Atrial fibrillation

Overview and definitions

Start with what the ECG shows. The 2024 ESC guideline defines AF as a supraventricular arrhythmia with uncoordinated atrial activation, which results in a loss of effective atrial contraction.^[1] On the surface ECG there are no discernible, regular P waves and the ventricles are activated irregularly, so the RR intervals follow no specific pattern in the absence of atrioventricular block.^[1] The 2023 ACC/AHA/ACCP/HRS guideline describes the atrial rhythm as chaotic, rapid (300–500 bpm) and irregular.^[2]

Classification by temporal pattern

The ESC classifies AF by temporal pattern. These categories reflect observed episodes of AF and do not suggest the underlying pathophysiological process.^[1]

^{[1] [1]}

Over time some patients with AF develop atrial and ventricular damage that can make rhythm control futile; for this reason, or when the patient and physician jointly choose rate control, AF is classified as permanent, the most common type in historical registries.^[1] The ACC/AHA proposes a classification using stages, which recognises AF as a progressive disease needing different strategies at different stages, from prevention and screening to rate and rhythm control; the stages are not mutually exclusive.^[2]

Epidemiology and risk factors

^[1]

AF is the most common sustained arrhythmia worldwide.^[1] The rise in overall prevalence is largely attributable to population growth, ageing and survival from other cardiac conditions, and age-standardised incidence rates are higher for men than women.^[1]

The ESC lists factors associated with incident AF (its Table 16); they are robustly associated with incident AF in observational studies, but it is not known whether the relationships are causal.^[1] Mendelian randomisation studies robustly implicate systolic blood pressure and higher body mass index as causal risk factors for incident AF.^[1]

• Heart failure and AF travel together: over 30 years of the Framingham cohort, 57% of those with new heart failure had concomitant AF and 37% of those with new AF had heart failure.^[1]

• Diabetes mellitus is present in around 25% of patients with AF.^[1]

• Risk factors for atrial flutter (AFL) and AF are similar, and more than half of all patients with AFL will develop AF.^[1]

• For primary prevention, ESC 2024 recommends, to prevent AF, maintaining optimal blood pressure in the general population with ACE inhibitors or ARBs as first-line therapy (Class I, level B); maintaining an active lifestyle, with the equivalent of 150–300 min per week of moderate-intensity or 75–150 min per week of vigorous-intensity aerobic physical activity (Class I, level B); and avoidance of binge drinking and alcohol excess in the general population (Class I, level B).^[1]

• The ACC/AHA recommends that patients at increased risk of AF receive comprehensive guideline-directed lifestyle and risk factor modification targeting obesity, physical inactivity, unhealthy alcohol consumption, smoking, diabetes and hypertension (COR 1, LOE B-NR).^[2]

Pathophysiology

AF often stems from waves of electrical activity from ectopic action potentials, most commonly generated in the pulmonary veins of the left atrium, or from re-entrant activity promoted by heterogeneous conduction due to interstitial fibrosis.^[2] The observation that ectopic firing from the pulmonary veins triggers AF revolutionised its treatment.^[2] Atrial ectopy can generate runs of tachycardia, whereas persistent AF needs a substrate that is large enough, or conduction heterogeneous enough, for re-entry to persist.^[2]

The atria of patients with AF tend to have shorter effective refractory periods and slower conduction, which enhance dispersion of repolarisation and favour re-entry; this substrate is sensitive to AF initiation, frequently after premature atrial contractions.^[2] Downregulation of connexin decreases gap junctions, leading to slow, heterogeneous atrial conduction velocity and repolarisation and promoting regional functional conduction block that can support re-entry.^[2]

• Atrial cardiomyopathy (ESC 2024, Table 6) is a combination of structural, electrical or functional atrial changes with clinical impact, such as progression or recurrence of AF, limited effectiveness of AF therapy or development of heart failure; it includes inflammatory and prothrombotic atrial remodelling, neurohormonal activation and myocardial fibrosis.^[1]
• The ACC/AHA adds that atrial cardiomyopathy is common, associated with ageing and with comorbidities causing metabolic or haemodynamic stress, and often leads to or results from AF; the ECG abnormalities of AF likely represent a shared phenotype of many distinct aetiologies (genetic, environmental and metabolic).^[2]

^[2]

Clinical presentation

Symptoms of AF are variable and broad, not just typical palpitations, and asymptomatic episodes can occur, although 90% of patients with AF describe symptoms of variable severity.^[1] Cardiac-specific symptoms such as palpitations are less common than non-specific symptoms such as fatigue, but they significantly impair quality of life.^[1] The available literature suggests that women with AF appear to be more symptomatic and have poorer quality of life.^[1]

• The presence or absence of symptoms is not related to incident stroke, systemic embolism or mortality.^[1]
• Symptoms may also relate to associated comorbidities and not just the AF component.^[1]
• Evaluating the impact of AF-related symptoms is recommended before and after major changes in treatment to inform shared decision-making and guide treatment choices (ESC 2024, Class I, level B).^[1]

^[1]

Differential diagnosis

The distinguishing features below come from the ACC/AHA definitions of other atrial arrhythmias and the ESC definition of AF.^[1][2]

^{[1] [2]}

Per the ESC, atrial high-rate episodes need visual inspection because some may be electrical artefacts or false positives.^[1] ECG confirmation of AF does not include non-ECG wearables and other devices that typically use photoplethysmography.^[1]

Clinical assessment and investigations

• Confirmation by an ECG (12-lead, multiple or single leads) is recommended to establish the diagnosis of clinical AF and commence risk stratification and treatment (ESC 2024, Class I, level A).^[1]

• The time period of AF needed for diagnosis on monitoring devices is not clear cut: a standard 12-lead ECG measures 10 s, while 30 s or more on single-lead or multiple-lead ECG devices has generally been the consensus opinion, albeit with limited evidence.^[1]

• The ACC/AHA recommends that, in individuals without a known history of AF, an initial AF diagnosis be made by a clinician using visual interpretation of the electrocardiographic signals, regardless of the type of rhythm or monitoring device (COR 1, LOE B-NR).^[2]

• In patients with an intracardiac rhythm device capable of diagnosing AF, such as from an atrial pacemaker lead, AF should only be diagnosed after visual confirmation of the intracardiac tracings to exclude signal artefacts and other arrhythmias (ACC/AHA 2023, COR 1, LOE B-NR).^[2]

^[1]

• The ESC states that a 12-lead ECG is warranted in all patients with AF to confirm rhythm, determine ventricular rate, and look for signs of structural heart disease, conduction defects or ischaemia.^[1]

• The blood tests are there to detect concomitant conditions that may exacerbate AF or increase the risk of bleeding and/or thromboembolism, and patients presenting with new-onset or recurrent AF should have thyroid-stimulating hormone measured (ESC 2024).^[1]

• In the diagnostic evaluation of new AF, a transthoracic echocardiogram is recommended in patients with an AF diagnosis where this will guide treatment decisions (ESC 2024, Class I, level C); the ESC task force recognises that access to TTE might be limited or delayed in primary care, but this should not delay oral anticoagulation or other components of AF-CARE where indicated.^[1]

• The ACC/AHA recommends, in newly diagnosed AF, a transthoracic echocardiogram to assess cardiac structure, laboratory testing including a complete blood count, metabolic panel and thyroid function, and, when clinical suspicion exists, targeted testing for other medical conditions associated with AF, to determine stroke and bleeding risk factors and underlying conditions that will guide further management (COR 1, LOE B-NR).^[2]

• Protocolised testing for ischaemia, acute coronary syndrome and pulmonary embolism should not routinely be performed to assess the aetiology of newly diagnosed AF unless additional signs or symptoms indicate those disorders (ACC/AHA 2023, COR 3: No benefit, LOE B-NR).^[2]

Screening

• When screening for AF, ESC 2024 recommends review of an ECG (12-lead, single or multiple leads) by a physician to provide a definite diagnosis of AF and commence appropriate management (Class I, level B).^[1]
• Routine heart rhythm assessment during healthcare contact is recommended in all individuals aged 65 years or more for earlier detection of AF (ESC 2024, Class I, level C).^[1]
• Population-based screening using a prolonged non-invasive ECG-based approach should be considered in individuals aged 75 years or more, or 65 years or more with additional CHA2DS2-VA risk factors, to ensure earlier detection of AF (ESC 2024, Class IIa, level B).^[1]
• Screening devices broadly divide into those that provide an ECG and those with non-ECG approaches such as photoplethysmography.^[1]

Management — the acute and unstable patient

• Unstable patients with AF include those with haemodynamic instability caused by the arrhythmia or acute cardiac conditions, and severely ill patients who develop AF (sepsis, trauma, surgery, and particularly cancer-related surgery).^[1]
• Electrical cardioversion is recommended in patients with AF and acute or worsening haemodynamic instability to improve immediate outcomes (ESC 2024, Class I, level C); the ACC/AHA recommends immediate electrical cardioversion to restore sinus rhythm in haemodynamic instability attributable to AF (COR 1, LOE C-LD).^[1][2]
• In the ESC text, emergency electrical cardioversion is still considered the first-choice treatment if sinus rhythm is thought to be beneficial, despite a high rate of immediate relapse; amiodarone is second-line because of its delayed activity but may be an appropriate alternative in the acute setting.^[1]

The ESC says physicians should always evaluate and manage the underlying causes of AF before or in parallel with acute rate or rhythm control, such as treating sepsis, addressing fluid overload or managing cardiogenic shock.^[1] Spontaneous restoration of sinus rhythm has been reported in up to 83% during the first 48 h after appropriate treatment of the underlying cause.^[1]

• The ESC states that, in general, for acute rate control, beta-blockers (for all LVEF) and diltiazem or verapamil (for LVEF above 40%) are preferred over digoxin because of their more rapid onset and dose-dependent effects; combining beta-blockers with diltiazem or verapamil should be avoided except in closely monitored situations.^[1]

• ESC 2024 states that intravenous amiodarone, digoxin or landiolol may be considered in patients with AF who have haemodynamic instability or severely depressed LVEF to achieve acute control of heart rate (Class IIb, level B; esmolol was removed from this row by the 2025 correction).^[1][3]

• Landiolol can safely control rapid AF in patients with low ejection fraction and acutely decompensated heart failure, with limited impact on myocardial contractility or blood pressure.^[1]

• The ACC/AHA recommends beta-blockers or non-dihydropyridine calcium channel blockers (verapamil, diltiazem; provided that EF is above 40%) for acute rate control in haemodynamically stable AF with rapid ventricular response (COR 1, LOE B-R).^[2]

• In AF with rapid ventricular response and known moderate or severe LV systolic dysfunction, with or without decompensated heart failure, intravenous non-dihydropyridine calcium channel blockers should not be administered (ACC/AHA 2023, COR 3: Harm, LOE B-NR).^[2]

• In AF with rapid ventricular response in patients who are critically ill and/or in decompensated heart failure, in whom beta-blockers and non-dihydropyridine calcium channel blockers are ineffective or contraindicated, intravenous amiodarone may be considered for acute rate control (ACC/AHA 2023, COR 2b, LOE B-NR), considering the risk of cardioversion and stroke.^[2]

• In AF with rapid ventricular response where beta-blockers and non-dihydropyridine calcium channel blockers are ineffective or contraindicated, the ACC/AHA says digoxin can be considered for acute rate control, alone or in combination (COR 2a, LOE B-R), and, in AF with rapid ventricular response, adding intravenous magnesium to standard rate-control measures is reasonable to achieve and maintain rate control (COR 2a, LOE A).^[2]

• In stable patients, a wait-and-see approach for spontaneous conversion to sinus rhythm within 48 h of AF onset should be considered as an alternative to immediate cardioversion in patients without haemodynamic compromise (ESC 2024, Class IIa, level B).^[1]
• The ESC reports that in RACE 7 ACWAS (recent-onset symptomatic AF without haemodynamic compromise), a wait-and-see approach for spontaneous conversion until 48 h after symptom onset was non-inferior to immediate cardioversion at 4 weeks follow-up; the ESC text does not name the endpoint.^[1]

Management — the AF-CARE pathway

AF-CARE builds on the 2016 ESC five-step approach and the ABC pathway of the 2020 ESC guideline, and joint management with each patient is its starting point.^[1] For the ESC, a careful search for comorbidities and risk factors [C] is critical and applies to all patients with AF; avoiding stroke and thromboembolism [A] in patients with risk factors comes next, focused on appropriate anticoagulation.^[1] Reducing symptoms and morbidity by rate and rhythm control [R] follows, and in selected patients may also reduce hospitalisation or improve prognosis; evaluation [E] is dynamic because AF and its comorbidities change over time.^[1]

• ESC 2024 recommends access to patient-centred management according to the AF-CARE principles in all patients with AF, regardless of gender, ethnicity and socioeconomic status, to ensure equality in healthcare provision and improve outcomes (Class I, level C); education for patients, family members, caregivers and healthcare professionals to optimise shared decision-making, facilitating open discussion of the benefits and risks of each treatment option (Class I, level C); and patient-centred management with a multidisciplinary approach should be considered in all patients with AF to optimise management and improve outcomes (Class IIa, level B).^[1]
• The ACC/AHA recommends that patients with AF receive comprehensive care addressing guideline-directed lifestyle and risk factor modification, AF symptoms, stroke risk and other associated medical conditions to reduce AF burden, progression or consequences (COR 1, LOE A).^[2]

^[1]

[C] Comorbidity and risk factor management

These are the full rows of ESC Recommendation Table 5.^[1]

^[1]

• The ESC target for treated systolic blood pressure in most adults is 120–129 mmHg.^[1]

• In an RCT among regular non-binge drinkers with AF, a goal of abstinence significantly reduced AF recurrence and burden, with intake in the intervention arm falling from 16.8 to 2.1 standard drinks per week and 61% attaining abstinence.^[1]

• For sleep apnoea, the ESC text suggests polysomnography or home sleep apnoea testing in preference to screening questionnaires.^[1]

• The ACC/AHA finds it may be reasonable to screen patients with AF for obstructive sleep apnoea, given its high prevalence, although the role of treating sleep-disordered breathing to maintain sinus rhythm is uncertain (COR 2b, LOE B-NR).^[2]

• The ACC/AHA has its own rows. In AF with overweight or obesity (BMI above 27 kg/m2), weight loss is recommended with an ideal target of at least 10% to reduce AF symptoms, burden, recurrence and progression to persistent AF (COR 1, LOE B-R).^[2]

• Moderate-to-vigorous exercise training to a target of 210 minutes per week is recommended in individuals with AF, except those whose AF is related to excessive exercise training, to reduce AF symptoms and burden, increase maintenance of sinus rhythm, increase functional capacity and improve quality of life (ACC/AHA 2023, COR 1, LOE B-R).^[2]

• Patients with AF seeking a rhythm-control strategy should minimise or eliminate alcohol to reduce AF recurrence and burden (ACC/AHA 2023, COR 1, LOE B-R).^[2]

• For AF with hypertension, the ACC/AHA recommends optimal blood pressure control to reduce AF recurrence and AF-related cardiovascular events (COR 1, LOE B-NR).^[2]

[A] Avoid stroke and thromboembolism

AF is a major risk factor for thromboembolism whether it is paroxysmal, persistent or permanent, so the ESC default is to provide oral anticoagulation (OAC) to all eligible patients except those at low risk of incident stroke or thromboembolism.^[1] Most available stroke risk scores are simple and practical, but their predictive value is generally modest.^[1]

These are the full rows of ESC Recommendation Table 6.^[1]

^[1]

CHA2DS2-VA: why the sex criterion went

• CHA2DS2-VASc gives points for congestive heart failure, hypertension, age 75 years or more (2 points), diabetes, prior stroke, TIA or thromboembolism (2 points), vascular disease, age 65–74 years and female sex.^[1]
• The ESC regards female sex as an age-dependent stroke risk modifier rather than a risk factor per se, notes that a sex criterion omits people who identify as non-binary or transgender or who are undergoing sex hormone therapy, and proposes CHA2DS2-VA, without a criterion for birth sex or gender, in the absence of other locally validated alternatives.^[1]

^[1]

• ESC Table 10 footnote: other markers that modify an individual’s risk should also be considered, including cancer, chronic kidney disease, ethnicity (black, Hispanic, Asian), biomarkers (troponin and BNP) and, in specific groups, atrial enlargement, hyperlipidaemia, smoking and obesity.^[1]
• The ACC/AHA says patients with AF at intermediate annual thromboembolic risk by risk scores (e.g. equivalent to CHA2DS2-VASc 1 in men or 2 in women) who remain uncertain about the benefit of anticoagulation can benefit from consideration of factors that might modify their stroke risk, to help inform the decision (COR 2a, LOE C-LD); its footnote says such factors may include AF burden or other features in its Table 3.^[2]
• Most available risk scores have a threshold of 0.6%–1.0% per year of thromboembolic events for clinical AF to warrant OAC.^[1]

^{[1] [2] [1] [2]}

Choosing the anticoagulant

• The DOACs (apixaban, dabigatran, edoxaban and rivaroxaban) have all shown at least non-inferior efficacy to warfarin for preventing thromboembolism, with a 50% reduction in intracranial haemorrhage.^[1]
• In meta-analyses of individual data from 71 683 RCT patients, standard full-dose DOACs compared with warfarin reduced stroke or systemic embolism (HR 0.81), all-cause mortality (HR 0.90) and intracranial bleeding (HR 0.48), with no significant difference in other major bleeding (HR 0.86).^[1]

^[1]

• The ACC/AHA recommends DOACs over warfarin in patients with AF who are candidates for anticoagulation and do not have a history of moderate to severe rheumatic mitral stenosis or a mechanical heart valve, to reduce mortality, stroke, systemic embolism and intracranial haemorrhage (COR 1, LOE A).^[2]

• In rheumatic mitral stenosis, or mitral stenosis of moderate or greater severity, with a history of AF, long-term warfarin is recommended over DOACs independent of the CHA2DS2-VASc score to prevent cardiovascular events, including stroke or death (ACC/AHA 2023, COR 1, LOE B-R).^[2]

• In patients with AF, the ACC/AHA classes non-evidence-based DOAC doses as harmful: they should be avoided to minimise risks of preventable thromboembolism or major bleeding and to improve survival (COR 3: Harm, LOE B-NR).^[2]

• The ESC notes that its DOAC restriction does not apply to bioprosthetic heart valves (including mitral) or to patients after transcatheter aortic valve implantation, that patients with other valve disease, such as mitral regurgitation, should preferentially receive a DOAC, and that the term ‘valvular’ AF is obsolete.^[1]

• The ACC/AHA likewise recommends DOACs over VKAs in patients with AF and valve disease other than moderate or greater mitral stenosis or a mechanical heart valve (COR 1, LOE B-NR).^[2]

• The ESC notes that dabigatran is more dependent on renal elimination and is contraindicated with an eGFR below 30 mL/min/1.73 m2.^[1]

^{[1] [3]}

Table 11 footnote: dose and dose adjustments are taken from the European Medicines Association Summary of Product Characteristics for each DOAC.^[1] There may be other patient-specific reasons for providing a reduced dose, but, in general, the standard full dose should be used to provide optimal prevention of thromboembolism related to AF.^[1] The footnote adds that antiplatelet agents should be stopped in most patients when commencing a DOAC, and that a number of drug interactions exist with each DOAC and should be taken into consideration.^[1] The 2025 correction changed the apixaban creatinine unit in this table from mmol/L to µmol/L.^[3] Inappropriate DOAC dose reductions are frequent in practice but increase the risk of stroke without decreasing bleeding risk.^[1]

• Vitamin K antagonist therapy reduces stroke risk by 64% and mortality by 26% in patients with AF at elevated thromboembolic risk (mostly warfarin in trials, compared with placebo or no treatment), and, according to the 2024 ESC guideline, VKAs are currently the only option in AF with mechanical heart valves or moderate-to-severe mitral stenosis.^[1]

Antiplatelets and stroke despite anticoagulation

• Antiplatelet drugs such as aspirin and clopidogrel are not an alternative to OAC: antiplatelet therapy is not recommended as an alternative to anticoagulation in patients with AF to prevent ischaemic stroke and thromboembolism (ESC 2024, Class III, level A), and adding antiplatelet treatment to OAC is not recommended in AF patients for the goal of preventing ischaemic stroke or thromboembolism (Class III, level B).^[1]

• The ACC/AHA classes aspirin alone or with clopidogrel as an alternative to anticoagulation, in patients with AF who are candidates for anticoagulation and without an indication for antiplatelet therapy, to reduce stroke risk as COR 3: Harm (LOE B-R).^[2]

• One-third of patients with AF presenting with an ischaemic stroke are already on anticoagulation.^[1]

• For them, ESC 2024 says a thorough diagnostic work-up should be considered in patients taking OAC who present with ischaemic stroke or thromboembolism, to prevent recurrent events, including assessment of non-cardioembolic causes, vascular risk factors, dosage and adherence (Class IIa, level B).^[1]

• After thromboembolism despite anticoagulation, adding antiplatelet treatment to anticoagulation (ESC 2024, Class III, level B), and switching from one DOAC to another or from a DOAC to a VKA without a clear indication (Class III, level B), are not recommended in patients with AF to prevent recurrent embolic stroke.^[1]

Left atrial appendage occlusion

• Percutaneous LAA occlusion may be considered in patients with AF and contraindications for long-term anticoagulant treatment to prevent ischaemic stroke and thromboembolism (ESC 2024, Class IIb, level C).^[1]

• Pending further RCTs, the ESC sees the most appropriate rationale in patients with a contraindication to all of the OAC options (the four DOACs and VKAs); implantation carries procedural risks including stroke, major bleeding, device-related thrombus, pericardial effusion, vascular complications and death, and device-related thrombi occur in 1.7%–7.2%.^[1]

• The ACC/AHA rates percutaneous LAAO as reasonable in AF with a moderate to high stroke risk (CHA2DS2-VASc score of 2 or more) and a contraindication to long-term OAC due to a non-reversible cause (COR 2a, LOE B-NR).^[2]

• In AF with a moderate to high stroke risk and a high risk of major bleeding on OAC, percutaneous LAAO may be a reasonable alternative to OAC based on patient preference, with careful consideration of procedural risk and the understanding that the evidence for OAC is more extensive (ACC/AHA 2023, COR 2b, LOE B-R).^[2]

• Surgical closure of the left atrial appendage is recommended as an adjunct to OAC in patients with AF undergoing cardiac surgery (ESC 2024, Class I, level B), should be considered as an adjunct to OAC during endoscopic or hybrid AF ablation (Class IIa, level C), and stand-alone endoscopic closure may be considered in patients with AF and contraindications for long-term anticoagulant treatment (Class IIb, level C), in each case to prevent ischaemic stroke and thromboembolism.^[1]

• In LAAOS III (4811 patients with AF having cardiac surgery for another indication), ischaemic stroke or systemic embolism occurred in 4.8% with occlusion vs 7.0% without (HR 0.67; 95% CI 0.53–0.85; P = 0.001) over a mean 3.8 years; the trial did not compare occlusion with anticoagulation, as 77% continued OAC.^[1]

• The ACC/AHA recommends surgical LAA exclusion, in addition to continued anticoagulation, in AF patients undergoing cardiac surgery with a CHA2DS2-VASc score of 2 or more or equivalent stroke risk, to reduce the risk of stroke and systemic embolism (COR 1, LOE A).^[2]

• In AF patients undergoing cardiac surgery with a CHA2DS2-VASc score of 2 or more or equivalent stroke risk, the ACC/AHA considers the benefit of surgical LAA exclusion without continued anticoagulation, to reduce stroke and systemic embolism, uncertain (COR 2b, LOE A).^[2]

Bleeding risk and bleeding on anticoagulation

• Assessment and management of modifiable bleeding risk factors is recommended in all patients eligible for OAC, as part of shared decision-making, to ensure safety and prevent bleeding (ESC 2024, Class I, level B); this includes strict hypertension control, reducing excess alcohol, avoiding unnecessary antiplatelet or anti-inflammatory drugs, and attention to OAC adherence, TTR on VKAs and interacting medications.^[1]

• Using bleeding risk scores to decide on starting or withdrawing OAC is not recommended in patients with AF, to avoid under-use of anticoagulation (ESC 2024, Class III, level B), and the ESC recommends no specific bleeding score.^[1]

• The ACC/AHA says that in patients at high stroke risk, bleeding risk scores should not be used in isolation to determine eligibility for OAC, but to identify and modify bleeding risk factors and inform decision-making (COR 3: No Benefit, LOE B-NR).^[2]

• The ESC states that bleeding risk factors are rarely a reason to withhold or withdraw OAC in eligible patients, as the stroke risk without anticoagulation often outweighs the risk of major bleeding; absolute contraindications are rare and include primary intracranial tumours and intracerebral bleeds related to amyloid angiopathy.^[1]

• In AF patients with active bleeding, interrupting anticoagulation and performing diagnostic or treatment interventions is recommended until the cause is identified and resolved (ESC 2024, Class I, level C).^[1]

• In AF patients with a life-threatening bleed or a bleed into a critical site, prothrombin complex concentrates should be considered on VKAs (ESC 2024, Class IIa, level C) and specific antidotes should be considered on a DOAC (Class IIa, level B), to reverse the antithrombotic effect.^[1]

• Idarucizumab can fully reverse dabigatran and help achieve haemostasis within 2–4 h in uncontrolled bleeding, and andexanet alfa rapidly reverses the factor Xa inhibitors (apixaban, edoxaban, rivaroxaban).^[1]

• For patients with AF who develop life-threatening bleeding, the ACC/AHA recommends idarucizumab on dabigatran to rapidly reverse its anticoagulant effect (COR 1, LOE B-NR); on a factor Xa inhibitor, either andexanet alfa (apixaban or rivaroxaban; edoxaban) or 4-factor prothrombin complex concentrate to rapidly reverse its effect (COR 1; the row marks andexanet alfa for apixaban or rivaroxaban with LOE B-NR, and andexanet alfa for edoxaban and 4-factor prothrombin complex concentrate with the C-LD footnote, which reads "C-LD LOE applies to data on edoxaban"); and, on warfarin, 4-factor prothrombin complex concentrate (if available) with intravenous vitamin K, over fresh frozen plasma and intravenous vitamin K, to rapidly achieve INR correction (COR 1, LOE A).^[2]

[R] Rate control

The ESC notes most patients need a combination of rate and rhythm control, consciously re-evaluated during follow-up.^[1] These are the full rows of ESC Recommendation Table 14, with the 2025 correction applied to the last row.^[1][3]

^{[1] [3]}

• In RACE II (permanent AF), lenient rate control (target heart rate under 110 b.p.m.) was non-inferior to a strict approach (under 80 b.p.m. at rest and under 110 b.p.m. during exercise) for a composite of clinical events, NYHA class or hospitalisation.^[1]
• The ESC concludes that lenient control is an acceptable initial approach unless there are ongoing symptoms or suspicion of tachycardia-induced cardiomyopathy, where stricter targets may be indicated.^[1]
• The ACC/AHA says that in suspected AF-induced cardiomyopathy or refractory heart failure symptoms on pharmacological rate control, a stricter rate-control strategy (target heart rate under 80 bpm at rest and under 110 bpm during moderate exercise) may be reasonable (COR 2b, LOE B-NR).^[2]
• The ACC/AHA states that in patients with AF without heart failure who are candidates for select rate-control strategies, the heart rate target should be guided by underlying symptoms, in general aiming at a resting heart rate under 100 to 110 bpm (COR 2a, LOE B-R).^[2]

^{[1] [3]}

In ESC Table 12 the Contraindicated column has one cell for each drug group, not one for each drug, so it is shown here once per group.^[1]

^[1]

• In ESC Table 12, N/A means not available or not widely available, and maximum doses are based on the summary of product characteristics of each drug.^[1]

• Footnotes to the ESC table: all rate control drugs are contraindicated in Wolff–Parkinson–White syndrome (also IV amiodarone); other beta-blockers are available but not recommended as specific rate control therapy in AF (e.g. propranolol and labetalol); atenolol has no data and should not be used in HFrEF or in pregnancy; and the amiodarone loading regimen may vary, and IV dosage should be considered when calculating the total load.^[1]

• The ESC advises that combining beta-blockers with verapamil or diltiazem should only be done in secondary care with regular monitoring of heart rate by 24 h ECG to check for bradycardia, and dronedarone should not be started for rate control because it increases heart failure, stroke and cardiovascular death in permanent AF.^[1]

• The ACC/AHA also classes dronedarone for long-term rate control in permanent AF with cardiovascular risk factors, and, in the long-term rate control table, non-dihydropyridine calcium channel blockers in AF with LVEF below 40%, given their potential to exacerbate heart failure, as COR 3: Harm (LOE B-R and C-LD respectively).^[2]

• For long-term rate control, the ACC/AHA recommends beta-blockers or non-dihydropyridine calcium channel blockers (diltiazem, verapamil), choosing the agent by underlying substrate and comorbid conditions (COR 1, LOE B-NR), and finds digoxin reasonable in AF with heart failure symptoms, combined with other rate-controlling agents or as monotherapy if other agents are not preferred, not tolerated or contraindicated (COR 2a, LOE B-R).^[2]

• In RATE-AF (symptomatic permanent AF), low-dose digoxin and bisoprolol did not differ in patient-reported quality of life at 6 months, but digoxin produced fewer adverse effects, greater improvement in mEHRA and NYHA scores and a reduction in BNP.^[1]

• The ESC reserves amiodarone as a last option when heart rate cannot be controlled even with maximal tolerated combination therapy, or for patients who do not qualify for atrioventricular node ablation and pacing.^[1]

• For ablate and pace, the ESC text advises implanting the pacemaker a few weeks before atrioventricular node ablation, with an initial pacing rate after ablation of 70–90 b.p.m.^[1]

• The ACC/AHA differs: in patients with AF planned to undergo AVNA, implantation of a pacemaker before the ablation (before or on the same day) is recommended to ensure adequacy of the pacing leads (COR 1, LOE B-NR), and with a persistently rapid ventricular response, initial pacemaker lower rate programming should be 80 to 90 bpm to reduce the risk of sudden death (COR 1, LOE C-LD).^[2]

• In APAF-CRT (narrow QRS complexes), atrioventricular node ablation with CRT was superior to rate control drugs for the primary outcomes (all-cause mortality, and death or heart failure hospitalisation) and for the secondary outcomes (symptom burden and physical limitation).^[1]

[R] Rhythm control

• Rhythm control refers to therapies that restore and maintain sinus rhythm, and the ESC stresses it is never a strategy on its own but always part of AF-CARE.^[1]

• Older studies found that a rhythm control strategy using antiarrhythmic drugs (AADs) did not reduce mortality and morbidity compared with rate control alone and may increase hospitalisation, whereas multiple studies have shown that rhythm control strategies have a positive effect on quality of life once sinus rhythm is maintained.^[1]

• In symptomatic patients, the ESC says patient factors that favour an attempt at rhythm control should be considered, including suspected tachycardiomyopathy, a brief AF history, a non-dilated left atrium or patient preference.^[1]

• Implementation of a rhythm control strategy should be considered within 12 months of diagnosis in selected patients with AF at risk of thromboembolic events to reduce the risk of cardiovascular death or hospitalisation (ESC 2024, Class IIa, level B).^[1]

• The ACC/AHA states that in recently diagnosed AF (under 1 year), rhythm control can be useful to reduce hospitalisations, stroke and mortality (COR 2a, LOE B-R), and that in reduced LV function with persistent (or high-burden) AF a trial of rhythm control should be recommended to evaluate whether AF is contributing (COR 1, LOE B-R).^[2]

^{[1] [5]}

• The ESC notes that in EAST-AFNET 4, rhythm control was pursued predominantly with antiarrhythmic drugs (80% of patients in the intervention arm).^[1]

• The ESC notes that CASTLE-AF (catheter ablation vs standard conventional treatment) demonstrated that a rhythm control strategy with catheter ablation can improve mortality and morbidity in selected patients with HFrEF and an implanted cardiac device, whereas CABANA could not demonstrate a significant difference in mortality and morbidity between catheter ablation and standard rhythm and/or rate control drugs in symptomatic AF patients older than 64 years, or younger than 65 years with risk factors for stroke.^[1]

Cardioversion and anticoagulation

Any rhythm control procedure has an inherent risk of thromboembolism; the core ESC rows follow.^[1]

^[1]

• The ESC adds that only patients without thromboembolic risk factors and with sinus rhythm restored within 24 h of AF onset have optional post-cardioversion OAC, and that if thrombus is detected, therapeutic anticoagulation should be given for at least 4 weeks followed by repeat TOE.^[1]

• The ACC/AHA uses 48 hours: with AF duration of 48 hours or more, 3 weeks of uninterrupted therapeutic anticoagulation or imaging to exclude intracardiac thrombus is recommended before elective cardioversion (COR 1, LOE B-R).^[2]

• In patients with AF undergoing cardioversion, the ACC/AHA says therapeutic anticoagulation should be established before cardioversion and continued for at least 4 weeks afterwards without interruption to prevent thromboembolism (COR 1, LOE B-NR).^[2]

• For AF of under 48 hours (not in the setting of cardiac surgery) without anticoagulation, the ACC/AHA says pre-cardioversion imaging to exclude intracardiac thrombus may be considered at elevated thromboembolic risk (CHA2DS2-VASc 2 or more or equivalent) (COR 2b, LOE C-LD); with low thromboembolic risk (CHA2DS2-VASc 0–1 or equivalent) and AF under 12 hours, the benefit of pre-cardioversion imaging or peri-cardioversion anticoagulation is uncertain given the low incidence of peri-cardioversion thromboembolic events (COR 2b, LOE C-LD).^[2]

• The ESC notes that electrical cardioversion can be safely applied in elective and acute settings with sedation by intravenous midazolam, propofol or etomidate; biphasic defibrillators are standard because of their superior efficacy compared with monophasic defibrillators, maximum fixed-energy shocks were more effective than low escalating energy in a randomised trial, and intravenous atropine or isoproterenol, or temporary transcutaneous pacing, should be available in case of post-cardioversion bradycardia.^[1]

• For patients with AF undergoing elective electrical cardioversion, the ACC/AHA says biphasic energy of at least 200 J as initial energy can be beneficial to improve success of the initial shock (COR 2a, LOE B-R).^[2]

• Electrical cardioversion as a diagnostic tool should be considered in persistent AF when the value of sinus rhythm for symptoms is uncertain, or to assess improvement in LV function (ESC 2024, Class IIa, level C).^[1]

Pharmacological cardioversion

• Pharmacological cardioversion is an elective procedure in haemodynamically stable patients and is less effective than electrical cardioversion; the limited contemporary efficacy data are likely biased by spontaneous restoration of sinus rhythm in 76%–83% of patients with recent-onset AF (69% within 48 h).^[1]

^{[1] [1]}

The ESC states that safe pill-in-the-pocket use needs screening to exclude sinus node dysfunction, atrioventricular conduction defects or Brugada syndrome, plus prior in-hospital validation of efficacy and safety.^[1]

• The ACC/AHA rows differ: for patients with AF, pharmacological cardioversion is reasonable as an alternative to electrical cardioversion in those who are haemodynamically stable, or when electrical cardioversion is preferred but cannot be performed (COR 2a, LOE C-LD); ibutilide is reasonable for pharmacological cardioversion in patients without depressed LV function (LVEF under 40%) (COR 2a, LOE A); and intravenous amiodarone is reasonable for pharmacological cardioversion, although time to conversion is generally longer than with other agents (8–12 hours) (COR 2a, LOE A).^[2]
• For recurrent AF outside hospital, the ACC/AHA finds the pill-in-the-pocket approach (a single oral dose of flecainide or propafenone with a concomitant atrioventricular nodal blocking agent) reasonable if previously tested in a monitored setting (COR 2a, LOE A); for patients with AF generally, intravenous procainamide may be considered for pharmacological cardioversion when other intravenous agents are contraindicated or not preferred (COR 2b, LOE B-R).^[2]

Antiarrhythmic drugs for long-term rhythm control

^[1]

• In a meta-analysis of 59 RCTs, AADs reduced AF recurrences by 20%–50% compared with no treatment, placebo or rate control drugs; they do not eliminate recurrence, but in paroxysmal or persistent AF a recurrence is not equivalent to treatment failure if episodes are less frequent, briefer or less symptomatic.^[1]
• The ESC says drug safety, rather than efficacy, should determine the choice; long-term oral dosing is 400 mg twice daily for dronedarone and 80–160 mg twice daily for sotalol.^[1]
• The ACC/AHA rows for long-term maintenance of sinus rhythm differ: dofetilide or amiodarone is reasonable in AF with HFrEF (LVEF 40% or less) (COR 2a; LOE A for dofetilide, B-NR for amiodarone); flecainide or propafenone is reasonable without previous MI, known or suspected significant structural heart disease, or ventricular scar or fibrosis (COR 2a, LOE A); and dronedarone is reasonable without recent decompensated heart failure or severe LV dysfunction (COR 2a, LOE A).^[2]
• The ACC/AHA also finds low-dose amiodarone (100–200 mg/day) reasonable in AF with normal LV function but, in view of its adverse effect profile, reserves it for patients in whom other rhythm-control strategies are ineffective, not preferred or contraindicated (COR 2a, LOE A); sotalol may be considered without significant baseline QT prolongation, hypokalaemia, hypomagnesaemia or bradycardia, with dose selection based on kidney function and close monitoring of the QT interval, heart rate, serum potassium and magnesium, and kidney function (COR 2b, LOE A).^[2]
• The ACC/AHA classes flecainide and propafenone in patients (including those with AF) with previous MI and/or significant structural heart disease, including HFrEF (LVEF 40% or less), as harmful because of the risk of worsening heart failure, potential proarrhythmia and increased mortality, and dronedarone for maintenance of sinus rhythm in AF with NYHA class III or IV heart failure or decompensated heart failure in the past 4 weeks as harmful because of increased early mortality with worsening heart failure (both COR 3: Harm, LOE B-R).^[2]

Catheter ablation

^[1]

• Pulmonary vein isolation (PVI) remains the cornerstone of AF catheter ablation; whether first-line ablation is superior to drug therapy in persistent AF is not clear.^[1]

• Registries and trials report peri-procedural serious adverse event rates of 2.9%–7.2%, with a very low 30-day mortality (under 0.1%).^[1]

• Because CABANA did not confirm a benefit of ablation over medical therapy (high crossover and low event rates may have diluted the effect), the ESC considers that only highly selected asymptomatic patients could be candidates for catheter ablation, after detailed discussion of risks and the potential benefit of delaying AF progression.^[1]

• The ACC/AHA rates ablation as useful (COR 1, LOE A) in symptomatic AF when antiarrhythmic drugs have been ineffective, contraindicated, not tolerated or not preferred and continued rhythm control is desired, and as first-line therapy in selected patients (generally younger with few comorbidities) with symptomatic paroxysmal AF in whom rhythm control is desired; the stated aims are to improve symptoms and, for first-line use, to reduce progression to persistent AF.^[2]

• In other patients with symptomatic paroxysmal or persistent AF managed with rhythm control, first-line ablation can be useful to improve symptoms (ACC/AHA 2023, COR 2a, LOE B-R), and in appropriate patients with AF and HFrEF on GDMT with a reasonable expectation of procedural benefit, ablation is beneficial to improve symptoms, quality of life, ventricular function and cardiovascular outcomes (COR 1, LOE A).^[2]

^[1]

• The ESC notes that left atrial thrombus is a contraindication to catheter ablation because of the risk of dislodgement causing ischaemic stroke.^[1]
• The ACC/AHA recommends continuing OAC for at least 3 months after ablation, with longer duration determined by underlying risk, and dictating longer-term OAC by stroke risk (e.g. CHA2DS2-VASc score of 2 or more) (both COR 1, LOE B-NR).^[2]
• In patients on a DOAC, the ACC/AHA recommends ablation with either continuous or minimally interrupted oral anticoagulation (COR 1, LOE A), whereas ESC 2024 recommends uninterrupted OAC to prevent peri-procedural ischaemic stroke and thromboembolism (Class I, level A).^[1][2]
• For recurrent symptomatic AF after catheter ablation, the ACC/AHA finds repeat catheter ablation or antiarrhythmic drug therapy useful to improve symptoms and freedom from AF (COR 1, LOE B-NR).^[2]

Surgical and hybrid ablation

• Concomitant surgical ablation is recommended in patients undergoing mitral valve surgery with AF suitable for a rhythm control strategy (ESC 2024, Class I, level A) and should be considered in patients undergoing non-mitral valve cardiac surgery with AF suitable for a rhythm control strategy (Class IIa, level B), both to prevent symptoms and recurrence of AF, with shared decision-making supported by an experienced team of electrophysiologists and arrhythmia surgeons; intraprocedural imaging for left atrial thrombus is recommended during surgical ablation to guide surgical strategy, independent of OAC use, to prevent peri-procedural ischaemic stroke and thromboembolism (Class I, level C).^[1]
• Endoscopic and hybrid ablation should be considered in symptomatic persistent AF refractory to AAD therapy (ESC 2024, Class IIa, level A) and may be considered in symptomatic paroxysmal AF refractory to AADs after a failed percutaneous catheter ablation strategy (Class IIb, level B), both to prevent symptoms, recurrence and progression of AF within a shared decision-making rhythm control team of electrophysiologists and surgeons; continuation of OAC is recommended in patients with AF at elevated thromboembolic risk after concomitant, endoscopic or hybrid ablation, independent of rhythm outcome or LAA exclusion, to prevent ischaemic stroke and thromboembolism (Class I, level C).^[1]
• The best validated surgical method is the Maze procedure.^[1]
• The ACC/AHA rows differ in strength: concomitant surgical ablation in patients with AF undergoing cardiac surgery can be beneficial to reduce the risk of recurrent AF (COR 2a, LOE B-R); anticoagulation for at least 3 months after surgical ablation is reasonable to reduce the risk of stroke or systemic embolism (COR 2a, LOE B-NR); and hybrid epicardial and endocardial ablation in symptomatic persistent AF refractory to antiarrhythmic drugs might be reasonable to reduce the risk of recurrent atrial arrhythmia (COR 2b, LOE B-R).^[2]

[E] Evaluation and dynamic reassessment

• In most cases the ESC advises re-evaluation 6 months after initial presentation and then at least annually by a healthcare professional in primary or secondary care.^[1]
• Individualised reassessment of thromboembolic risk is recommended at periodic intervals in patients with AF to ensure anticoagulation is started in appropriate patients (ESC 2024, Class I, level B).^[1]

Specific scenarios

AF with acute coronary syndrome or PCI

• The incidence of AF in acute coronary syndromes ranges from 2% to 23%, and ESC 2024 calls AF a common precipitant of type 2 MI, a term from the numbered classification of the Fourth Universal Definition of MI (UDMI) that is now dated.^[1][8]
• The current Fifth UDMI (2026) has replaced the numerical classification: secondary MI results from myocardial oxygen supply–demand imbalance due to another acute condition.^[8]
• Under the Fifth UDMI, secondary MI is considered when acute myocardial injury from such a supply–demand mismatch causes symptoms or signs of myocardial ischaemia; where coronary and/or cardiac imaging is feasible and appropriate, it is confirmed if the ischaemia results from obstructive coronary artery disease without evidence of an acute coronary pathology, or the resulting myocardial injury is sufficient to cause a new regional wall motion abnormality or loss of myocardial viability.^[8]
• When imaging is not appropriate, as in some patients with comorbidities, advanced frailty or limited life expectancy due to the underlying acute condition, the Fifth UDMI says it may not be possible to definitively confirm secondary MI and clinical judgment is required.^[8]
• In the ESC text, acute coronary syndromes treated by PCI require DAPT, so a peri-procedural triple regimen of OAC, aspirin and a P2Y12 inhibitor should be the default strategy for most patients; OAC (preferably a DOAC) with a P2Y12 inhibitor causes less major bleeding than triple therapy with aspirin, and clopidogrel is the preferred P2Y12 inhibitor.^[1]

^[1]

• The ESC text recommends short-term triple therapy (1 week or less) for all patients without diabetes after ACS or PCI; in AF patients with ACS or CCS and diabetes undergoing coronary stent implantation, prolonging triple therapy with low-dose aspirin, clopidogrel and an OAC up to 3 months may be of benefit if thrombotic risk outweighs bleeding risk in the individual patient.^[1]
• The ESC text states that in patients with stable CCS for more than 12 months, a DOAC alone is sufficient.^[1]
• The ACC/AHA recommends, in AF with chronic coronary disease (beyond 1 year after revascularisation, or CAD not requiring revascularisation) without a history of stent thrombosis, OAC monotherapy over OAC plus a single antiplatelet agent (aspirin or a P2Y12 inhibitor) to decrease the risk of major bleeding (COR 1, LOE B-R).^[2]
• The ACC/AHA prefers DOACs over VKAs with antiplatelet therapy in AF at increased stroke risk undergoing PCI (COR 1, LOE A), and in most patients with AF on OAC undergoing PCI prefers early discontinuation of aspirin (1–4 weeks) with continued OAC plus a P2Y12 inhibitor over triple therapy (COR 1, LOE A), both to reduce the risk of clinically relevant bleeding.^[2]

Device-detected subclinical AF

• DOAC therapy may be considered in asymptomatic device-detected subclinical AF with elevated thromboembolic risk to prevent ischaemic stroke and thromboembolism, excluding patients at high risk of bleeding (ESC 2024, Class IIb, level B).^[1]
• NOAH (2536 patients with device-detected atrial high-rate episodes) was stopped prematurely for safety concerns and futility for the efficacy of edoxaban, and hence provides limited information: over a median follow-up of 21 months it found no difference between edoxaban and placebo in a composite of cardiovascular death, stroke or embolism (HR 0.81; 95% CI 0.60–1.08; P = 0.15), and those randomised to edoxaban had a higher rate of the composite of death or major bleeding than placebo (HR 1.31; 95% CI 1.02–1.67; P = 0.03); patients in ARTESiA and NOAH had a low burden of device-detected subclinical AF (median duration 1.5 h and 2.8 h, respectively), with thromboembolism rates (around 1% per patient-year) lower than would be expected for an equivalent cohort of patients with clinical AF and a CHA2DS2-VASc score of 4.^[1]
• Whatever the OAC decision, the ESC says these patients need AF-CARE follow-up because the risk of developing clinical AF is high (6%–9% per year).^[1]

^{[1] [6]}

• The ACC/AHA, for device-detected atrial high-rate episodes without previously diagnosed AF, finds it reasonable to start OAC for episodes of 24 hours or more with a CHA2DS2-VASc score of 2 or more or equivalent stroke risk (COR 2a, LOE B-NR); it may be reasonable for episodes of 5 minutes to 24 hours with a score of 3 or more or equivalent stroke risk (COR 2b, LOE B-NR), both within shared decision-making that considers episode duration and individual risk.^[2]
• Episodes under 5 minutes without another indication for OAC should not be anticoagulated (ACC/AHA 2023, COR 3: No Benefit, LOE B-NR).^[2]

Post-operative AF

• Post-operative AF, new-onset AF in the immediate post-operative period, occurs in 30%–50% of patients after cardiac surgery and 5%–30% after non-cardiac surgery, and is associated with a 4–5 times increase in recurrent AF over the next 5 years and with stroke, myocardial infarction, heart failure and death.^[1]

^[1]

• The ESC recommends considering long-term OAC after acute bleeding risk has settled, according to thromboembolic risk factors.^[1]
• In patients who develop post-operative AF after cardiac surgery, the ACC/AHA finds it reasonable to give anticoagulation, when deemed safe with respect to surgical bleeding, for 60 days after surgery unless complications develop, and to reassess the need for longer-term anticoagulation then (COR 2a, LOE B-NR).^[2]
• In patients undergoing cardiac surgery who are at high risk of post-operative AF, the ACC/AHA finds short-term prophylactic beta-blockers or amiodarone reasonable to reduce the incidence of post-operative AF (COR 2a, LOE B-R).^[2]

Trigger-induced AF

• Trigger-induced AF is a new AF episode in close proximity to a precipitating and potentially reversible factor; the most common precipitant unmasking a tendency to AF is acute sepsis, and longer-term data suggest that sepsis-triggered AF recurs after discharge in between a third and a half of patients.^[1]
• Retrospective and observational data suggest these patients carry the same thromboembolic risk as primary AF, and long-term OAC should be considered in suitable patients with trigger-induced AF at elevated thromboembolic risk to prevent ischaemic stroke and systemic thromboembolism (ESC 2024, Class IIa, level C), starting after the acute trigger has been corrected.^[1]
• The ACC/AHA rows for AF identified during acute medical illness or surgery: patients should be counselled about the significant risk of recurrent AF after the acute illness resolves (COR 1, LOE B-NR); outpatient follow-up for thromboembolic risk stratification, decisions on starting or continuing OAC, and AF surveillance can be beneficial given a high risk of recurrence (COR 2a, LOE B-NR); and in critical illness due to sepsis, the benefits of anticoagulation for stroke prevention are uncertain (COR 2b, LOE B-NR).^[2]

Embolic stroke of undetermined source, flutter, cardiomyopathy and thyroid disease

• ESUS: AF is reported to be the mechanism in 30% of patients; prolonged monitoring for AF is recommended to inform AF treatment decisions (ESC 2024, Class I, level B), and starting OAC in ESUS without documented AF is not recommended because of lack of efficacy in preventing ischaemic stroke and thromboembolism (Class III, level A).^[1]
• In stroke or TIA of undetermined cause, the ACC/AHA finds initial cardiac monitoring and, if needed, extended monitoring with an implantable loop recorder reasonable to improve detection of AF (COR 2a, LOE B-R).^[2]
• Atrial flutter: OAC is recommended at elevated thromboembolic risk to prevent ischaemic stroke and thromboembolism (ESC 2024, Class I, level B); rate control can be difficult, and in observational studies 50%–70% of patients have manifested AF during long-term follow-up after flutter ablation.^[1]
• For typical (right-sided, cavotricuspid isthmus-dependent) atrial flutter, the ACC/AHA recommends anticoagulation according to the same risk profile used for AF (COR 1, LOE B-NR), continuation for at least 4 weeks after successful cardioversion or ablation restoring sinus rhythm (COR 1, LOE C-LD), and ongoing OAC as indicated for AF after successful cavotricuspid isthmus ablation of typical flutter when AF was detected before the ablation (COR 1, LOE A).^[2]
• Its footnote adds that left-sided flutter or atrial tachycardia developing after AF ablation should be anticoagulated and managed like AF (ACC/AHA 2023).^[2]
• Hypertrophic cardiomyopathy or cardiac amyloidosis: OAC is recommended in all patients with AF regardless of CHA2DS2-VA score, to prevent ischaemic stroke and thromboembolism (ESC 2024, Class I, level B).^[1]
• Thyroid: the ESC says patients with new-onset or recurrent AF should be tested for TSH; amiodarone induces thyroid dysfunction in 15%–20% of treated patients.^[1]
• In hyperthyroidism with AF and an elevated stroke risk on a standard clinical risk score, the ACC/AHA recommends anticoagulation until thyroid function has returned to normal and sinus rhythm can be maintained (COR 1, LOE B-NR).^[2]
• After stroke: two trials found no difference in clinical outcomes between early and delayed DOAC prescription after ischaemic stroke, and the ESC finds the evidence insufficient to recommend whether OAC should be started or restarted after intracranial haemorrhage.^[1]
• In AF after intracranial haemorrhage, the ACC/AHA finds early (1–2 weeks) resumption of anticoagulation reasonable with very high thromboembolic risk (over 5% per year), such as rheumatic heart disease or a mechanical heart valve, to reduce thromboembolic events (COR 2a, LOE C-LD); delayed (4–8 weeks) resumption may be considered after careful risk–benefit assessment, to balance thromboembolic and haemorrhagic complications (COR 2b, LOE C-LD); and with high risk of recurrent ICH (e.g. cerebral amyloid angiopathy), anticoagulation-sparing strategies such as LAAO may be considered to reduce the risk of recurrent haemorrhage (COR 2b, LOE B-NR).^[2]

Complications and pitfalls

• Heart failure is the most common non-fatal outcome in those with AF, occurring in around half of patients over time, with a four- to five-fold increase in relative risk compared with those without AF in two meta-analyses.^[1]
• Next come ischaemic stroke (RR 2.3), ischaemic heart disease (RR 1.61) and other thromboembolic events, and patients with AF have an increased risk of cognitive impairment (adjusted HR 1.39) and dementia (OR 1.6).^[1]

Prognosis and follow-up

AF carries up to a two-fold increased risk of all-cause mortality (RR 1.95) and cardiovascular mortality (RR 2.03) compared with sinus rhythm, and the most frequent cause of death is heart failure related.^[1] Review is dynamic: in most cases the ESC advises re-evaluation 6 months after initial presentation and then at least annually.^[1]

Special populations

Pregnancy

^[1]

• In the ESC text, LMWH and unfractionated heparin are preferred because they do not cross the placenta; VKAs should be avoided in the first trimester (risk of miscarriage, teratogenicity) and from week 36 onwards (risk of foetal intracranial bleeding if early unexpected delivery), DOACs are not recommended in pregnancy, and atenolol can cause intrauterine growth retardation.^[1]
• The ACC/AHA states that in pregnant patients with AF, DC cardioversion is safe for patient and fetus and should be performed as in patients who are not pregnant (COR 1, LOE B-NR).^[2]
• The ACC/AHA pregnancy rows on drugs: in persistent AF, rate-control agents with a record of safety in pregnancy, such as beta-blockers (e.g. propranolol or metoprolol) and digoxin, alone or with beta-blockers, are reasonable first-line agents (COR 2a, LOE B-NR); without structural heart disease, antiarrhythmic agents with a history of safe use in pregnancy (e.g. flecainide and sotalol) are reasonable for maintaining sinus rhythm (COR 2a, LOE C-LD); with a structurally normal heart and haemodynamically stable AF, pharmacological cardioversion with agents with a history of safe use in pregnancy, such as intravenous procainamide, may be considered (COR 2b, LOE C-LD).^[2]
• Pregnant individuals with AF and elevated stroke risk may be considered for anticoagulation, recognising that no anticoagulation strategy is completely safe for both mother and fetus, with shared decision-making about the risks to both (ACC/AHA 2023, COR 2b, LOE C-LD).^[2]

Older, frail, renal, congenital and cancer patients

• Maintaining VKA treatment rather than switching to a DOAC may be considered in patients aged 75 years or more on clinically stable therapeutic VKA with polypharmacy, to prevent excess bleeding risk (ESC 2024, Class IIb, level B); the ESC text applies this to frail patients aged 75 years or more with polypharmacy who are stable on a VKA.^[1]
• Kidney disease: the ESC notes dabigatran is contraindicated with an eGFR below 30 mL/min/1.73 m2, and ESC Table 11 gives the renal dose-reduction criteria for each DOAC.^[1]
• In AF at elevated stroke risk, the ACC/AHA recommends warfarin or, preferably, evidence-based doses of direct thrombin or factor Xa inhibitors in CKD stage 3 to reduce stroke risk (COR 1, LOE B-R); finds warfarin or labelled DOAC doses reasonable in stage 4 to reduce stroke risk (COR 2a, LOE B-NR); and says warfarin (INR 2.0–3.0) or an evidence-based dose of apixaban might be reasonable in end-stage CKD (CrCl under 15 mL/min) or on dialysis to reduce stroke risk (COR 2b, LOE B-NR).^[2]
• Adult congenital heart disease: OAC should be considered in all adults with AF or flutter and intracardiac repair, cyanosis, Fontan palliation or a systemic right ventricle, to prevent ischaemic stroke and thromboembolism, regardless of other thromboembolic risk factors (ESC 2024, Class IIa, level C).^[1]
• The ACC/AHA says that in adults with AF and moderate or severe congenital heart disease, particularly low-flow states such as Fontan circulation, blind-ending cardiac chambers and cyanosis, anticoagulation independent of conventional risk scores may be reasonable to reduce thromboembolic events (COR 2b, LOE C-LD).^[2]
• Cancer: as in patients without cancer, DOACs have similar efficacy and better safety than VKAs.^[1]
• In most patients with AF and cancer (remote history or receiving active treatment), the ACC/AHA finds DOACs reasonable to choose over VKAs for stroke risk reduction (COR 2a, LOE B-NR).^[2]
• Children: the 2018 NHFA/CSANZ Australian guideline is written for adult patients with AF, and no paediatric recommendations are used on this page.^[4]

Guidelines and regional differences

^{[1] [2]}

In Australia and New Zealand

The 2018 National Heart Foundation of Australia and CSANZ guideline was written to assist Australian practitioners in the diagnosis and management of adult patients with AF; the points below are from its published summary.^[4]

• Opportunistic screening in the clinic or community is recommended for patients over 65 years of age.^[4]
• Deciding between rate and rhythm control at diagnosis and periodically thereafter is highlighted.^[4]
• Beta-blockers or non-dihydropyridine calcium channel antagonists remain the first-line choice for acute and chronic rate control, and cardioversion remains first-line for acute rhythm control when clinically indicated.^[4]
• Flecainide is preferable to amiodarone for acute and chronic rhythm control, and failure of rate or rhythm control should prompt consideration of percutaneous or surgical ablation.^[4]
• The sexless CHA2DS2-VA score is recommended to assess stroke risk; anticoagulation is not recommended for a score of 0 and is recommended for a score of 2 or more.^[4]
• If anticoagulation is indicated, non-vitamin K oral anticoagulants are recommended in preference to warfarin.^[4]
• An integrated care approach delivered by multidisciplinary teams, with patient education and eHealth tools where available, should be adopted.^[4]

For ablation services, the CSANZ published a 2023 expert position statement on catheter and surgical ablation for AF that recognises healthcare factors relevant to the Australian and New Zealand healthcare environments and is tailored to local conditions and populations.^[7]

Exam pearls

• Using the temporal pattern of clinical AF (paroxysmal, persistent or permanent) is not recommended to determine the need for OAC (ESC 2024, Class III, level B): AF is a major thromboembolic risk factor whether paroxysmal, persistent or permanent.^[1]
• ESC 2024 uses CHA2DS2-VA: a score of 2 or more is recommended as an indicator of elevated thromboembolic risk for decisions on initiating OAC (Class I, level C) and a score of 1 should be considered an indicator of elevated thromboembolic risk for decisions on initiating OAC (Class IIa, level C); ACC/AHA 2023 bases anticoagulation in patients with AF on estimated annual thromboembolic risk from a validated score such as CHA2DS2-VASc: recommended at 2% or more per year (e.g. a score of 2 or more in men, 3 or more in women; COR 1, LOE A) and reasonable at 1% to under 2% (equivalent to 1 in men, 2 in women; COR 2a, LOE A), to prevent stroke and systemic thromboembolism.^[1][2]
• Hypertrophic cardiomyopathy and cardiac amyloidosis: OAC is recommended in all patients with AF regardless of CHA2DS2-VA score, to prevent ischaemic stroke and thromboembolism (ESC 2024, Class I, level B).^[1]
• In AF with LVEF above 40%, beta-blockers, diltiazem, verapamil or digoxin are recommended as first-choice drugs; with LVEF 40% or less, beta-blockers and/or digoxin are recommended; both to control heart rate and reduce symptoms (ESC 2024, both Class I, level B); for long-term rate control the ACC/AHA recommends beta-blockers or non-dihydropyridine calcium channel blockers (diltiazem, verapamil), with the choice of agent according to underlying substrate and comorbid conditions (COR 1, LOE B-NR) and finds digoxin reasonable in AF with heart failure symptoms, combined with other rate-controlling agents or as monotherapy if other agents are not preferred, not tolerated or contraindicated (COR 2a, LOE B-R).^[1][2]
• Atrioventricular node ablation with CRT should be considered in severely symptomatic permanent AF with at least one heart failure hospitalisation, to reduce symptoms, physical limitations, recurrent heart failure hospitalisation and mortality (ESC 2024, Class IIa, level B).^[1]
• After ablation, OAC is recommended for at least 2 months in all patients irrespective of rhythm outcome or score, to reduce the risk of peri-procedural ischaemic stroke and thromboembolism, and then according to the CHA2DS2-VA score, to prevent ischaemic stroke and thromboembolism (ESC 2024, both Class I, level C); the ACC/AHA says OAC should be continued for at least 3 months after ablation, with a longer duration determined by underlying risk (COR 1, LOE B-NR).^[1][2]

References8ShowHide

1. [1]Van Gelder IC, Rienstra M, Bunting KV, et al. 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J, 2024.PMID 39210723
2. [2]Joglar JA, Chung MK, Armbruster AL, et al. 2023 ACC/AHA/ACCP/HRS Guideline for the Diagnosis and Management of Atrial Fibrillation: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation, 2024.PMID 38033089
3. [3]European Society of Cardiology Correction to: 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): Developed by the task force for the management of atrial fibrillation of the European Society of Cardiology (ESC), with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Endorsed by the European Stroke Organisation (ESO). Eur Heart J, 2025.PMID 40622753
4. [4]Brieger D, Amerena J, Attia JR, et al. National Heart Foundation of Australia and Cardiac Society of Australia and New Zealand: Australian clinical guidelines for the diagnosis and management of atrial fibrillation 2018. Med J Aust, 2018.PMID 30067936
5. [5]Kirchhof P, Camm AJ, Goette A, et al. Early Rhythm-Control Therapy in Patients with Atrial Fibrillation. N Engl J Med, 2020.PMID 32865375
6. [6]Healey JS, Lopes RD, Granger CB, et al. Apixaban for Stroke Prevention in Subclinical Atrial Fibrillation. N Engl J Med, 2024.PMID 37952132
7. [7]Kistler PM, Sanders P, Amarena JV, et al. 2023 Cardiac Society of Australia and New Zealand Expert Position Statement on Catheter and Surgical Ablation for Atrial Fibrillation. Heart Lung Circ, 2024.PMID 38702234
8. [8]Mills NL, Newby LK, Zaman S, et al. Fifth Universal Definition of Myocardial Infarction (2026). Glob Heart, 2026.PMID 42666939