bleeding events may require blood transfusions and fluid replacement. Specific diagnostic and treatment interventions directed against the cause of the bleeding (e.g. gastroscopy) should be performed promptly. If the intake of NOAC was recent (<2–4 h), charcoal administration and/or gastric lavage will reduce further exposure. Dialysis clears dabigatran but is less effective for the other NOACs.

Figure 11

Management of active bleeding in patients receiving anticoagulation. Institutions should have an agreed procedure in place.

Immediate reversal of the antithrombotic effect is indicated in severe or life-threatening bleeding events. An agreed institutional procedure for the management of life-threatening bleeds should be documented and accessible at all times to ensure adequate initial management. For VKAs, administration of fresh frozen plasma restores coagulation more rapidly than vitamin K, and prothrombin complex concentrates achieve even faster blood coagulation.⁵⁰⁵ Registry data suggest that the combination of plasma and prothrombin complex concentrates is associated with the lowest case fatality following intracranial haemorrhage on VKA treatment with an INR ≥1.3.⁵⁰⁶ In a multicentre randomized trial of 188 patients, four-factor prothrombin complex concentrates achieved more rapid INR reversal and effective haemostasis than plasma in patients undergoing urgent surgical or invasive procedures.⁵⁰⁷ Administration of prothrombin complex concentrates may also be considered for severe bleeding on NOAC treatment if specific antidotes are not available.

Several antidotes to NOACs are under development. Idarucizumab (approved in 2015 by the US Food and Drug Administration and the European Medicines Agency) is a clinically available humanized antibody fragment that binds dabigatran and rapidly and dose-dependently reverses its effects without over-correction or thrombin generation.⁴⁷⁵ Andexanet alpha, a modified recombinant human factor Xa that lacks enzymatic activity, reverses the anticoagulant activity of factor Xa antagonists in healthy subjects within minutes after administration and for the duration of infusion, with a transient increase in markers of coagulation activity of uncertain clinical relevance.⁵⁰⁸ Another agent under development is ciraparantag (PER977), an antidote designed to reverse both direct thrombin and factor Xa inhibitors as well as the indirect inhibitor enoxaparin.⁵⁰⁹ The clinical usefulness of these specific antidotes needs further evaluation.

9.6.2 Oral anticoagulation in atrial fibrillation patients at risk of or having a bleeding event

While anticoagulation therapy should be paused to control active bleeding, absolute contraindications to long-term OAC after a bleeding episode are rare. When nuisance bleeds are the reason to stop OAC, a change from one anticoagulant to another seems reasonable. Many causes or triggers of major bleeding events can be treated and/or eliminated, including uncontrolled hypertension, gastrointestinal ulcers, and intracranial aneurysms. Reinitiation of anticoagulation after a bleeding event is often clinically justified.^460,510 Difficult decisions, including the discontinuation and recommencement of OAC, should be taken by a multidisciplinary team, balancing the estimated risk of recurrent stroke and bleeding, and considering the bleeding risk of different stroke prevention therapies. LAA exclusion or occlusion might be an alternative in selected patients.

Recommendations for management of bleeding

AF = atrial fibrillation; NOAC = non-vitamin K antagonist oral anticoagulant; OAC = oral anticoagulation; VKA = vitamin K antagonist.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

9.7 Combination therapy with oral anticoagulants and antiplatelets

Approximately 15% of AF patients in contemporary trials⁵¹³ and registries^514–516 have a history of myocardial infarction. Between 5–15% of AF patients will require stenting at some point in their lives. This scenario requires careful consideration of antithrombotic therapy, balancing bleeding risk, stroke risk, and risk of acute coronary syndromes (ACS).⁵¹⁶ Co-prescription of OAC with antiplatelet therapy, in particular triple therapy, increases the absolute risk of major haemorrhage.^445,517,518 A recent meta-analysis involving 30 866 patients with a recent ACS evaluated the effects of adding NOAC therapy to single (4135 patients) or dual (26 731 patients) antiplatelet therapy.⁵¹⁹ The addition of a NOAC increased the bleeding risk by 79–134%, while reducing recurrent ischaemic events only marginally in patients without AF. OAC monotherapy, and not combination therapy with antiplatelets, is recommended in AF patients with stable CAD but without an ACS and/or coronary intervention in the previous 12 months. In patients treated for ACS, and in those receiving a coronary stent, short-term triple combination therapy of OAC, clopidogrel, and aspirin seems warranted (Figure 12).

Figure 12

Antithrombotic therapy after an acute coronary syndrome in atrial fibrillation patients requiring anticoagulation.

9.7.1 Antithrombotic therapy after acute coronary syndromes and percutaneous coronary intervention in patients requiring oral anticoagulation

The optimal combination antithrombotic therapy or duration of combination therapy for AF patients undergoing percutaneous coronary intervention is not known, but the continued bleeding risk suggests a short duration. Expert consensus,⁵²⁰ reviewed and reconsidered by this Task Force, suggests the following principles: AF patients at risk for stroke, patients with mechanical valves, and patients with recent or recurrent deep vein thrombosis or pulmonary embolism should continue OAC during and after stenting. In general, a short period of triple therapy (OAC, aspirin, clopidogrel) is recommended, followed by a period of dual therapy (OAC plus a single antiplatelet) (Figure 13). When a NOAC is used, the consensus recommendation is that the lowest dose effective for stroke prevention in AF should be considered. Dose reduction beyond the approved dosing tested in phase III trials (see Table 13) is not currently recommended, and awaits assessment in ongoing controlled trials. The combination of aspirin, clopidogrel, and low-dose rivaroxaban (2.5 mg twice daily) is not recommended for stroke prevention in AF.⁵²¹

Figure 13

Antithrombotic therapy after elective percutaneous intervention in atrial fibrillation patients requiring anticoagulation.

The use of prasugrel or ticagrelor as part of triple therapy should be avoided unless there is a clear need for these agents (e.g. stent thrombosis on aspirin plus clopidogrel), given the lack of evidence and the greater risk of major bleeding compared with clopidogrel.^522,523 Ongoing trials will inform about such combination therapies in the future.

The omission of aspirin while maintaining clopidogrel and OAC has been evaluated in the WOEST (What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing) trial, in which 573 anticoagulated patients undergoing percutaneous coronary intervention (70% with AF) were randomized to either dual therapy with OAC and clopidogrel (75 mg once daily) or to triple therapy with OAC, clopidogrel, and aspirin.⁵²⁴ Bleeding was lower in the dual vs. triple therapy arm, driven by fewer minor bleeding events. The rates of myocardial infarction, stroke, target vessel revascularization, and stent thrombosis did not differ (albeit with low event numbers), but all-cause mortality was lower in the dual therapy group at 1 year (2.5% vs. triple therapy 6.4%). Although the trial was too small to assess ischaemic outcomes, dual therapy with OAC and clopidogrel may emerge in the future as an alternative to triple therapy in patients with AF and ACS and/or coronary intervention.⁵²⁵

Recommendations for combination therapy with oral anticoagulants and antiplatelets

ACS = acute coronary syndromes; AF = atrial fibrillation; OAC = oral anticoagulant.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

10. Rate control therapy in atrial fibrillation

Rate control is an integral part of the management of AF patients, and is often sufficient to improve AF-related symptoms. Compared with stroke prevention and rhythm control, very little robust evidence exists to inform the best type and intensity of rate control treatment, with the majority of data derived from short-term crossover trials and observational studies.^41,526–528 Pharmacological rate control can be achieved for acute or long-term rate control with beta-blockers, digoxin, the calcium channel blockers diltiazem and verapamil, or combination therapy (Table 15). A number of antiarrhythmic drugs also have rate-limiting properties (amiodarone, dronedarone, sotalol, and to some extent propafenone), but they should only be used in patients needing rhythm control therapy (see Chapter 11).

Table 15

Rate control therapy in atrial fibrillation

AF = atrial fibrillation; CKD = chronic kidney disease; i.v. = intravenous; LV = left ventricular; LVEF = left ventricular ejection fraction.

^aA number of other beta-blockers are also available, but are not recommended as specific rate control therapy in AF. These include atenolol (25–100 mg once daily with a short biological half-life), propranolol [non-selective, 1 mg over 1 min and repeat up to 3 mg at 2-min intervals (acute) or 10–40 mg three times daily (long-term)], or labetalol [non-selective, 1–2 mg/min (acute)].

^bIf ongoing requirement for amiodarone, follow with 900 mg i.v. over 24 h diluted in 500–1000 mL via a central venous cannula.

10.1 Acute rate control

In the setting of acute new-onset AF, patients are often in need of heart rate control. Physicians should evaluate underlying causes of elevated heart rate, such as infection, endocrine imbalance, anaemia, and pulmonary embolism. For acute rate control, beta-blockers and diltiazem/verapamil are preferred over digoxin because of their rapid onset of action and effectiveness at high sympathetic tone.^528–532 The choice of drug (Table 15) and target heart rate will depend on patient characteristics, symptoms, LVEF and haemodynamics, but a lenient initial approach to heart rate seems acceptable. Combination therapy may be required (Figure 14). In patients with HFrEF, beta-blockers, digitalis (digoxin or digitoxin), or their combination should be used,^218,533 as diltiazem and verapamil can have negative inotropic effects in patients with LVEF <40%.^222,534,535 In critically ill patients and those with severely impaired LV systolic function, intravenous amiodarone can be used where excess heart rate is leading to haemodynamic instability.^536–538 Urgent cardioversion should be considered in unstable patients (see chapter 11.1).

Figure 14

Acute heart rate control in patients with atrial fibrillation.

10.2 Long-term pharmacological rate control

10.2.1 Beta-blockers

Beta-adrenoreceptor blocker monotherapy is often the first-line rate-controlling agent,⁵³⁹ largely based on observations of better acute heart rate control than digoxin. Interestingly, the prognostic benefit of beta-blockers seen in HFrEF patients with sinus rhythm is lost in those with AF. In an individual patient-level meta-analysis of RCTs, beta-blockers did not reduce all-cause mortality compared to placebo in those with AF at baseline (HR 0.97; 95% CI 0.83–1.14; P = 0.73), whereas there was a clear benefit in patients with sinus rhythm (HR 0.73; 95% CI 0.67–0.80; P < 0.001).²³ The analysis, which included 3066 participants with HFrEF and AF, showed consistency across all subgroups and outcomes, with no heterogeneity between the 10 RCTs included (I² = 0%). Despite this lack of prognostic benefit in HFrEF, this Task Force still considers beta-blockers as a useful first-line rate control agent across all AF patients, based on the potential for symptomatic and functional improvement as a result of rate control, the lack of harm from published studies, and the good tolerability profile across all ages in sinus rhythm and in AF.^23,540

10.2.2 Non-dihydropyridine calcium channel blockers

Verapamil or diltiazem provide reasonable rate control in AF patients.⁵⁴¹ They should be avoided in patients with HFrEF because of their negative inotropic effects.^222,534,535 Verapamil or diltiazem can improve arrhythmia-related symptoms,⁵²⁶ in comparison with beta-blockers, which reduced exercise capacity and increased B-type natriuretic peptide in one small trial of low-risk patients with preserved LVEF.⁵⁴²

10.2.3 Digitalis

Cardiac glycosides such as digoxin and digitoxin have been in use for over two centuries, although prescriptions have been declining steadily over the past 15 years.⁵⁴³ In the randomized Digitalis Investigation Group (DIG) trial, digoxin had no effect on mortality compared to placebo in HFrEF patients in sinus rhythm (RR 0.99; 95% CI 0.91–1.07), but reduced hospital admissions (RR 0.72; 95% CI 0.66–0.79).^544,545 There have been no head-to-head RCTs of digoxin in AF patients.⁵⁴⁶ Observational studies have associated digoxin use with excess mortality in AF patients,^547–549 but this association is likely due to selection and prescription biases rather than harm caused by digoxin,^550–553 particularly as digoxin is commonly prescribed to sicker patients.²²⁵ In a crossover mechanistic trial of 47 patients with HFrEF and AF, there were no differences in heart rate, blood pressure, walking distance, or LVEF between carvedilol and digoxin, although beta-blockers did result in higher B-type natriuretic peptide levels, combination carvedilol/digoxin improved LVEF, and digoxin withdrawal reduced LVEF.⁵⁵⁴ Comparisons with other rate control therapies are based on small, short-duration studies that identify no or marginal differences in exercise capacity, quality of life, or LVEF compared to digoxin.^{526,554–558} Lower doses of digoxin (≤250 μg once daily), corresponding to serum digoxin levels of 0.5–0.9 ng/mL, may be associated with better prognosis.²²⁵

10.2.4 Amiodarone

Amiodarone can be useful for rate control as a last resort. The wide array of extracardiac adverse effects associated with amiodarone renders it a reserve agent in patients whose heart rate cannot be controlled with combination therapy (e.g. beta-blocker or verapamil/diltiazem combined with digoxin).

In summary, there is equipoise for the use of different rate control agents in AF. The choice of beta-blocker, diltiazem/verapamil, digoxin, or combination therapy should be made on an individual basis, after consideration of patient characteristics and patient preference. All available therapies have the potential for adverse effects and patients should initially be treated with a low dose and uptitrated to achieve symptom improvement. In practice, achieving a heart rate <110 b.p.m. will often require combination therapy (Figure 15). The benefit of different rate control strategies on symptoms, quality of life, and other intermediate outcomes is under investigation.⁵⁵⁹

Figure 15

Long-term heart rate control in patients with atrial fibrillation.

10.3 Heart rate targets in atrial fibrillation

The optimal heart rate target in AF patients is unclear. The RACE (Rate Control Efficacy in Permanent Atrial Fibrillation) II study randomized 614 patients with permanent AF to either a target heart rate <80 b.p.m. at rest and <110 b.p.m. during moderate exercise, or to a lenient heart rate target of <110 b.p.m. There was no difference in a composite of clinical events (14.9% in the strict rate control group, 12.9% in the lenient group),⁵⁶⁰ NYHA class, or hospitalizations.^560,561 Similar results were found in a pooled analysis of the AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) and RACE trials (1091 participants), albeit with smaller heart rate differences and without randomization.⁵⁶² It is worthwhile to note that many ‘adequately rate-controlled’ patients (resting heart rate 60–100 b.p.m.) are severely symptomatic, calling for additional management.¹⁹⁴ Nonetheless, lenient rate control is an acceptable initial approach, regardless of heart failure status, unless symptoms call for stricter rate control.

10.4 Atrioventricular node ablation and pacing

Ablation of the atrioventricular node/His bundle and implantation of a VVI pacemaker can control ventricular rate when medications fail to control rate and symptoms. It is a relatively simple procedure with a low complication rate and low long-term mortality risk,^563,564 especially when the pacemaker is implanted a few weeks before the AV nodal ablation and the initial pacing rate after ablation is set at 70–90 b.p.m.^565,566 The procedure does not worsen LV function⁵⁶⁷ and may even improve LVEF in selected patients.^568–570 In selected HFrEF patients treated with biventricular pacing (cardiac resynchronization therapy), AF can terminate,⁵⁷¹ although such a ‘rhythm control’ effect of cardiac resynchronization therapy is likely to be small and clearly needs confirmation.⁵⁷² AV nodal ablation renders patients pacemaker-dependent for the rest of their lives, limiting AV nodal ablation and pacing to patients whose symptoms cannot be managed by rate-controlling medication or by reasonable rhythm control interventions (see AF Heart Team, chapter 11.6). The choice of pacing therapy (right ventricular or biventricular pacing with or without an implantable defibrillator) will depend on individual patient characteristics, including LVEF.^573,574

Recommendations for rate control

AF = atrial fibrillation; bpm = beats per minute; LVEF = left ventricular ejection fraction.

Digitoxin is a suitable alternative to digoxin, where available. In patients with heart failure with reduced ejection fraction (LVEF <40%), recommended beta-blockers are bisoprolol, carvedilol, long-acting metoprolol, and nebivolol.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

11. Rhythm control therapy in atrial fibrillation

Restoring and maintaining sinus rhythm is an integral part of AF management. Antiarrhythmic drugs approximately double the rate of sinus rhythm compared with placebo.^580–584 Catheter ablation or combination therapy is often effective when antiarrhythmic drugs fail.^{226,585–587} Although many clinicians believe that maintaining sinus rhythm can improve outcomes in AF patients,⁵⁸⁸ all trials that have compared rhythm control and rate control to rate control alone (with appropriate anticoagulation) have resulted in neutral outcomes.^{41,578,579,582,589–593} Whether modern rhythm control management involving catheter ablation, combination therapy, and early therapy leads to a reduction in major cardiovascular events is currently under investigation, e.g. in the EAST – AFNET 4 (Early treatment of Atrial fibrillation for Stroke prevention Trial)⁴⁰ and CABANA (Catheter Ablation vs. Anti-arrhythmic Drug Therapy for Atrial Fibrillation Trial)⁵⁹⁴ trials. For now, rhythm control therapy is indicated to improve symptoms in AF patients who remain symptomatic on adequate rate control therapy.

11.1 Acute restoration of sinus rhythm

11.1.1 Antiarrhythmic drugs for acute restoration of sinus rhythm (‘pharmacological cardioversion’)

Antiarrhythmic drugs can restore sinus rhythm in patients with AF (pharmacological cardioversion) as shown in small controlled trials, meta-analyses,^{41,584,595,596} and in a few larger controlled trials.^597–605 Outside of Europe, dofetilide is available and can convert recent-onset AF.⁶⁰⁶ Pharmacological cardioversion restores sinus rhythm in approximately 50% of patients with recent-onset AF (Table 16).^607–609 In the short-term, electrical cardioversion restores sinus rhythm quicker and more effectively than pharmacological cardioversion and is associated with shorter hospitalization.^609–613 Pharmacological cardioversion, conversely, does not require sedation or fasting (Figure 16).

Table 16

Antiarrhythmic drugs for pharmacological cardioversion

ACS = acute coronary syndromes; AV = atrio-ventricular; IHD = ischaemic heart disease; i.v. = intravenous; LVH = left ventricular hypertrophy; NYHA = New York Heart Association; SBP = systolic blood pressure.

^aUse a large peripheral vessel and change to oral amiodarone within 24 h of i.v. (central line) administration.

^bIbutilide is only available in selected European countries.

Figure 16

Rhythm control management of recent onset atrial fibrillation.

Flecainide and propafenone are effective for pharmacological cardioversion,^{595,602–605,614,615} but their use is restricted to patients without structural  heart disease. Ibutilide is an alternative where available, but carries a risk of torsades de pointes.⁶¹⁵ Vernakalant^602–605 can be given to patients with mild heart failure (NYHA Class I or II), including those with ischaemic heart disease, provided they do not present with hypotension or severe aortic stenosis.^616–618 Amiodarone can be used in patients with heart failure and in patients with ischaemic heart disease (although patients with severe heart failure were excluded from most of the AF cardioversion trials).⁵⁹⁶ Amiodarone also slows heart rate by 10–12 b.p.m. after 8–12 h when given intravenously.⁵⁹⁶ Both amiodarone and flecainide appear more effective than sotalol in restoring sinus rhythm.^600,601,619

11.1.2 ‘Pill in the pocket’ cardioversion performed by patients

In selected patients with infrequent symptomatic episodes of paroxysmal AF, a single bolus of oral flecainide (200–300 mg) or propafenone (450–600 mg) can be self-administered by the patient at home (‘pill in the pocket’ therapy) to restore sinus rhythm, after safety has been established in the hospital setting.⁶²⁰ This approach seems marginally less effective than hospital-based cardioversion,⁶²¹ but is practical and provides control and reassurance to selected patients.

11.1.3 Electrical cardioversion

Synchronized direct current electrical cardioversion quickly and effectively converts AF to sinus rhythm, and is the method of choice in severely haemodynamically compromised patients with new-onset AF (Figure 16).^626–628 Electrical cardioversion can be performed safely in sedated patients treated with intravenous midazolam and/or propofol. Continuous monitoring of blood pressure and oximetry during the procedure is important.⁶²⁹ Skin burns may occasionally be observed. Intravenous atropine or isoproterenol, or temporary transcutaneous pacing, should be available to mitigate post-cardioversion bradycardia. Biphasic defibrillators are more effective than monophasic waveforms, and have become the industry standard.^626,628 Anterior–posterior electrode positions generate a stronger shock field in the left atrium than anterolaterally positioned electrodes, and restore sinus rhythm more effectively.^626,627,630

Pre-treatment with amiodarone (requiring a few weeks of therapy),^631,632 sotalol,⁶³¹ ibutilide,⁶³³ or vernakalant⁶³⁴ can improve the efficacy of electrical cardioversion, and similar effects are likely for flecainide⁵⁸⁴ and propafenone.⁶³⁵ Beta-blockers,⁶³⁶ verapamil, diltiazem,^637–639 and digoxin^640,641 do not reliably terminate AF or facilitate electrical cardioversion. When antiarrhythmic drug therapy is planned to maintain sinus rhythm after cardioversion, it seems prudent to start therapy 1–3 days before cardioversion (amiodarone: a few weeks) to promote pharmacological conversion and to achieve effective drug levels.^584,601

11.1.4 Anticoagulation in patients undergoing cardioversion

Cardioversion carries an inherent risk of stroke in non-anticoagulated patients,⁶⁴² which is reduced substantially by the administration of anticoagulation.⁶⁴³ Immediate initiation of anticoagulation is important in all patients scheduled for cardioversion.^644–646 Patients who have been in AF for longer than 48 h should start OAC at least 3 weeks before cardioversion and continue it for 4 weeks afterwards (in patients without a need for long-term anticoagulation). OAC should be continued indefinitely in patients at risk of stroke. This practice has never been evaluated in controlled trials, but seemed safe in a large observational data set from Finland.⁶⁴⁷ When early cardioversion is desired, TOE can exclude the majority of left atrial thrombi, allowing immediate cardioversion.^648,649 Ongoing studies will inform about the safety and efficacy of newly initiated anticoagulation using NOACs in patients scheduled for cardioversion.

11.2 Long-term antiarrhythmic drug therapy

The aim of antiarrhythmic drug therapy is improvement in AF-related symptoms.^41,580 Hence, the decision to initiate long-term antiarrhythmic drug therapy needs to balance symptom burden, possible adverse drug reactions, and patient preferences. The principles of antiarrhythmic drug therapy outlined in the 2010 ESC AF guidelines³⁶⁹ are still relevant and should be observed:

1. Treatment is aimed at reducing AF-related symptoms;

2. Efficacy of antiarrhythmic drugs to maintain sinus rhythm is modest;

3. Clinically successful antiarrhythmic drug therapy may reduce rather than eliminate the recurrence of AF;

4. If one antiarrhythmic drug ‘fails’, a clinically acceptable response may be achieved with another agent;

5. Drug-induced pro-arrhythmia or extracardiac side-effects are frequent;

6. Safety rather than efficacy considerations should primarily guide the choice of antiarrhythmic drug.

Antiarrhythmic drug therapy approximately doubles sinus rhythm maintenance compared with no therapy.⁵⁸⁰ There is no appreciable effect on mortality or cardiovascular complications, but rhythm control therapy can slightly increase the risk of hospitalizations (often for AF).^{41,578,579,582,589–593} To reduce the risk of side-effects,^201,580 a shorter duration of antiarrhythmic drug therapy seems desirable. As an example, short-term treatment (4 weeks) with flecainide for 4 weeks after cardioversion of AF was well-tolerated and prevented most (80%) AF recurrences when compared with long-term treatment.⁵⁸⁴ Short-term antiarrhythmic drug therapy is also used to avoid early AF recurrences after catheter ablation,⁶⁵⁰ and may be reasonable in patients deemed at increased risk of antiarrhythmic drug side-effects or in those with a low perceived risk of recurrent AF.

In addition to antiarrhythmic drug therapy and catheter ablation (see Chapter 11.3), management of concomitant cardiovascular conditions can reduce symptom burden in AF and facilitate the maintenance of sinus rhythm.^{203,204,296,312} This includes weight reduction, blood pressure control, heart failure treatment, increasing cardiorespiratory fitness, and other measures (see Chapter 7).

11.2.1 Selection of antiarrhythmic drugs for long-term therapy: safety first!

Usually, the safety of antiarrhythmic drug therapy determines the initial choice of antiarrhythmic drugs (Figure 17). The following major antiarrhythmic drugs are available to prevent AF:

Figure 17

Initiation of long-term rhythm control therapy in symptomatic patients with atrial fibrillation.

11.2.1.1 Amiodarone

Amiodarone is an effective multichannel blocker, reduces ventricular rate, and is safe in patients with heart failure.^582,651 Torsades de pointes pro-arrhythmia can occur, and QT interval and TU waves should be monitored on therapy (see Table 17).⁶⁵² Amiodarone often causes extracardiac side-effects, especially on long-term therapy,^653,654 rendering it a second-line treatment in patients who are suitable for other antiarrhythmic drugs. Amiodarone appears less suitable to episodic short-term therapy (unless after catheter ablation),⁶⁵⁵ probably because of its long biological half-life.

Table 17

Oral antiarrhythmic drugs used for maintaining sinus rhythm after cardioversion

AF = atrial fibrillation; AV = atrio-ventricular; bpm = beats per minute; CrCl = creatinine clearance; CYP2D6 = cytochrome P450 2D6; CYP3A4 = cytochrome P450 3A4; ECG = electrocardiogram; IHD = ischaemic heart disease; LV = left ventricular; LVEF = left ventricular ejection fraction; NYHA = New York Heart Association; SAN = sino-atrial node; VKA = vitamin K antagonist.

11.2.1.2 Dronedarone

Dronedarone maintains sinus rhythm, reduces ventricular rate, and prevents cardiovascular hospitalizations (mostly due to AF) and cardiovascular death in patients with paroxysmal or persistent AF or flutter who had at least one relevant cardiovascular comorbidity.^583,588,656 Dronedarone increases mortality in patients with recently decompensated heart failure (with or without AF),⁶⁵⁷ and in patients with permanent AF in whom sinus rhythm is not restored.⁶⁵⁸ Dronedarone moderately increases serum creatinine, reflecting a reduction in creatinine excretion rather than a decline in kidney function.⁶⁵⁹

11.2.1.3 Flecainide and propafenone

Flecainide and propafenone are effective in preventing recurrent AF.^581,584,620 They should only be used in patients without significant ischaemic heart disease or heart failure to avoid the risk of life-threatening ventricular arrhythmias.⁶⁶⁰ High ventricular rates resulting from the conversion of AF into atrial flutter with 1:1 conduction by flecainide or propafenone can be prevented by pre-administering a beta-blocker, verapamil, or diltiazem.

11.2.1.4 Quinidine and disopyramide

Quinidine and disopyramide have been associated with an increase in all-cause mortality (OR 2.39; 95% CI 1.03–5.59; number needed to harm 109; 95% CI 34–4985) at 1-year follow-up,^580,661 likely due to ventricular arrhythmias (torsades de pointes).^580,661 Although this pro-arrhythmic effect is more common at higher doses, they are less commonly used for rhythm control in AF. Disopyramide may be useful in ‘vagally mediated’ AF (e.g. AF occurring in athletes and/or during sleep⁷⁶), and has been shown to reduce LV outflow gradient and improve symptoms in patients with hypertrophic cardiomyopathy.^662–664

11.2.1.5 Sotalol

Sotalol has a relevant risk of torsades de pointes [1% in the Prevention of Atrial Fibrillation After Cardioversion (PAFAC) trial¹¹⁸]. Its d-enantiomer is associated with increased mortality compared to placebo in patients with LV dysfunction after a myocardial infarction,⁶⁶⁵ probably due to ventricular arrhythmias (OR 2.47; 95% CI 1.2–5.05; number needed to harm 166; 95% CI 61–1159).^580,665 On the other hand, d,l-sotalol has been used in AF patients without safety signals in two controlled trials.^581,601

11.2.1.6 Dofetilide

Dofetilide is another potassium channel blocker that is mainly available outside of Europe. Dofetilide restores and maintains sinus rhythm in heart failure patients,⁶⁶⁶ and occasionally in patients refractory to other antiarrhythmic drugs.⁶⁶⁷

Overall, it seems prudent to limit the use of quinidine, disopyramide, dofetilide, and sotalol to specific situations. Furthermore, combinations of QT-prolonging antiarrhythmic drugs should generally be avoided for rhythm control in AF (Table 17).

11.2.2 The twelve-lead electrocardiogram as a tool to identify patients at risk of pro-arrhythmia

Identifying patients at risk of pro-arrhythmia can help to mitigate the pro-arrhythmic risk of antiarrhythmic drugs.⁶⁶⁸ In addition to the clinical characteristics mentioned above, monitoring PR, QT, and QRS durations during initiation of antiarrhythmic drug therapy can identify patients at higher risk of drug-induced pro-arrhythmia on longer-term treatment.^669–671 In addition, the presence of ‘abnormal TU waves’ is a sign of imminent torsades de pointes.⁶⁵² Periodic ECG analysis for pro-arrhythmia signs has been used successfully in recent antiarrhythmic drug trials.^118,584,672 Specifically, ECG monitoring was used systematically on days 1–3 in patients receiving flecainide, propafenone, or sotalol to identify those at risk of pro-arrhythmia.^118,584,601 Based on this evaluated practice, we suggest to record an ECG in all patients before initiation of antiarrhythmic drugs. Scheduled ECGs during the initiation period seem reasonable (Table 17).

11.2.3 New antiarrhythmic drugs

Several compounds that inhibit the ultrarapid potassium current (I_Kur) and other inhibitors of atypical ion channels are in clinical development.^673–675 They are not available for clinical use at present. The antianginal compound ranolazine inhibits potassium and sodium currents and increases glucose metabolism at the expense of free fatty acid metabolism, thereby enhancing the efficient use of oxygen.^676,677 Ranolazine was safe in patients with non–ST-segment elevation myocardial infarction and unstable angina evaluated in the MERLIN (Metabolic Efficiency With Ranolazine for Less Ischemia in Non ST-Elevation Acute Coronary Syndrome) trial.⁶⁷⁸ In a post hoc analysis of continuous ECG recordings obtained during the first 7 days after randomization, patients assigned to ranolazine had a trend towards fewer episodes of AF than those on placebo [75 (2.4%) vs. 55 (1.7%) patients; P = 0.08].⁶⁷⁹ In the HARMONY (A Study to Evaluate the Effect of Ranolazine and Dronedarone When Given Alone and in Combination in Patients With Paroxysmal Atrial Fibrillation) trial, the highest tested dose of a combination of ranolazine (750 mg twice daily) and dronedarone (225 mg twice daily) slightly reduced AF burden in 134 subjects with paroxysmal AF and dual-chamber pacemakers.⁶⁸⁰ Small, open-label studies suggest that ranolazine might enhance the antiarrhythmic effect of amiodarone for cardioversion,^681–683 whereas the results from a controlled trial of ranolazine and the ranolazine–dronedarone combination to prevent AHRE in pacemaker patients were ambiguous.⁶⁸⁴ At present, there is insufficient evidence to recommend ranolazine as an antiarrhythmic drug, alone or in combination with other antiarrhythmic drugs. Of note, the ‘funny channel blocker’ ivabradine, which is used for angina and heart failure, increases the risk of AF.⁶⁸⁵

11.2.4 Antiarrhythmic effects of non-antiarrhythmic drugs

ACE inhibitors or ARBs appear to prevent new-onset AF in patients with LV dysfunction and in hypertensive patients with LV hypertrophy.^{219,236,237,239,246,250,686} Neprilysin inhibition needs to be studied further, but does not seem to enhance this effect.²²⁴ A Danish cohort study also suggested that initial treatment of uncomplicated hypertension with ACE inhibitors or ARBs reduces incident AF compared with other hypertensive agents.²⁴⁵ ARB therapy did not reduce the AF burden in patients with AF without structural heart disease.²⁴¹ Thus, ACE inhibitors or ARBs are unlikely to have a relevant direct antiarrhythmic effect. However, it might be justified to consider adding ACE inhibitors or ARB therapy to antiarrhythmic drugs to reduce AF recurrences after cardioversion.^248,249,687

Compared with placebo, beta-blockers are associated with a reduced risk of new-onset AF in patients with HFrEF and sinus rhythm.²³ Beta-blockers have also been reported to reduce symptomatic AF recurrences,^580,636,688 but this finding may be driven by the beneficial effect of rate control, which will render AF more often asymptomatic.

Peri-operative statin therapy appeared to reduce the risk of post-operative AF in a number of small RCTs^689,690; however, an adequately powered placebo-controlled trial has shown no effect of peri-operative rosuvastatin therapy on post-operative AF.⁶⁹¹ Statin treatment does not prevent AF in other settings.^692,693 Similarly, polyunsaturated fatty acids failed to show convincing benefit.^{241,694–698} The role of aldosterone antagonists in the management of AF has not been extensively investigated in humans. Although preliminary evidence from trials of eplerenone is encouraging for primary prevention,²⁴³ at present there is no robust evidence to make any recommendation for the use of aldosterone antagonists for secondary prevention of AF.^699–701

Recommendations for rhythm control therapy

Recommendations for rhythm control therapy (continued)

ACE = angiotensin-converting enzyme; AF = atrial fibrillation; ARB = angiotensin receptor blocker; CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female); ECG = electrocardiogram; LV = left ventricular; LVH = left ventricular hypertrophy; NOAC = non-vitamin K antagonist oral anticoagulant; TOE = transoesophageal echocardiography.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

11.3 Catheter ablation

Since the initial description of triggers in the pulmonary veins that initiate paroxysmal AF,¹⁰⁸ catheter ablation of AF has developed from a specialized, experimental procedure into a common treatment to prevent recurrent AF.^587,715 This is primarily achieved through isolation of the pulmonary veins, probably requiring complete isolation for full effectiveness,⁷¹⁶ and additional ablation in the posterior left atrial wall. AF ablation, when performed in experienced centres by adequately trained teams, is more effective than antiarrhythmic drug therapy in maintaining sinus rhythm, and the complication rate, though not negligible, is similar to the complication rate for antiarrhythmic drugs.^585,717

11.3.1 Indications

Catheter ablation of AF is effective in restoring and maintaining sinus rhythm in patients with symptomatic paroxysmal, persistent, and probably long-standing persistent AF, in general as second-line treatment after failure of or intolerance to antiarrhythmic drug therapy. In such patients, catheter ablation is more effective than antiarrhythmic drug therapy.^{185,586,713,717–720} As first-line treatment for paroxysmal AF, randomized trials showed only modestly improved rhythm outcome with catheter ablation compared to antiarrhythmic drug therapy.^{585,721–723} Complication rates were similar, when ablation was performed in expert centres, justifying catheter ablation as first-line therapy in selected patients with paroxysmal AF who ask for interventional therapy. Fewer data are available reporting the effectiveness and safety of catheter ablation in patients with persistent or long-standing persistent AF, but all point to lower recurrence rates after catheter ablation compared to antiarrhythmic drug therapy with or without cardioversion (Web Figure 2).^{185,717,723–726,1039} In patients who experience symptomatic recurrences of AF despite antiarrhythmic drug therapy, all RCTs showed better sinus rhythm maintenance with catheter ablation than on antiarrhythmic drugs.^{586,713,727,728} There is no current indication for catheter ablation to prevent cardiovascular outcomes (or desired withdrawal of anticoagulation), or to reduce hospitalization.^40,594

11.3.2 Techniques and technologies

Complete pulmonary vein isolation (PVI) on an atrial level is the best documented target for catheter ablation,^{716,729–731} achievable by point-by-point radiofrequency ablation, linear lesions encircling the pulmonary veins, or cryoballoon ablation, with similar outcomes.^732–734 Complete isolation of the pulmonary veins has better rhythm outcomes than incomplete isolation.⁷¹⁶ PVI was initially tested in patients with paroxysmal AF, but appears to be non-inferior to more extensive ablation in persistent AF as well.^729,735 More extensive ablations have been used in patients with persistent AF, but there are insufficient data to guide the use of these at present.^{117,718,719,735–737} Extended ablation procedures (beyond PVI) consistently require longer procedures and more ionizing radiation, potentially creating risk for patients. Left atrial macro re-entrant tachycardia is relatively uncommon after PVI (≈5%). It also seems rare after cryoballoon ablation,⁷³⁴ but may occur in up to 25% of patients after left atrial substrate modification ablation, often due to incomplete ablation lines. Thus, for patients with persistent AF, ablation of complex fractionated electrograms, ablation of rotors, or routine deployment of linear lesions or other additional ablations does not seem justified in the first procedure.^735,738,739 However, additional ablation on top of complete PVI⁷¹⁶ may be considered in patients with recurrent AF after the initial ablation procedure.^719,740,741 In patients with documented right atrial isthmus-dependent flutter undergoing AF ablation, right atrial isthmus ablation is recommended. Adenosine testing to identify patients in need of additional ablation remains controversial after evaluation in several reports.^{739,742–744} Ablation of so-called ‘rotors’, guided by body surface mapping or endocardial mapping, is under evaluation and cannot be recommended for routine clinical use at present.

11.3.3 Outcome and complications

11.3.3.1 Outcome of catheter ablation for atrial fibrillation

The rhythm outcome after catheter ablation of AF is difficult to predict in individual patients.^{173,227,713,728} Most patients require more than one procedure to achieve symptom control.^713,726,728 In general, better rhythm outcome and lower procedure-related complications can be expected in younger patients with a short history of AF and frequent, short AF episodes in the absence of significant structural heart disease.⁷⁴⁵ Catheter ablation is more effective than antiarrhythmic drug therapy in maintaining sinus rhythm (Web Figure 2).^746,1039 Sinus rhythm without severely symptomatic recurrences of AF is found in up to 70% of patients with paroxysmal AF, and around 50% in persistent AF.^713,728,735 Very late recurrence of AF after years of sinus rhythm is not uncommon and may reflect disease progression, with important implications for continuation of AF therapies.⁷²⁸ Multiple variables have been identified as risk factors for recurrence after catheter ablation of AF, but their predictive power is weak. The decision for catheter ablation, thus, should be based on a shared decision-making process⁷⁴⁷ (see Chapter 8), following thorough explanation of the potential benefits and risks, and of the alternatives such as antiarrhythmic drugs or acceptance of the current symptoms without rhythm control therapy.¹⁷⁵

11.3.3.2 Complications of catheter ablation for atrial fibrillation

There is a clear need to systematically capture complications in clinical practice to improve the quality of AF ablation procedures.¹⁷⁵ The median length of hospital stay in AF patients undergoing their first ablation as part of the EURObservational Research Programme (EORP) was 3 days (interquartile range 2–4 days), based on data from 1391 patients from hospitals performing at least 50 ablations per year. Five to seven per cent of patients will suffer severe complications after catheter ablation of AF, and 2–3% will experience life-threatening but usually manageable complications.^{727,748–750} Intraprocedural death has been reported, but is rare (<0.2%).⁷⁵¹ The most important severe complications are stroke/TIA (<1%), cardiac tamponade (1–2%), pulmonary vein stenosis, and severe oesophageal injury leading to atrio-oesophageal fistula weeks after ablation (Table 18). ‘Silent strokes’ (i.e. white matter lesions detectable by brain MRI) have been observed in around 10% of patients treated with radiofrequency and cryoballoon ablation.⁷⁵² The clinical relevance of this observation is unclear.⁷⁴⁹ Post-procedural complications include stroke, with the highest risk within the first week,⁷⁵³ late pericardial tamponade several days after catheter ablation,⁷⁵¹ and oesophageal fistulas, which usually become apparent 7–30 days after ablation. Timely detection of atrio-oesophageal fistulas can be life-saving and should be based on the typical triad of infection without a clear focus, retrosternal pain, and stroke or TIA.⁷⁴⁸

Table 18

Complications related to catheter ablation of atrial fibrillation

TIA = transient ischaemic attack.

^aOesophageal fistula should be suspected in patients presenting with the triad of unspecific signs of infection, chest pain, and stroke or TIA in the first weeks after an ablation procedure. It requires immediate therapy.

^b<10% for cryoablation or radiofrequency ablation, >20% for phased radiofrequency ablation.

11.3.4 Anticoagulation: before, during, and after ablation

Patients anticoagulated with VKAs should continue therapy during ablation (with an INR of 2–3).⁷⁶⁰ Anticoagulation with NOACs is an alternative to warfarin.^{478,761–765} There is no safety signal from observational cohorts treated with uninterrupted NOAC therapy undergoing catheter ablation in experienced centres.^{761,763,766,767} The first controlled trial comparing continuous NOAC and VKA therapy in AF ablation patients, enrolling around 200 patients, has recently been published,⁷⁶⁸ as well as several observational data sets.^761,769,770 Ongoing studies compare uninterrupted VKA with NOAC therapy in AF patients undergoing ablation [e.g. AXAFA – AFNET 5 (Anticoagulation using the direct factor Xa inhibitor apixaban during Atrial Fibrillation catheter Ablation: Comparison to vitamin K antagonist therapy; NCT02227550) and RE-CIRCUIT (Randomized Evaluation of dabigatran etexilate Compared to warfarin in pulmonaRy vein ablation: assessment of different peri-proCedUral anticoagulation sTrategies; NCT02348723)]. During ablation, heparin should be given to maintain an activated clotting time >300 s. Anticoagulation should be maintained for at least 8 weeks after ablation for all patients. The true incidence of thrombo-embolic events after catheter ablation has never been systematically studied and the expected stroke risk has been adopted from non-ablation AF cohorts. Although observational studies suggest a relatively low stroke rate in the first few years after catheter ablation of AF,^{737,771–776} the long-term risk of recurrent AF and the safety profile of anticoagulation in ablated patients need to be considered. In the absence of controlled trial data, OAC after catheter ablation should follow general anticoagulation recommendations, regardless of the presumed rhythm outcome.

11.3.5 Ablation of atrial fibrillation in heart failure patients

Catheter ablation, compared with amiodarone therapy, significantly reduces recurrent AF in AF patients with HFrEF.⁷⁷⁷ Selected patients with HFrEF and AF can achieve recovery of LV systolic function after catheter ablation (probably reflecting tachycardiomyopathy). Several smaller trials suggest improved LV function after catheter ablation in HFrEF patients^{185,226–228,778,779} and reduced hospitalizations,^720,777 especially in patients without a previous myocardial infarction.⁷⁸⁰ Larger trials are warranted to confirm these findings. Catheter ablation can be demanding in these patients. Thus, indications for catheter ablation in HFrEF patients should be carefully balanced, and the procedures performed in experienced centres.

11.3.6 Follow-up after catheter ablation

Patients and physicians involved in the follow-up after catheter ablation should know the signs and symptoms of late complications to allow swift referral for treatment (Table 18). Patients should also be aware that symptomatic and asymptomatic AF recurrences are frequent after catheter ablation.^119,781,782 In line with the primary goal of rhythm control therapy, asymptomatic episodes should generally not trigger further rhythm control therapy in routine care. Patients should be seen at least once by a rhythm specialist in the first 12 months after ablation. Further rhythm control options should be considered in patients with symptomatic recurrences, including discussion in a Heart Team (Figure 17, Figure 19).

11.4 Atrial fibrillation surgery

11.4.1 Concomitant atrial fibrillation surgery

The Cox maze procedure was first performed 30 years ago as a ‘cut-and-sew’ technique, including isolation of the posterior left atrium, a connection to the posterior mitral annulus, a cavotricuspid connection, a cavocaval connection, and exclusion of the LAA (Figure 18).⁷⁸³ Thereby, the Cox maze procedure creates an electrical labyrinth (maze) of passages through which the sinoatrial node impulse finds a route to the atrioventricular node while preventing fibrillatory conduction. The Cox maze procedure and other, often simpler, forms of AF surgery have mainly been used in patients undergoing other open heart surgical procedures.^{461,466,784–798} In a systematic review commissioned for these guidelines, performing concomitant AF surgery resulted in increased freedom from AF, atrial flutter, and atrial tachycardia compared to no concomitant AF surgery (RR 1.94; 95% CI 1.51–2.49; n = 554 from seven RCTs) (Web Figure 3).¹⁰⁴⁰ Patients undergoing the Cox maze procedure required pacemaker implantation more often (RR 1.69; 95% CI 1.12–2.54; n = 1631 from 17 RCTs), without a detectable difference in other outcomes or complications. These findings are underpinned by an analysis of the Society of Thoracic Surgeons database comprising 67 389 patients in AF undergoing open heart surgery: mortality or major morbidity was not affected by concomitant AF surgery (adjusted OR 1.00; 95% CI 0.83–1.20), but pacemaker implantation was more frequent (adjusted OR 1.26; 95% CI 1.07–1.49).⁷⁹⁹ Predictors of AF recurrence after surgery include left atrial dilatation, older age, >10-year history of AF, and non-paroxysmal AF.^800–804 Regarding AF type, surgical PVI seems effective in paroxysmal AF.⁸⁰⁵ Biatrial lesion patterns may be more effective in persistent and long-standing persistent AF.^797,803,806 The suggested management of patients with AF-related symptoms undergoing cardiac surgery is displayed in Figure 19, with an important contribution of the AF Heart Team to advise and inform patient choice.

Figure 18

A: Surgical lesion sets for the biatrial Cox maze procedure. Surgeon's view showing left atrial lesions (left panel) and right atrial lesions (middle and right panel). B: Left atrial lesions in a thoracoscopic minimally invasive surgical procedure (dashed lines), including left atrial appendage exclusion (double line).

Figure 19

Surgical rhythm control in patients with atrial fibrillation undergoing cardiac surgery.

11.4.2 Stand-alone rhythm control surgery

Current technology (e.g. bipolar radiofrequency or cryothermy) renders the Cox maze procedurec easier, and more reproducible and feasible, via a mini-thoracotomy.^786,807,808 Thoracoscopic PVI with bipolar radiofrequency prevents recurrence of paroxysmal AF (69–91% freedom from arrhythmias at 1 year, see Figure 18B for lesion set),^468,809,810 and seems effective in patients refractory to catheter ablation.⁸¹¹ The average length of hospital stay for thoracoscopic ablation varies from 3.6 to 6.0 days.^468,812,813 The FAST (Atrial Fibrillation Catheter Ablation vs. Surgical Ablation Treatment) trial,⁴⁶⁸ and another smaller trial,⁸¹⁴ suggested that thoracoscopic AF surgery could be more effective than catheter ablation for the maintenance of sinus rhythm,^468,814 while also causing more complications (Table 19).⁸¹⁵ To improve results,^{468,816–818} more extensive lesion sets have been performed, including connecting lines between the PVI (“box lesion”) and lines towards the mitral annulus.^{812,819–822} To improve the generation of transmural lesions,⁷¹⁶ endo-epicardial ablation strategies have recently been proposed.^{812,823–825} Although preliminary experience with hybrid simultaneous ablation shows promise, procedural time and rates of bleeding complications are higher.^812,823

Table 19

Complications of thoracoscopic atrial fibrillation surgery

^aThe rate of asymptomatic cerebral embolism is unknown.

11.5 Choice of rhythm control following treatment failure

There is insufficient evidence to underpin clear recommendations on how to treat patients with recurrent AF after catheter ablation. Early recurrences of AF or atrial tachycardias after ablation (occurring within 8 weeks) may be treated with cardioversion. Many of the published series of patients undergoing AF ablation included those who failed antiarrhythmic drug therapy. Thus, considering ablation therapy in patients who have symptomatic recurrences on antiarrhythmic drug therapy is often reasonable. Alternatively, trialling another antiarrhythmic drug can be considered. Combining an antiarrhythmic drug with ablation (‘hybrid therapy’, see Chapter 12) should be considered based on the different and possibly synergistic effects of these drugs with AF ablation, possibly benefitting patients in whom either treatment alone was previously ineffective. Rate control without rhythm control, surgical ablation, or repeat catheter ablation should be considered as well (Figure 20). Patient preferences and local access to therapy are important considerations to inform the therapy choice in patients who are in need of further rhythm control therapy after an initial therapy failure.

Figure 20

Choice of rhythm control therapy following treatment failure.

11.6 The Atrial Fibrillation Heart Team

In view of the complexity of the different treatment options in patients with failed rhythm control therapy who still require or demand further rhythm control therapy, this Task Force proposes that decisions involving AF surgery or extensive AF ablation should be based on advice from an AF Heart Team. This will also apply to reversal to a rate control strategy in patients with severe (EHRA III or IV) AF symptoms. An AF Heart Team should consist of a cardiologist with expertise in antiarrhythmic drug therapy, an interventional electrophysiologist, and a cardiac surgeon with expertise in appropriate patient selection, techniques, and technologies for interventional or surgical AF ablation (Figure 20). Such AF Heart Teams—and a collaborative infrastructure supporting a continued interaction between physicians delivering continued care, AF cardiologists, interventional electrophysiologists, and AF surgeons—should be established to provide optimal advice, and ultimately to improve rhythm outcomes for patients in need of advanced and complex rhythm control interventions.

Recommendations for catheter ablation of atrial fibrillation and atrial fibrillation surgery

AF = atrial fibrillation; NOAC = non-vitamin K antagonist oral anticoagulant; VKA = vitamin K antagonist.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

12. Hybrid rhythm control therapy

AF has many different drivers, which are only partially targeted by antiarrhythmic drugs or catheter ablation.⁹⁶ Hence, combination or ‘hybrid’ rhythm control therapy seems reasonable, although there is little evidence from controlled trials supporting its use.

12.1 Combining antiarrhythmic drugs and catheter ablation

Antiarrhythmic drug therapy is commonly given for 8–12 weeks after ablation to reduce early recurrences of AF after catheter ablation, supported by a recent controlled trial where amiodarone halved early AF recurrences compared with placebo.⁶⁵⁰ Prospective studies have not been done, but a meta-analysis of the available (weak) evidence suggests slightly better prevention of recurrent AF in patients treated with antiarrhythmic drugs after catheter ablation.⁷¹³ Many patients are treated with antiarrhythmic drug therapy after catheter ablation (most often amiodarone or flecainide),⁵⁸⁷ and this seems a reasonable option in patients with recurrent AF after ablation. It seems common sense to consider antiarrhythmic drug therapy in patients who are in need of further rhythm control therapy after catheter ablation, but controlled trials to confirm this are desirable.

Combining cavotricuspid isthmus ablation and antiarrhythmic drugs may lead to improved rhythm control without the need for left atrial ablation in patients who develop ‘drug-induced atrial flutter’ on therapy with flecainide, propafenone, or amiodarone,^834–836 although recurrent AF is a concern in the long-term.^837,838

12.2 Combining antiarrhythmic drugs and pacemakers

In selected patients with sick sinus syndrome and fast ventricular response during AF paroxysms requiring rate control therapy, the addition of a pacemaker not only optimizes rate control but may also help to control rhythm.^711,712 Moreover, when antiarrhythmic drug treatment leads to sinus node dysfunction and bradycardia, pacing may permit uptitration of the antiarrhythmic drug dose. Such strategies have never been prospectively investigated and the existing populations studied are highly selected.^839,840 Some patients with AF-induced bradycardia may benefit from catheter ablation of AF, obviating the need for antiarrhythmic drugs and pacemaker implantation.^829,830

13. Specific situations

13.1 Frail and ‘elderly’ patients

Many AF patients present at older age (e.g. >75 or >80 years). There are no studies suggesting that cardiovascular risk reduction is less effective in these ‘elderly’ AF patients than in younger patients. Rather, age is one of the strongest predictors/risk factors for ischaemic stroke in AF.³⁸² Good data are available to support the use of anticoagulants in older patients from BAFTA (Birmingham Atrial Fibrillation Treatment of the Aged Study),³⁶² the NOAC trials,³⁹ and from analyses in elderly Americans (Medicare).³⁹⁶ Elderly AF patients are at higher risk of stroke and, thus, are more likely to benefit from OAC than younger patients,⁸⁴¹ and yet OAC is still underutilized in the elderly.^220,842 Although the evidence base is smaller for other treatment options in AF, the available data support the use of available rate and rhythm control interventions, including pacemakers and catheter ablation, without justification to discriminate by age group. Individual patients at older age may present with multiple comorbidities including dementia, a tendency to falls, CKD, anaemia, hypertension, diabetes, and cognitive dysfunction. Such conditions may limit quality of life more than AF-related symptoms. Impairment of renal and hepatic function and multiple simultaneous medications make drug interactions and adverse drug reactions more likely. Integrated AF management and careful adaptation of drug dosing seem reasonable to reduce the complications of AF therapy in such patients.⁸⁴³

13.2 Inherited cardiomyopathies, channelopathies, and accessory pathways

Several inherited cardiac conditions are associated with early-onset AF (Table 20). Treatment of the underlying cardiac condition is an important contribution to AF management in these young patients (see also ESC guidelines on sudden cardiac death⁸⁴⁴ and hypertrophic cardiomyopathy⁸⁴⁵).

Table 20

Inherited cardiomyopathies, channelopathies, and pathways associated with atrial fibrillation

AF = atrial fibrillation; VT = ventricular tachycardia.

13.2.1 Wolff–Parkinson–White syndrome

Patients with pre-excitation and AF are at risk of rapid conduction across the accessory pathway, resulting in a fast ventricular rate, possible ventricular fibrillation, and sudden death. In AF patients with evidence of an antegrade accessory pathway, catheter ablation of the pathway is recommended.^869,870 This procedure is safe and effective and may be considered as a prophylactic treatment strategy.^871,872 In AF patients surviving a sudden death event with evidence of an accessory pathway, urgent catheter ablation of the pathway is recommended.⁸⁶⁹ A documented short pre-excited RR interval (<250 ms) during spontaneous or induced AF is one of the risk markers for sudden death in Wolff–Parkinson–White syndrome (WPW) syndrome, in addition to a history of symptomatic tachycardia, the presence of multiple accessory pathways, and Ebstein's anomaly. Intravenous procainamide, propafenone, or ajmaline can be used to acutely slow ventricular rate,^873,874 whereas digoxin, verapamil, and diltiazem are contraindicated.⁸⁷⁵ Intravenous amiodarone should be used with caution, as there are case reports of accelerated ventricular rhythms and ventricular fibrillation in patients with pre-excited AF receiving intravenous amiodarone infusion.⁸⁷⁶

13.2.2 Hypertrophic cardiomyopathy

AF is the most common arrhythmia in patients with hypertrophic cardiomyopathy, affecting approximately one-quarter of this population.⁸⁷⁷ Observational data highlight a high stroke risk in hypertrophic cardiomyopathy patients with AF, confirming the need for OAC.⁸⁷⁸ While there is more experience with VKAs, there are no data to suggest that NOACs cannot be used in these patients.⁸⁴⁵ Studies of rate or rhythm control medications in patients with hypertrophic cardiomyopathy are relatively scarce. Beta-blockers and diltiazem or verapamil seem reasonable treatment options for rate control in these patients. In the absence of significant LV outflow tract obstruction, digoxin can be used alone or in combination with beta-blockers.⁸⁴⁵ Amiodarone seems a safe antiarrhythmic drug in AF patients with hypertrophic cardiomyopathy,⁸⁷⁹ and expert opinion suggests that disopyramide may be beneficial in those with outflow tract obstruction. AF ablation is effective to suppress symptomatic AF recurrences.^880–884 Surgical treatment of AF may be appropriate in patients with hypertrophic cardiomyopathy undergoing surgery (e.g. for LV outflow tract obstruction or mitral valve surgery), but experience is limited.

13.2.3 Channelopathies and arrhythmogenic right ventricular cardiomyopathy

Many channelopathies and inherited cardiomyopathies are associated with AF. AF prevalence ranges from 5–20% in patients with long QT syndrome or Brugada syndrome, and is up to 70% in short QT syndrome (Table 20).^{853,856–858} Penetrance of disease phenotype including AF is variable.^{61,852,885,886} Both shortening as well as prolongation of the atrial action potential have been demonstrated as likely mechanisms underlying AF in these diseases. It seems reasonable to consider antiarrhythmic drugs that reverse the suspected channel defect in AF patients with inherited cardiomyopathies (e.g. a sodium channel blocker in LQT3⁸⁵², or quinidine in Brugada syndrome⁸⁸⁷). More importantly, new-onset AF in young, otherwise healthy individuals should trigger a careful search for such inherited conditions, including clinical history, family history, ECG phenotype, and echocardiography and/or other cardiac imaging.

Monogenic defects only account for 3–5% of all patients with AF, even in younger populations.^{846,848,888–890} Furthermore, there is no clear link between detected mutations and specific outcomes or therapeutic needs. For these reasons, genetic testing is not recommended in the general AF population.⁷⁷ Other guidelines have described the indications for genetic testing in patients with inherited arrhythmogenic diseases.^844,891

Recommendations for inherited cardiomyopathies

AF = atrial fibrillation; HCM = hypertrophic cardiomyopathy; LV = left ventricular; WPW = Wolff–Parkinson–White syndrome.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

13.3 Sports and atrial fibrillation

Physical activity improves cardiovascular health, which translates into a lower risk of AF.⁸⁹⁸ Therefore, physical activity is a cornerstone of preventing AF. Intensive sports practice, especially endurance sports (>1500 h of endurance sports practice),⁸⁹⁹ increases the risk of AF later in life,^900–902 probably mediated by altered autonomic tone, volume load during exercise, atrial hypertrophy, and dilatation.^903,904 This results in a U-shaped relationship of physical activity and incident AF.^{214,898,902,905,906} Detraining can reduce AF in models⁹⁰⁴ and reduces ventricular arrhythmias in athletes,⁹⁰⁷ but the role of detraining for AF in human athletes is unknown. The management of athletes with AF is similar to general AF management, but requires a few special considerations. Clinical risk factors will determine the need for anticoagulation. Sports with direct bodily contact or prone to trauma should be avoided in patients on OAC. Beta-blockers are not well tolerated and at times prohibited, and digoxin, verapamil, and diltiazem are often not potent enough to slow heart rate during exertional AF. Catheter ablation for AF probably has similar outcomes in athletes as in non-athletes,^908,909 but further data are needed. Pill-in-the-pocket therapy has been used as well.⁶²⁰ After ingestion of flecainide or propafenone as pill-in-the-pocket, patients should refrain from sports as long as AF persists and until two half-lives of the antiarrhythmic drug have elapsed. Prophylactic ablation of the flutter circuit may be considered in athletes treated with sodium channel blockers.⁹¹⁰

Recommendations for physical activity in patients with atrial fibrillation

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

13.4 Pregnancy

AF in pregnant women is rare and is usually associated with pre-existing heart disease. AF is associated with increased complications for the mother and foetus.^911,912 Better treatment of congenital heart diseases will probably increase the incidence of AF during pregnancy in the future.⁹¹³ Pregnant women with AF should be managed as high-risk pregnancies in close collaboration with cardiologists, obstetricians, and neonatologists.

13.4.1 Rate control

Owing to a lack of specific data, beta-blockers, verapamil, diltiazem, and digoxin all carry a US Food and Drug Administration pregnancy safety category of C (benefits may outweigh risk), except for atenolol (category D: positive evidence of risk). Their use should be at the lowest dose and for the shortest time required. None of the agents are teratogenic, but they readily cross the placenta.⁹¹⁴ Beta-blockers are commonly used in pregnant women with cardiovascular conditions (e.g. for management of gestational hypertension and pre-eclampsia), but may be associated with intrauterine growth retardation,⁹¹⁵ and hence growth scans after 20 weeks' gestation are recommended. Digoxin is considered safe for maternal and foetal arrhythmias.⁹¹⁶ There are insufficient data to comment on verapamil or diltiazem, hence rate control using beta-blockers and/or digoxin is recommended.⁹¹⁷ With regards to breastfeeding, all rate control agents are present in breast milk, although levels of beta-blockers, digoxin, and verapamil are too low to be considered harmful. Diltiazem will be present at high levels and should be considered second-line treatment.⁹¹⁸

13.4.2 Rhythm control

Rhythm control therapy in pregnant patients with AF has only been reported in case studies. Amiodarone is associated with severe adverse foetal side-effects and should only be considered for emergency situations.⁹¹⁹ Flecainide and sotalol can both be used for conversion of foetal arrhythmias without major adverse effects,⁹²⁰ and thus are likely to be safe to treat maternal symptomatic AF. Electrical cardioversion can be effective for restoration of sinus rhythm when tachyarrhythmia is causing haemodynamic instability, with low rates of adverse outcomes for both mother and foetus.⁹²¹ However, in view of the risk of foetal distress, electrical cardioversion should only be carried out where facilities are available for foetal monitoring and emergency caesarean section. As with other emergencies during pregnancy, patients should receive 100% oxygen, intravenous access should be established early, and the mother should be positioned in the left lateral position to improve venous return.⁹²²

13.4.3 Anticoagulation

VKAs should be avoided in the first trimester because of teratogenic effects, and in the 2–4 weeks preceding delivery to avoid foetal bleeding. Low-molecular-weight heparins are a safe substitute, as they do not cross the placenta.⁹²³ In the third trimester, frequent laboratory checks for adequate anticoagulation (e.g. every 10–14 days) and corresponding dose adjustments are advised, given that in some women high doses of both VKA and heparin may be needed to maintain adequate anticoagulation. Pregnant patients with AF and mechanical prosthetic valves who elect to stop VKA treatment in consultation with their specialist team between 6–12 weeks of gestation, should receive continuous, dose-adjusted unfractionated heparin or dose-adjusted subcutaneous low-molecular-weight heparin. As only limited data are available about teratogenesis for NOACs, these drugs should be avoided during pregnancy.

Recommendations during pregnancy

NOAC = non-vitamin K antagonist oral anticoagulants.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

13.5 Post-operative atrial fibrillation

AF is common after cardiac surgery (occurring in 15–45% of patients),^924–926 and is associated with increased length of hospital stay and higher rates of complications and mortality.⁹²⁷ Post-operative AF is also not uncommon after other major surgery, especially in elderly patients. The treatment of post-operative AF is mainly based on studies of patients undergoing cardiac surgery, with much less evidence in the non-cardiac surgery setting.

13.5.1 Prevention of post-operative atrial fibrillation

Beta-blockers reduce post-operative AF and supraventricular tachycardias, albeit with high heterogeneity and moderate risk of bias in a systematic review of published studies. The most commonly studied drug was propranolol, with AF in 16.3% of the treatment group vs. 31.7% in the control group.⁹²⁵ In the majority of these studies, beta-blockers were administered post-operatively, a regimen supported in a recent meta-analysis.⁹²⁸ Amiodarone reduced the incidence of post-operative AF compared to beta-blocker therapy in several meta-analyses, also reducing hospital stay.^{925,929–931}

Despite initial reports from meta-analyses,^689,932,933 pre-operative treatment with statins did not prevent post-operative AF in a prospective controlled trial.⁹³⁴ Other therapies have also been studied in small, hypothesis-generating trials, but have not demonstrated clear beneficial effects. These include magnesium,^925,935,936 n-3 polyunsaturated fatty acids,^937–945 colchicine,⁹⁴⁶ corticosteroids,^947,948 and posterior pericardectomy.⁹⁴⁹ Post-operative overdrive biatrial pacing has not gained widespread use despite some suggestion of prophylactic effects.^925,950

13.5.2 Anticoagulation

Post-operative AF is associated with an increased early stroke risk, increased morbidity, and 30-day mortality.^927,951,952 In the long-term, patients with an episode of post-operative AF have a two-fold increase in cardiovascular mortality, and a substantially increased risk of future AF and ischaemic stroke, compared with patients that remain in sinus rhythm after surgery.^952–958 OAC at discharge has been associated with a reduced long-term mortality in patients with post,operative AF,⁹⁵⁹ without evidence from controlled trials. Good quality data are needed to determine whether long-term anticoagulation can prevent strokes in patients with post-operative AF at high stroke risk,^368,386 and to assess whether short episodes of post-operative AF (e.g. <48 h) carry a similar risk as longer episodes.⁹⁶⁰ The indication and timing of OAC in post-operative AF patients should take into consideration the risk of post-operative bleeding.

13.5.3 Rhythm control therapy in post-operative atrial fibrillation

In haemodynamically unstable patients, cardioversion and consideration of antiarrhythmic drugs is recommended. Amiodarone or vernakalant have been efficient in converting post-operative AF to sinus rhythm.^603,950,961 A recent medium-sized trial randomizing patients with post-operative AF to either rhythm control therapy with amiodarone or to rate control did not find a difference in hospital admissions during a 60-day follow-up,⁹⁶² underpinning that the aim of rhythm control therapy should be to improve AF-related symptoms in post-operative AF. In asymptomatic patients and in those with acceptable symptoms, rate control or deferred cardioversion preceded by anticoagulation is a reasonable approach.

Recommendations for preventing postoperative atrial fibrillation

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

13.6 Atrial arrhythmias in grown-up patients with congenital heart disease

Atrial arrhythmias (AF, atrial flutter, atrial tachycardias) often occur late after surgical repair of congenital heart defects, occurring in 15–40% of grown-up patients with congenital heart disease (GUCH). They are associated with heart failure, syncope, thrombo-embolic events, and sudden death.^963–967 The pathophysiological substrate is complex, associated with hypertrophy, fibrosis, hypoxaemia, chronic haemodynamic overload, and surgical scars and patches. Additionally, related primary anomalies in the conduction pathways can lead to reentrant atrial and ventricular tachycardia, heart block, and sinus node dysfunction.⁹⁶³ Macroreentrant atrial tachycardia or atypical atrial flutter may be seen after nearly any surgical procedure involving atriotomy or atrial patches.

13.6.1 General management of atrial arrhythmias in grown-up patients with congenital heart disease

The conventional stroke risk factors should be used to inform decisions on long-term anticoagulation in GUCH patients with AF. In addition, anticoagulation should be considered in GUCH patients with atrial arrhythmias when they present with intracardiac repair, cyanosis, Fontan palliation, or systemic right ventricle.⁹⁶⁸ Beta-blockers, verapamil, diltiazem, and digitalis can be used. Care should be taken to avoid bradycardia and hypotension.

Sodium channel blockers suppress approximately half of atrial arrhythmias in Fontan patients.⁹⁶⁹ Amiodarone is more effective, but long-term treatment with an antiarrhythmic drug carries a high risk of extracardiac side-effects in this relatively young population. Intracardiac thrombi are common in GUCH patients undergoing cardioversion for AF, but also in patients with atrial tachycardias or atrial flutter.⁹⁷⁰ Therefore, both a TOE and anticoagulation for a few weeks before the planned cardioversion should be considered.⁹⁶⁴ Radiofrequency ablation may be a good option for symptomatic GUCH patients with atrial arrhythmias, especially in those with atrial flutter and other macro re-entrant tachycardias. Interventions should be performed in adequately qualified centres by specialized teams.

13.6.2 Atrial tachyarrhythmias and atrial septal defects

Atrial flutter and fibrillation occur in 14–22% of adults with unoperated atrial septal defects, especially in older patients,⁹⁷¹ and can lead to heart failure.⁹⁷² Early repair can reduce but not eliminate the risk of AF.⁹⁷³ Biatrial volume overload,⁹⁷⁴ pulmonary hypertension,⁹⁷⁵ and possibly the arrhythmogenic effect of atrial patches can contribute to these arrhythmias.⁹⁷⁶ Anticoagulation should be decided upon based on stroke risk factors. In patients with a history of paroxysmal or persistent AF, AF surgery could be considered at the time of surgical closure, or catheter ablation at the time of interventional atrial septal defect closure. Catheter ablation of late atrial arrhythmias has been shown to be effective in small cohorts of patients after surgical atrial septal defect.⁹⁷⁷

13.6.3 Atrial tachyarrhythmias after Fontan operation

Atrial arrhythmias occur in up to 40% of patients with a Fontan circulation, and can manifest as atrial flutter, primary atrial tachycardia, AF, and accelerated junctional rhythm or junctional tachycardia⁹⁷⁸ with or without sinoatrial node dysfunction.⁹⁷⁹ Patients with atriopulmonary anastomoses (possibly due to higher atrial volume and pressure load) and those with early post-operative atrial arrhythmias are more likely to develop long-term atrial arrhythmias.⁹⁸⁰ Atrial arrhythmias can also be the first manifestation of obstruction of the atriopulmonary anastomosis, a complication that must be identified. Right atrial thrombus formation is common in Fontan patients with atrial arrhythmias and requires oral anticoagulation.⁹⁸¹ Operative conversion to total cavopulmonary artery connection with concomitant arrhythmia surgery can, in some patients, improve heart failure symptoms and reduce recurrent arrhythmias,^969,982 with low recurrence rates of clinically apparent atrial arrhythmias in the first few years after repeat surgery.^983–985 Catheter ablation of atrial arrhythmia in Fontan patients has been successful in selected patients.⁹⁸⁶

13.6.4 Atrial tachyarrhythmias after tetralogy of Fallot correction

After repair of tetralogy of Fallot, approximately one-third of patients develop atrial arrhythmias, including intra-atrial reentrant tachycardia, focal atrial tachycardia, and AF.⁹⁸⁷ Circuits involving the cavotricuspid isthmus and areas of presumed surgical right atrial scarring have been described as responsible for atrial arrhythmias.

Recommendations in patients with grown-up congenital heart disease

AF = atrial fibrillation; CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female); GUCH = grown-up patients with congenital heart disease; OAC = oral anticoagulation; TOE = transoesophageal echocardiography.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

13.7 Management of atrial flutter

The goals for the management of atrial flutter are similar to those for AF.⁹⁹² Based on the available evidence, the stroke risk in patients with atrial flutter is not much different from that in AF.⁸²⁷ Furthermore, many patients diagnosed with atrial flutter develop AF.^993–995 Thus, anticoagulation should  be used in patients with atrial flutter similar to that in patients with AF. Rate control in atrial flutter is achieved with the same medications as in AF, but is often more difficult to achieve. Flecainide, propafenone, dofetilide, and intravenous ibutilide are useful for cardioversion of atrial flutter. They should be combined with a rate-controlling agent to avoid 1:1 conduction of slowing flutter waves to the ventricles. Ibutilide is more effective for conversion of atrial flutter than AF, whereas vernakalant is less effective in converting typical atrial flutter.^996,997 Electrical cardioversion of atrial flutter can be performed using lower energies (50–100 J) than for AF.^998,999 Atrial overdrive pacing through pacemaker leads or endocardial or transesophageal catheters can convert atrial flutter to sinus rhythm.^1000,1001 Anticoagulation and TOE around cardioversion or overdrive pacing should be used similar to that in AF.

Ablation of the cavotricuspid isthmus for isthmus-dependent right atrial flutter (either the common counter-clockwise atrial flutter or the less-common clockwise atrial flutter) restores and maintains sinus rhythm with a success rate of 90–95%.¹⁰⁰² It may also reduce AF recurrences in selected patients,^1003,1004 and help to prevent hospitalizations.^1004,1005 Isthmus ablation is comparably safe and more effective than antiarrhythmic drug therapy, and is recommended for recurrent atrial flutter.^{585–587,713} Catheter ablation of left atrial macroreentrant tachycardia is more complex, with lower success rates and higher recurrence rates.^1006,1007

Recommendations for management of atrial flutter

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

14. Patient involvement, education, and self-management

14.1 Patient-centred care

Autonomous, informed patients are better placed to adhere to long-term therapy, and it is very likely that long-term management of chronic conditions such as AF will benefit from informed patients who are aware of their own responsibilities in the disease management process.³²⁸ Shared decision-making⁷⁴⁷ and patient-centred organization of care can help to ensure adherence to management and empower patients, and respect individual patient preferences, needs, and values (see Chapter 8.2).^{326,1008,1009} Patients in an active role tend to have better health outcomes and care experiences, and engagement itself can be considered as an intermediate outcome.¹⁰¹⁰

14.2 Integrated patient education

Education is a prerequisite for informed, involved patients and patient-centred care. However, lack of AF-related knowledge in patients is common, even in those who have received verbal and written information,^32,1011,1012 indicating the need to further develop structured patient education. Several patient information tools have been developed, largely focusing on oral anticoagulation.^1013–1016 This task force has developed a dedicated app for AF patients to support patient information and education.Understanding patients' perceptions and attitudes towards AF and its management can improve AF management and related outcomes.¹⁰¹⁷ This includes tailored patient education focusing on the disease, symptom recognition, therapy, modifiable risk factors for AF, and self-management activities.^1018,1019

14.3 Self-management and shared decision-making

Self-management is primarily focused on tasks to manage the condition, such as adhering to a therapeutic regimen or modifying behaviour (e.g. resulting in smoking cessation or weight loss).¹⁰²⁰ It requires understanding of the treatment modalities and goals.³⁵⁰ Within a multidisciplinary team, allied health professionals can guide this interactive process in which communication, trust, and reciprocal respect foster patient engagement.¹⁰²¹ Shared decision-making should be considered as a routine part of the decision-making process,⁷⁴⁷ supported by decision aids where applicable.¹⁰²² Models of care that integrate education, engagement, and shared decision making are now available,¹⁰²³ and may be of particular value in the management of AF.

Recommendations for patient involvement, education, and self-management

AF = atrial fibrillation.

^aClass of recommendation.

^bLevel of evidence.

^cReference(s) supporting recommendations.

15. Gaps in evidence

There are some areas of AF management that are supported by excellent evidence from multiple, adequately powered randomized trials (e.g. oral anticoagulation). Other areas, such as rhythm control therapy, integrated AF management, and lifestyle modifications are clearly developing the required evidence, while areas such as rate control are in dire need of better studies to underpin future guidelines. Here, we identify areas in need of further research.

15.1 Major health modifiers causing atrial fibrillation

Atrial fibrillation has different causes in different patients. More research is needed into the major causes (and electrophysiological mechanisms) of AF in different patient groups.^176,1024 Such research should consider the major comorbidities associated with AF, and characterize the response to AF therapy in patients with different, pathophysiologically distinct types of AF.

15.2 How much atrial fibrillation constitutes a mandate for therapy?

Technological advances allow screening for an irregular pulse using patient-operated ECG devices, smartphones, and a variety of other technologies. These may be very useful to detect silent, undiagnosed AF.¹⁵⁷ Adequately powered studies evaluating the diagnostic accuracy of such technologies, the diagnostic yield in different populations, the shortest duration and pattern of atrial arrhythmias conveying a stroke risk, and the effect of ECG screening on outcomes are needed.

15.3 Atrial high-rate episodes (AHRE) and need for anticoagulation

All of the information on the benefit of OAC has been generated in patients with AF diagnosed by ECG. Technological advances allow ready detection of AHRE in patients with implanted devices and an atrial lead. Such patients are at increased stroke risk, but it is unclear whether they benefit from OAC. Controlled trials evaluating OAC in AHRE patients are ongoing and will provide evidence on the best antithrombotic therapy in these patients.

15.4 Stroke risk in specific populations

Several specific AF groups should be studied to better characterize their risk for AF, stroke, and other AF-related complications (e.g. patients with one stroke risk factor, and non-Caucasian patients). Confounding factors (e.g. different therapy of concomitant cardiovascular diseases) may help to explain the variability in the reported rates of incident AF, prevalent AF, and AF complications. This also applies to the effect of gender in AF patients.⁴⁷

15.5 Anticoagulation in patients with severe chronic kidney disease

The use of NOACs has not been tested in patients with creatinine clearance <30 mL/min, and there is very little evidence on the effects of OAC in patients on haemodialysis or on other forms of renal replacement therapy. Studies evaluating OAC in patients with severe CKD are needed to inform the best management in this patient group at high risk for stroke and bleeding.

15.6 Left atrial appendage occlusion for stroke prevention

The most common justification for LAA occlusion devices in clinical practice is a perceived high bleeding risk and, less often, contraindications for OAC.⁴⁵⁹ Unfortunately, LAA occluders have not been tested in such populations. Furthermore, LAA occluders have not been compared with NOAC therapy in patients at risk for bleeding, or with thoracoscopic LAA clipping. There is a clear need to conduct adequately designed and powered trials to define the clinical role of LAA occluders compared with NOAC therapy in patients with relative or absolute contraindications for anticoagulation, and/or in those suffering from an ischaemic stroke on anticoagulant therapy.

15.7 Anticoagulation in atrial fibrillation patients after a bleeding or stroke event

At least 2% of anticoagulated patients with AF will experience a serious bleeding event per year. Observational data suggest that OAC can be reinitiated even after an intracerebral bleeding event.^460,484 Controlled studies evaluating different anticoagulation and stroke prevention interventions are urgently needed to provide evidence on the best management of patients who have suffered a bleeding event that would usually lead to withholding OAC. Some studies (e.g. APACHE-AF [Apixaban versus Antiplatelet drugs or no antithrombotic drugs after anticoagulation-associated intraCerebral HaEmorrhage in patients with Atrial Fibrillation]¹⁰²⁵) are ongoing, but adequately powered trials are needed. Similarly, prospectively collected data are needed on the stroke prevention and bleeding risk following (re-)initiation of OAC after stroke or intracranial bleeding.

15.8 Anticoagulation and optimal timing of non-acute cardioversion

Based on retrospective data, previous recommendations on the safe time-window in which a cardioversion can be performed in new-onset AF used ≤48 h as the ‘gold standard’ for non-protected cardioversion. However, new evidence has emerged that initiating pre-cardioversion anticoagulation in patients with AF episodes of <24 h or even <12 h would provide even better safety.^{642,647,1026–1028} Further research is needed to establish a clear safety margin in this clinical situation.

15.9 Competing causes of stroke or transient ischaemic attack in atrial fibrillation patients

Prospective RCTs have demonstrated the superiority of carotid endarterectomy compared to stenting in patients with symptomatic high-degree stenosis of the internal carotid artery.¹⁰²⁹ As endartectomy minimizes the need for combination therapy with OAC and antiplatelets,¹⁰³⁰ this approach has appeal in patients with AF to reduce bleeding risk. However, few of these studies included patients with AF. In a large observational study, the composite of in-hospital mortality, post-procedural stroke, and cardiac complications was higher in AF patients undergoing carotid stenting (457/7668; 6.0%) compared with endarterectomy (4438/51 320; 8.6%; P < 0.0001).¹⁰³¹ Despite adjustment for baseline risk, this may just reflect the type of patients referred for each procedure, and further randomized studies are needed to confirm the optimal treatment strategy in AF patients with carotid disease.

15.10 Anticoagulation in patients with biological heart valves (including transcatheter aortic valve implantation) and specific forms of valvular heart disease

The optimal antithrombotic therapy in the first months after biological valve replacement (including after catheter-based valve replacement) is not known. VKAs remain the mainstay during the initial post-operative period; NOACs probably deliver the same protection. In patients without AF, many centres use platelet inhibitors only. NOACs appear to be equally effective as VKAs in patients with moderate aortic stenosis, based on a subanalysis from the ROCKET-AF trial,¹⁰³² as well as the Loire Valley AF project.¹⁰³³ Further data would be helpful to confirm these observations.¹⁰³⁴ The safety and efficacy of NOACs in patients with rheumatic mitral valve disease has not been evaluated and should be studied.

15.11 Anticoagulation after ‘successful’ catheter ablation

In view of the long-term recurrence rates of AF, this Task Force recommends that OAC is continued in AF patients after ‘successful’ catheter ablation. Nonetheless, observational data suggest that the stroke risk may be lower after catheter ablation of AF compared with other AF patients. The ongoing EAST – AFNET 4 trial will inform, in a more general way, whether rhythm control therapy can reduce stroke rates in anticoagulated AF patients. In addition, there seems to be a place for a controlled trial evaluating the termination of OAC therapy at an interval after ‘successful’ catheter ablation.

15.12 Comparison of rate control agents

Although the use of rate control therapy is very common in AF patients, robust data comparing rate control therapies are scant, with the majority of studies being small uncontrolled trials over short periods of follow-up. Some studies are ongoing [e.g. RATE-AF (Rate Control Therapy Evaluation in Permanent Atrial Fibrillation)⁵⁵⁹] and will investigate the potential benefits of different rate-controlling agents, characteristics, or biomarkers that can help to personalize the use of rate control, and the adverse event profile of specific drugs in defined groups of patients.

15.13 Catheter ablation in persistent and long-standing persistent atrial fibrillation

While a few recent randomized studies support the use of catheter or surgical ablation in patients with persistent AF and long-standing persistent AF,¹⁰⁴² there is a clear need for more data evaluating this intervention in adequately powered randomized trials.

15.14 Optimal technique for repeat catheter ablation

PVI emerges as the most important target for catheter ablation of AF. Although a plethora of different additional ablation techniques have been published, their added value is questionable in patients undergoing a first catheter ablation, including those with persistent AF.^735,1042 Many patients are in need of multiple catheter ablation procedures, and such interventions often follow local or operator-specific protocols without clear evidence to support the choice of ablation target or intervention. There is a clear clinical need to define the best approach in patients who are in need of a second ablation procedure.

15.15 Combination therapy for maintenance of sinus rhythm

In the follow-up after initially successful catheter ablation, even when done in experienced centres, many patients will experience symptomatic recurrences of AF. These patients are often managed with antiarrhythmic drugs. There is a surprising paucity of data evaluating different rhythm control interventions in patients with recurrent AF after catheter ablation. Such studies seem reasonable and feasible.

15.16 Can rhythm control therapy convey a prognostic benefit in atrial fibrillation patients?

The progress in rhythm control therapy (catheter ablation, new antiarrhythmic drugs) and observational long-term analyses suggest that rhythm control therapy may have a prognostic benefit in anticoagulated AF patients. Ongoing trials such as CABANA and EAST – AFNET 4 will provide initial answers to this important question, but more data are needed, including trials of surgical ablation techniques.

15.17 Thoracoscopic ‘stand-alone’ atrial fibrillation surgery

Minimally invasive epicardial ablation surgery for the treatment of stand-alone AF was reported a decade ago.¹⁰³⁵ The procedure has since evolved towards a totally thoracoscopic procedure,¹⁰³⁶ and lesion sets were extended to a complete left atrial maze.⁸²² Randomized trials using a standardized procedure are urgently needed to clearly define the benefits and risks of thoracoscopic AF ablation, and to further support decisions of the AF Heart Team.

15.18 Surgical exclusion of the left atrial appendage

Exclusion of the LAA has been performed by cardiothoracic surgeons for decades, but prospective randomized studies comparing the rate of ischaemic stroke with or without left appendage exclusion are presently lacking. The LAAOS (Left Atrial Appendage Occlusion Study) III is currently randomizing cardiac surgery patients with AF to undergo concomitant occlusion or no occlusion of the appendage.⁴⁶⁷ More data are also needed to confirm the safety and efficacy of thoracoscopic exclusion, following early positive observational data.¹⁰³⁷

15.19 Concomitant atrial fibrillation surgery

Adequately powered randomized trials are needed, employing systematic follow-up with uniform lesion sets and energy sources, to evaluate the benefits and risks of concomitant AF surgery in symptomatic AF patients. An RCT on non-uniform lesion sets with long-term follow-up is due to publish shortly.¹⁰³⁸ Such trials will assist the AF Heart Team to decide on optimal therapy for individual patients, including the full repertoire of medical and surgical options for the treatment of AF.

16. To do and not to do messages from the Guidelines

ACE = angiotensin-converting enzyme; AF = atrial fibrillation; AHRE = atrial high rate episodes; ARB = angiotensin receptor blocker; CHA₂DS₂-VASc = Congestive Heart failure, hypertension, Age ≥75 (doubled), Diabetes, Stroke (doubled), Vascular disease, Age 65–74, and Sex (female); ECG = electrocardiogram; EHRA = European Heart Rhythm Association; ICD = implantable cardioverter defibrillator; INR = international normalized ratio; LV = left ventricular; LVEF = left ventricular ejection fraction; LVH = left ventricular hypertrophy; NOAC = non-vitamin K antagonist oral anticoagulant; OAC = oral anticoagulation; TIA = transient ischaemic attack; TOE = transoesophageal echocardiography; TTR = time in therapeutic range; VKA = vitamin K antagonist.

17. A short summary of the management of atrial fibrillation patients

Here, we provide 17 simple rules to guide the diagnosis and management of AF patients according to the 2016 ESC Guidelines for the management of atrial fibrillation developed in cooperation with EACTS.

• Use ECG screening in at-risk populations for AF, especially stroke survivors and the elderly.

• Document AF by ECG before starting treatment.

• Evaluate all AF patients by clinical evaluation, ECG, and echocardiogram for underlying cardiovascular conditions such as hypertension, heart failure, valvular heart disease, and others.

• Provide tailored information and education to AF patients to empower them to support AF management.

• Propose lifestyle changes to all suitable AF patients to make their management more effective.

• Treat underlying cardiovascular conditions adequately, e.g. valve repair or replacement in AF patients with significant valvular heart disease, treatment of heart failure, or management of hypertension, among others.

• Use oral anticoagulation in all AF patients unless they are at low risk for stroke based on the CHA₂DS₂VASc score or have true contraindications for anticoagulant therapy.

• Anticoagulate patients with atrial flutter similar to AF. Offer isthmus ablation to symptomatic flutter patients.

• Reduce all modifiable bleeding risk factors in all AF patients on oral anticoagulation, e.g. by treating hypertension, minimizing the duration and intensity of concomitant antiplatelet and non-steroidal anti-inflammatory drug therapy, treating anaemia and eliminating causes for blood loss, maintaining stable INR values in patients on VKAs, and moderating alcohol intake.

• Check ventricular rate in all AF patients and use rate control medications to achieve lenient rate control.

• Evaluate AF-related symptoms in all AF patients using the modified EHRA symptoms scale. Whenever patients have AF-related symptoms, aim to improve symptoms by adjustment of rate control therapy and by offering antiarrhythmic drugs, cardioversion, or catheter or surgical ablation.

• Select antiarrhythmic drugs based on their safety profile and consider catheter or surgical ablation when antiarrhythmic drugs fail.

• Do not offer routine genetic testing in AF patients unless there is suspicion of an inherited cardiac condition.

• Do not use antiplatelet therapy for stroke prevention in AF.

• Do not permanently discontinue oral anticoagulation in AF patients at increased risk of stroke unless such a decision is taken by a multidisciplinary team.

• Do not use rhythm control therapy in asymptomatic AF patients, nor in patients with permanent AF.

• Do not perform cardioversion or catheter ablation without anticoagulation, unless an atrial thrombus has been ruled out transoesophageal echocardiogram.

18. Web addenda

Three additional Web figures and two additional Web tables can be accessed in the Web addenda to the 2016 ESC AF Guidelines, available at European Heart Journal online and also via the ESC Website (www.escardio.org/guidelines).

19. Appendix

ESC Committee for Practice Guidelines (CPG): Jose Luis Zamorano (Chairperson) (Spain), Victor Aboyans (France), Stephan Achenbach (Germany), Stefan Agewall (Norway), Lina Badimon (Spain), Gonzalo Barón-Esquivias (Spain), Helmut Baumgartner (Germany), Jeroen J. Bax (The Netherlands), Héctor Bueno (Spain), Scipione Carerj (Italy), Veronica Dean (France), Çetin Erol (Turkey), Donna Fitzsimons (UK), Oliver Gaemperli (Switzerland), Paulus Kirchhof (UK/Germany), Philippe Kolh (Belgium), Patrizio Lancellotti (Belgium), Gregory Y. H. Lip (UK), Petros Nihoyannopoulos (UK), Massimo F. Piepoli (Italy), Piotr Ponikowski (Poland), Marco Roffi (Switzerland), Adam Torbicki (Poland), António Vaz Carneiro (Portugal), Stephan Windecker (Switzerland).

ESC National Cardiac Societies actively involved in the review process of the 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS:

Armenia: Armenian Cardiologists Association, Hamlet G. Hayrapetyan; Austria: Austrian Society of Cardiology, Franz Xaver Roithinger; Azerbaijan: Azerbaijan Society of Cardiology, Farid Aliyev; Belarus: Belorussian Scientific Society of Cardiologists, Alexandr Chasnoits; Belgium: Belgian Society of Cardiology, Georges H. Mairesse; Bosnia and Herzegovina: Association of Cardiologists of Bosnia and Herzegovina, Daniela Loncar Matičević; Bulgaria: Bulgarian Society of Cardiology, Tchavdar Shalganov; Croatia: Croatian Cardiac Society, Boško Skorić; Cyprus: Cyprus Society of Cardiology, Loizos Antoniades; Czech Republic: Czech Society of Cardiology, Milos Taborsky; Denmark: Danish Society of Cardiology, Steen Pehrson; Egypt: Egyptian Society of Cardiology, Said Khaled; Estonia: Estonian Society of Cardiology, Priit Kampus; Finland: Finnish Cardiac Society, Antti Hedman; The Former Yugoslav Republic of Macedonia: Macedonian Society of Cardiology, Lidija Poposka; France: French Society of Cardiology, Jean-Yves Le Heuzey; Georgia: Georgian Society of Cardiology, Kakhaber Estadashvili; Germany: German Cardiac Society, Dietmar Bänsch; Hungary: Hungarian Society of Cardiology, Zoltán Csanádi; Ireland: Irish Cardiac Society, David Keane; Israel: Israel Heart Society, Roy Beinart; Italy: Italian Federation of Cardiology, Francesco Romeo; Kazakhstan: Association of Cardiologists of Kazakhstan, Kulzida Koshumbayeva; Kosovo: Kosovo Society of Cardiology, Gani Bajraktari; Kyrgyzstan: Kyrgyz Society of Cardiology, Aibek Mirrakhimov, Latvia: Latvian Society of Cardiology, Oskars Kalejs; Lebanon: Lebanese Society of Cardiology, Samer Nasr; Lithuania: Lithuanian Society of Cardiology, Germanas Marinskis; Luxembourg: Luxembourg Society of Cardiology, Carlo Dimmer; Malta: Maltese Cardiac Society, Mark Sammut; Moldova: Moldavian Society of Cardiology, Aurel Grosu; Morocco: Moroccan Society of Cardiology, Salima Abdelali; The Netherlands: Netherlands Society of Cardiology, Martin E. W. Hemels; Norway: Norwegian Society of Cardiology, Ole-Gunnar Anfinsen; Poland: Polish Cardiac Society, Beata Średniawa; Portugal: Portuguese Society of Cardiology, Pedro Adragao; Romania: Romanian Society of Cardiology, Gheorghe-Andrei Dan; Russian Federation: Russian Society of Cardiology, Evgeny N. Mikhaylov; San Marino: San Marino Society of Cardiology, Marco Zavatta; Serbia: Cardiology Society of Serbia, Tatjana Potpara; Slovakia: Slovak Society of Cardiology, Peter Hlivak Slovenia: Slovenian Society of Cardiology, Igor Zupan; Spain: Spanish Society of Cardiology, Angel Arenal; Sweden: Swedish Society of Cardiology, Frieder Braunschweig; Switzerland: Swiss Society of Cardiology, Dipen Shah; Tunisia: Tunisian Society of Cardiology and Cardio-Vascular Surgery, Ag Sana Ouali; Turkey: Turkish Society of Cardiology, Mesut Demir; Ukraine: Ukrainian Association of Cardiology, Oleg Sychov; United Kingdom: British Cardiovascular Society, Ed Duncan.

Author notes

The disclosure forms of all experts involved in the development of these guidelines are available on the ESC website http://www.escardio.org/guidelines.

Representing the European Association for Cardio-Thoracic Surgery (EACTS)

Representing the European Stroke Association (ESO)

ESC Committeefor Practice Guidelines (CPG) and National Cardiac Societies Reviewers can be found in the Appendix.

ESC entities having participated in the development of this document:

Associations: European Association for Cardiovascular Prevention and Rehabilitation (EACPR), European Association of Cardiovascular Imaging (EACVI), European Heart Rhythm Association (EHRA), Heart Failure Association (HFA).

Councils: Council on Cardiovascular Nursing and Allied Professions, Council for Cardiology Practice, Council on Cardiovascular Primary Care, Council on Hypertension.

Working Groups: Cardiac Cellular Electrophysiology, Cardiovascular Pharmacotherapy, Grown-up Congenital Heart Disease, Thrombosis, Valvular Heart Disease.

The content of these European Society of Cardiology (ESC) Guidelines has been published for personal and educational use only. No commercial use is authorized. No part of the ESC Guidelines may be translated or reproduced in any form without written permission from the ESC. Permission can be obtained upon submission of a written request to Oxford University Press, the publisher of the European Heart Journal and the party authorized to handle such permissions on behalf of the ESC ([e-mail]).

Disclaimer. The ESC Guidelines represent the views of the ESC and were produced after careful consideration of the scientific and medical knowledge and the evidence available at the time of their publication. The ESC is not responsible in the event of any contradiction, discrepancy and/or ambiguity between the ESC Guidelines and any other official recommendations or guidelines issued by the relevant public health authorities, in particular in relation to good use of healthcare or therapeutic strategies. Health professionals are encouraged to take the ESC Guidelines fully into account when exercising their clinical judgment, as well as in the determination and the implementation of preventive, diagnostic or therapeutic medical strategies; however, the ESC Guidelines do not override, in any way whatsoever, the individual responsibility of health professionals to make appropriate and accurate decisions in consideration of each patient's health condition and in consultation with that patient and, where appropriate and/or necessary, the patient's caregiver. Nor do the ESC Guidelines exempt health professionals from taking into full and careful consideration the relevant official updated recommendations or guidelines issued by the competent public health authorities, in order to manage each patient's case in light of the scientifically accepted data pursuant to their respective ethical and professional obligations. It is also the health professional's responsibility to verify the applicable rules and regulations relating to drugs and medical devices at the time of prescription.