Atrial fibrillation: an irregular rhythm, the clot it makes, and two decisions that are made separately

Topic: atrial fibrillation: an irregular rhythm, the clot it makes, and the two decisions that are made separately · Since 1980 · Grounded citations only · Published 2026-08-31

A normal heartbeat starts in the right atrium, spreads across both upper chambers so they contract together, and passes through a single electrical gate to the ventricles below. In atrial fibrillation, that coordination fails. The atria are driven by continuous disorganised electrical activity and no longer contract as a unit — they quiver. The gate passes impulses through erratically, so the ventricles beat irregularly and often fast. On the wrist this is an irregularly irregular pulse, and it is the oldest sign of the condition.

Two consequences follow, and they are almost entirely separate problems.

The first is symptoms. Palpitations, breathlessness, fatigue, reduced exercise tolerance. Losing the atrial contribution to ventricular filling and running at an uncontrolled rate makes the heart a less efficient pump. Some people are severely limited. Many feel nothing at all.

The second is stroke, and it is the one that kills. Blood that is not being pushed cleanly out of a contracting atrium can stagnate — particularly in the left atrial appendage, a small blind-ended pouch off the left atrium whose structure and flow characteristics make it the dominant site of thrombus formation [1] [2] [3] [4]. A clot that forms there and embolises travels up the carotid arteries to the brain.

The size of that risk is the reason this review exists. In the Framingham cohort followed 34 years, atrial fibrillation carried a near fivefold excess of stroke — and, uniquely among the cardiac risk factors examined, its effect did not weaken with age. The attributable risk of stroke from hypertension, coronary disease and cardiac failure all fell with advancing age; for atrial fibrillation it rose from 1.5 percent at ages 50–59 to 23.5 percent at ages 80–89 [5]. An earlier Framingham analysis put the proportion of strokes associated with the arrhythmia at 14.7 percent overall — 68 of 462 first strokes — rising from 6.7 percent at ages 50–59 to 36.2 percent at ages 80–89 [6]. In the very old, this one arrhythmia is behind more than a third of strokes.

It is common and getting commoner. United States prevalence rose from 0.1 percent below age 55 to 9.0 percent at 80 and over, with an estimated 2.3 million affected adults projected to exceed 5.6 million by 2050 [7]; the number of European Union adults aged 55 and over with atrial fibrillation is projected to more than double between 2010 and 2060 [8]. Global burden, incidence, prevalence and associated mortality all rose between 1990 and 2010, with AF-associated mortality roughly doubling in both sexes over that period [9] [10]. Incidence and prevalence have been characterised in Framingham [11] [12], Rotterdam [13], Olmsted County [14], Renfrew/Paisley [15], older adults [16] and long before all of them [17].

Here is the number people misquote. The lifetime risk of developing atrial fibrillation is 1 in 4 for men and women aged 40 and over — and 1 in 6 even in the absence of prior heart failure or myocardial infarction [18]. That is the risk of getting the arrhythmia, not the fraction of strokes it causes. The two "one in four" figures are different quantities and are frequently confused.

And much of it is silent. Atrial fibrillation is often paroxysmal — coming and going — and often asymptomatic, which means the first presentation can be the stroke. That is what makes detection a subject in its own right, and it is covered below.

The organising fact of management, which this review returns to repeatedly: the decision to anticoagulate and the decision about rhythm are made separately, on different evidence, for different reasons. Restoring sinus rhythm does not earn a patient the right to stop anticoagulation. Getting that wrong causes strokes.

Centerpiece: turning a list of risk factors into an annual stroke risk

0 1 2 3 4 5 6 risk score 0.5 1 2 5 10 20 stroke / thromboembolism per 100 person-years (log scale) 2 per 100 person-years f i t t e d   C H A D S - V A S c   c u r v e 2 2 crosses 2%/year at a score of 1.52 — between 1 and 2, where the threshold sits fitted x1.46 per point (the paper states x1.5) fitted x1.77 per point (never stated anywhere) Straight lines on a log axis mean a constant multiplier per point. Both scores are close to straight: t h e   l o g - l i n e a r   f i t   e x p l a i n s   1 0 0 . 0 %   o f   t h e   v a r i a n c e   f o r   C H A D S   a n d   9 7 . 5 %   f o r   C H A D S - V A S c .   S h a d e d   b a n d s   a r e 2 2 2 the reported confidence intervals. Neither cohort was anticoagulated — these are untreated rates. A list of risk factors becomes an annual risk C H A D S   ( M e d i c a r e   c o h o r t ,   o f f   a n t i t h r o m b o t i c s ) 2 C H A D S - V A S c   ( D a n i s h   r e g i s t r i e s ,   n o t   a n t i c o a g u l a t e d ) 2 2 0 2 4 6 8 observed stroke rate per 100 patient-years (lower bars: % of patients, scaled ÷10) called "LOW risk" 0–2.3 per 100 py called "HIGH risk" 2.5–7.9 per 100 py 0.2 percentage points separate the two labels 9–49% of the SAME patients called low risk 11–77% of the SAME patients called high risk Twelve published schemes, one set of patients The upper pair is the observed risk behind each label; the lower pair is how many of the same patients each label captures. A score is only as useful as the scheme generating it — an argument for standardising on one, not for trusting any of them absolutely. Which score you pick decides who gets treated
Left: annual stroke rate against risk score in two untreated cohorts, on a log scale — straight lines mean a constant multiplier per point, and the fitted CHA2DS2-VASc curve crosses 2 per 100 person-years right where the treatment threshold sits. Right: what happens when twelve different published schemes are applied to the same patients.

Everyone with atrial fibrillation has an increased stroke risk. Not everyone has enough of one to justify lifelong anticoagulation, because anticoagulation causes bleeding. The clinical problem is therefore quantitative, and it is solved with a score.

CHADS2 was the first widely used version. It assigns one point each for congestive heart failure, hypertension, age 75 or over, and diabetes, and two points for prior stroke or transient ischaemic attack — the doubling being why the acronym carries a subscript 2 [19]. Validated in 1,733 Medicare beneficiaries with nonrheumatic atrial fibrillation who were not prescribed warfarin, across 2,121 patient-years and 94 ischaemic strokes, it achieved a c statistic of 0.82 (0.80–0.84), better than the two schemes it was built from [19].

Its output is the left panel of the figure: stroke rate per 100 patient-years without antithrombotic therapy of 1.9 at score 0, 2.8 at 1, 4.0 at 2, 5.9 at 3, 8.5 at 4, 12.5 at 5 and 18.2 at 6 [19]. The paper also states the pattern as a single number — the rate increased by a factor of 1.5 (1.3–1.7) for each 1-point increase [19].

The first check is on that summary. Fitting a constant multiplier to the seven printed rates recovers ×1.46 per point, which rounds to the stated 1.5, and the log-linear fit explains 99.99 percent of the variance. Those seven numbers lie almost exactly on an exponential. A clinical score built by adding up integers turns out to behave like a clean multiplicative model of risk — which is not obvious, and is what makes the score interpretable rather than merely predictive.

CHA2DS2-VASc refined it by adding what CHADS2 left out [20]: vascular disease, age 65–74, and female sex, with age 75 or over promoted to two points. It was developed by reclassifying the Birmingham/NICE schema and tested in 1,084 patients from the Euro Heart Survey [20]. The Danish nationwide registries then quantified it in 47,576 patients with a CHADS2 score of 0–1 who were not on a vitamin K antagonist or heparin: stroke or thromboembolism per 100 person-years at one year of 0.84 (0.65–1.08) at score 0, 1.79 (1.53–2.09) at 1, 3.67 (3.34–4.03) at 2, 5.75 (5.33–6.21) at 3 and 8.18 (6.68–10.02) at 4 [21].

The check the arithmetic was never given. The Danish paper prints five rates and never fits a multiplier to them. It is ×1.77 per point — a gradient about 21 percent steeper than the ×1.5 CHADS2 states for itself. The finer score does not merely re-rank patients into more boxes; it separates them faster per box. And its fit to a constant multiplier is slightly worse than CHADS2's (R² 0.974 against 0.9999) in a specific and useful way: the steps from 0→1 and 1→2 are larger than the steps from 2→3 and 3→4. The curve bends most steeply at the bottom — exactly the range in which the treatment decision is actually being made.

And the threshold falls out of the curve. Extrapolating the fitted CHA2DS2-VASc relationship, annual stroke risk crosses 2 per 100 person-years at a score of 1.52 — between 1 and 2, which is where guidelines place the anticoagulation threshold for men. The threshold is not an arbitrary line on an integer scale. It is the score at which the untreated risk crosses a level that the treatment's benefit reliably exceeds its bleeding cost.

Why the refinement mattered, in one comparison. The Euro Heart Survey analysis found the newer scheme's overall discrimination only modestly better (c statistic 0.606 against CHADS2's). What it did far better was the thing that matters at the threshold: patients it classified as low risk had no thromboembolic events at all, while 1.4 percent of CHADS2 low-risk subjects had one [20]. It also cut the proportion parked in the uninformative intermediate stratum from 61.9 percent to 15.1 percent [20]. And in the Danish cohort, patients with a CHADS2 score of 0 — nominally all low risk — had one-year event rates ranging from 0.84 to 3.2 percent depending on their CHA2DS2-VASc score [21]. CHADS2 was not identifying a low-risk group; it was averaging over one. The value of CHA2DS2-VASc is in ruling patients out of treatment safely, not in ranking high-risk patients more precisely. Further validations followed in low-risk groups [22], in Asian populations [23] and against clinical outcomes generally [24] [25], along with the effect on anticoagulation recommendations [26].

The right panel is the honest limit. Twelve published stratification schemes were applied to test cohorts. Observed stroke rates for patients labelled low risk ranged from 0 to 2.3 per 100 patient-years, and for high risk from 2.5 to 7.9 — the two labels separated by 0.2 percentage points of actual risk at their boundary. The fraction of the same patients called low risk varied from 9 to 49 percent, and high risk from 11 to 77 percent [27]. Alternative schemes continue to be proposed and some outperform both on discrimination — the ATRIA score reached c-indices of 0.73 in derivation and 0.70 in external validation, with positive net reclassification against CHADS2 and CHA2DS2-VASc [28] — as do echocardiographic and biomarker-based refinements [29] [30] [31] [32] [33] [34] [35].

Three limits on the figure itself. The Danish cohort was restricted to patients with CHADS2 0–1, so the CHA2DS2-VASc curve only spans scores 0–4 and the highest scores are not represented; the extrapolated crossing point at 1.52 is a fit, not an observation; and the two curves come from different populations, countries and eras, so their vertical offset should not be read as one score being safer than the other.

The teaching point. The score converts an unordered list of clinical facts into a single estimated annual probability, and treatment is offered once that probability crosses a threshold set by what the treatment can achieve. That is risk-based decision-making in its cleanest clinical form — and the right panel is the reminder that the estimate is only as good as the scheme producing it.

Pillar 1: measurement and diagnosis

The ECG, which is the whole diagnosis

Atrial fibrillation is an electrocardiographic diagnosis and nothing else will do. The tracing shows an irregularly irregular ventricular rhythm with no discernible P waves — no organised atrial depolarisation to produce them. No symptom, pulse check or imaging finding substitutes; a recording of the rhythm is required, which is why the practical problem is not interpretation but capture.

Catching it when it comes and goes

Atrial fibrillation is classified by how it behaves over time — paroxysmal (terminating spontaneously), persistent (requiring intervention to terminate) and permanent (accepted as the ongoing rhythm) — and this classification, set out in the successive management guidelines [36] [37] [38], drives everything from monitoring strategy to the realistic goals of rhythm control. Its practical consequence is that a normal ECG in clinic excludes nothing.

Ambulatory monitoring extends the recording window, from 24-hour Holter through multi-day patches to implantable loop recorders, which have been used to quantify arrhythmia burden after ablation with a precision intermittent recording cannot match [39].

Screening detects a real and treatable population. In a cluster-randomised trial across 50 English primary care centres and 14,802 patients aged 65 or over, active screening raised the detection rate of new atrial fibrillation to 1.63 percent per year against 1.04 percent in control practices — and systematic invitation to electrocardiography and opportunistic pulse-taking performed almost identically (1.62 versus 1.64 percent) [40]. Stepwise screening of an entire 75–76-year-old Swedish municipality is more striking still: of 848 participants, 12-lead ECG found previously undiagnosed silent atrial fibrillation in 10 (1 percent), and among 403 people with two or more CHADS2 risk factors who completed two weeks of twice-daily handheld ECG recording, 30 (7.4 percent) were diagnosed with paroxysmal atrial fibrillation — with 9 percent of the whole screened population identified as candidates for anticoagulation. The same study found that 35 of 81 people with already-known atrial fibrillation (43 percent) were not on oral anticoagulation [41]. The undertreatment of known disease was as large a problem as the undiagnosed disease. Systematic reviews of screening yield [42] and primary-care approaches [43] reach compatible conclusions.

Consumer wearables changed the scale of this. In the Apple Heart Study, 419,297 participants used a smartwatch irregular-pulse notification algorithm over a median 117 days. Only 2,161 (0.52 percent) were ever notified [44]. Among the 450 who returned analysable ECG patches, atrial fibrillation was present in 34 percent (97.5% CI 29–39), and the positive predictive value of a notification against simultaneous ECG was 0.84 for a subsequent irregular pulse notification and 0.71 for an irregular tachogram [44]. Smartphone photoplethysmography has been assessed for the same purpose [45], and the wider use of cardiac wearables reviewed [46]. Read carefully, those numbers cut both ways: the notification rate is low and its predictive value respectable, but two-thirds of notified participants who returned a patch did not have atrial fibrillation captured on it.

Device-detected atrial fibrillation, and a finding that complicates the whole story

Implanted pacemakers and defibrillators monitor the atrium continuously, which created a population nobody had been able to study: people with subclinical atrial fibrillation and no clinical arrhythmia. In ASSERT, 2,580 patients aged 65 or over with hypertension and no history of atrial fibrillation were monitored after device implantation [47]. Subclinical atrial tachyarrhythmias occurred in 261 (10.1 percent) by 3 months, and were associated with a hazard ratio of 2.49 (1.28–4.85) for ischaemic stroke or systemic embolism, with a population attributable risk of 13 percent [47]. Device-detected atrial fibrillation predicts stroke across large registries too [48], and the entity now has a formal European consensus definition [49].

But the temporal analysis of that same trial is the finding to sit with. Of the patients who had a stroke or systemic embolism, very few had subclinical atrial fibrillation in the month beforehand. Among the 14 with an episode detected more than 30 days before the event, the most recent episode had occurred a median of 339 days earlier. And 8 patients (16 percent) had subclinical atrial fibrillation detected only after their stroke, despite continuous monitoring for a median of 228 days before it [50].

That is hard to reconcile with the simple mechanical story of this review's opening. If clots formed in a fibrillating atrium and embolised promptly, episodes should cluster before events. They do not. Atrial fibrillation may be as much a marker of an abnormal atrium — an atrial cardiomyopathy with its own thrombotic tendency — as the proximate cause of each clot. The clinical implication is direct, and it is why ARTESiA tested apixaban specifically in subclinical device-detected atrial fibrillation [51] [52]. That trial's abstract is not available in OpenAlex — only its title — so this review does not quote its results, and readers should go to the trial report.

What else the workup is for

An atrial fibrillation diagnosis is the start of a search for drivers and consequences: hypertension, obstructive sleep apnoea, obesity, alcohol, thyroid disease, and structural heart disease on echocardiography. Left atrial size and structural remodelling both predict the arrhythmia and are produced by it [53] [54] [12]. Atrial fibrillation complicating acute myocardial infarction is its own well-described entity [55], as is atrial fibrillation after cardiac surgery [56] [57]. Bleeding risk is assessed alongside stroke risk, most commonly with HAS-BLED — hypertension, abnormal renal or liver function, stroke, bleeding history, labile INR, elderly over 65, drugs or alcohol — derived from 3,978 Euro Heart Survey patients with a c statistic of 0.72 overall [58] [59] [60]. Its purpose is frequently misunderstood: a high bleeding score identifies modifiable bleeding risks to correct, not a reason to withhold anticoagulation from someone whose stroke risk warrants it.

Pillar 2: management — two decisions, made separately

This is the section where the structure matters more than any individual fact.

Decision one: stroke prevention

Anticoagulation works, and the effect size is large. A meta-analysis of the randomised trials found that adjusted-dose warfarin reduces stroke by approximately 60 percent and antiplatelet agents by approximately 20 percent, with warfarin about 40 percent more efficacious than antiplatelet therapy (relative risk reduction 39 percent, CI 22–52, across 12 trials and 12,963 participants) — and absolute increases in major extracranial haemorrhage of 0.3 percent per year or less, smaller than the absolute stroke reductions [61]. The founding trials established this against placebo [62] [63] [64] [65], including in elderly community populations [66], and clopidogrel plus aspirin was tested as an alternative and found inferior to oral anticoagulation [67].

Warfarin's problem was never efficacy. It is a vitamin K antagonist, acting on VKORC1 — the enzyme whose variants cause both warfarin resistance [68] and much of the between-person dose variation [69], alongside CYP2C9 variants affecting dose requirement and bleeding risk [70], to the point that pharmacogenetic dosing algorithms were developed [71]. Add a long list of drug and food interactions [72], a narrow therapeutic window requiring regular INR testing, and bleeding that rises steeply outside it [73] [74], and the practical result was that a highly effective drug was widely underprescribed and often poorly controlled.

The direct oral anticoagulants solved the practical problem. Two mechanisms, both downstream of warfarin's: dabigatran directly inhibits thrombin, the active form of prothrombin [75], while rivaroxaban, apixaban and edoxaban inhibit coagulation factor X in its activated form. The four pivotal trials, each against dose-adjusted warfarin:

A meta-analysis of these trials collected the class effect [80], and registry data confirmed the pattern outside trials [81]. The consistent finding across all four is less intracranial haemorrhage, which is what changes the risk-benefit calculation most for the patients at highest stroke risk. Practical guidance on their use is extensive [82], as is the management of bleeding on them [83], including specific reversal with idarucizumab for dabigatran [84] [85] and prothrombin complex concentrate [86]. Dosing in extremes of body weight has been studied [87] [88], and the boundary of their applicability was found the hard way: dabigatran was worse than warfarin in patients with mechanical heart valves [89], which is why "nonvalvular" appears in the indication.

For patients who genuinely cannot take an anticoagulant, apixaban was compared against aspirin in exactly that group and reduced stroke or systemic embolism by more than half (hazard ratio 0.45, 0.32–0.62) without significantly increasing major bleeding (1.4 against 1.2 percent per year) [90]. That trial largely removed aspirin monotherapy as a defensible option.

Left atrial appendage closure is the mechanical alternative — occlude the pouch where the clots form [91]. In PROTECT AF, 707 patients randomised 2:1 to a device or warfarin over a mean 3.8 years had a primary event rate of 2.3 against 3.8 per 100 patient-years (rate ratio 0.60), meeting both non-inferiority and superiority, with lower cardiovascular mortality (1.0 against 2.4 per 100 patient-years, hazard ratio 0.40) and all-cause mortality (3.2 against 4.8, hazard ratio 0.66) [92] [93]. Procedural safety improved with experience [94] [95], the approach has been used specifically in patients with contraindications to anticoagulation [96] [97], compared against direct oral anticoagulants [98] [99], and deployed across multicentre practice [100] [101] [102], with imaging protocols for sizing and planning [103] [104] [105].

And surgical occlusion, tested at scale in LAAOS III: among patients with atrial fibrillation undergoing cardiac surgery for another reason, with a mean CHA2DS2-VASc score of 4.2 and 76.8 percent still on oral anticoagulation at 3 years, appendage occlusion reduced stroke or systemic embolism from 7.0 to 4.8 percent (hazard ratio 0.67, 0.53–0.85, P = 0.001) over a mean 3.8 years, without excess perioperative bleeding, heart failure or death [106]. That result is important beyond the operating theatre: the benefit was on top of continued anticoagulation, which means the appendage contributes residual risk that drugs do not fully cover.

Decision two: rate or rhythm

Having settled anticoagulation, the separate question is what to do about the rhythm itself.

For two decades the answer was: it does not much matter, so keep it simple. AFFIRM randomised 4,060 patients with atrial fibrillation and high stroke or death risk to rhythm control or rate control. Five-year mortality was 23.8 percent with rhythm control against 21.3 percent with rate control (hazard ratio 1.15, 0.99–1.34, P = 0.08), with more hospitalisations in the rhythm-control arm [107]. RACE reached the same conclusion in 522 patients with recurrent persistent atrial fibrillation, where rate control was not inferior on a composite endpoint despite only 10 percent of the rate-control group being in sinus rhythm at 2.3 years against 39 percent of the rhythm-control group [108]. Further trials [109] [110], quality-of-life analyses [111], reviews [112] and secondary analyses of AFFIRM's mortality [113] and strokes [114] followed. A key AFFIRM on-treatment analysis found that the presence of sinus rhythm was associated with a lower risk of death, and antiarrhythmic drugs with increased mortality once sinus rhythm was adjusted for — the authors concluding that any beneficial antiarrhythmic effect of the available drugs was offset by their adverse effects. The same analysis found warfarin use improved survival [115], which is the two-decisions principle emerging from inside a rhythm trial.

Rate control also turned out not to need to be strict. RACE II randomised 614 patients with permanent atrial fibrillation to a lenient target (resting heart rate under 110) or a strict one (resting under 80 and under 110 on moderate exercise) [116]. Three-year cumulative incidence of the primary composite was 12.9 percent with lenient against 14.9 percent with strict control, meeting non-inferiority [116]. The easier target was no worse, and was far easier to reach.

Then EAST-AFNET 4 changed the picture, by changing when. It randomised 2,789 patients with early atrial fibrillation — median 36 days since diagnosis — and cardiovascular conditions to early rhythm control (antiarrhythmic drugs or ablation) or usual care in which rhythm control was reserved for symptoms [117]. Stopped early for efficacy after a median 5.1 years, the first primary outcome — cardiovascular death, stroke, or hospitalisation for worsening heart failure or acute coronary syndrome — occurred at 3.9 per 100 person-years with early rhythm control against 5.0 with usual care (hazard ratio 0.79, 96% CI 0.66–0.94, P = 0.005) [117]. Nights in hospital did not differ, and symptoms and left ventricular function at 2 years did not differ either — so the benefit was not symptomatic. Serious adverse events related to rhythm-control therapy occurred in 4.9 percent against 1.4 percent [117], which is the cost. The strategy also benefited patients with and without symptoms [118] and those with heart failure [119].

The reconciliation with AFFIRM is about timing and tools: EAST-AFNET treated patients within weeks of diagnosis, using safer drug options and ablation, rather than after years of established arrhythmia with the antiarrhythmics of the 1990s.

Catheter ablation rests on one 1998 observation. Mapping the spontaneous initiation of atrial fibrillation in 45 patients found 69 ectopic foci, of which 65 (94 percent) were in the pulmonary veins [120]. Electrically isolating those veins from the left atrium became the procedure, with defined electrophysiological endpoints [121] [122] [123], approaches for persistent disease [124] [125], the non-pulmonary-vein triggers that account for the rest [126], substrate imaging with fibrosis on delayed-enhancement MRI [127] [128], and worldwide surveys of methods, efficacy and safety [129] [130]. It has real complications, including the rare and often fatal atrio-oesophageal fistula [131], phrenic nerve injury [132] and stiff left atrial syndrome [133].

What ablation delivers is rhythm, and — in the trials — not clearly mortality. Against antiarrhythmic drugs it consistently wins on arrhythmia outcomes [134] [135] [136] [137], though as first-line therapy in paroxysmal atrial fibrillation MANTRA-PAF found no significant difference in cumulative arrhythmia burden over 2 years, with ablation better on the secondary measures (freedom from any atrial fibrillation 85 against 71 percent) and carrying procedural risk including one death from a procedure-related stroke [138]. CABANA randomised 2,204 patients to ablation or drug therapy: over a median 48.5 months the primary composite of death, disabling stroke, serious bleeding or cardiac arrest occurred in 8.0 against 9.2 percent (hazard ratio 0.86, 0.65–1.15, P = 0.30) — not significant — while death or cardiovascular hospitalisation (51.7 against 58.1 percent, P = 0.001) and atrial fibrillation recurrence (49.9 against 69.5 percent, P < 0.001) both favoured ablation [139]. The trial had 27.5 percent crossover from drugs to ablation [139], which is the standard caveat on its null primary result.

The exception is heart failure, where the effect is large. CASTLE-AF randomised 363 patients with symptomatic atrial fibrillation, NYHA class II–IV heart failure, ejection fraction 35 percent or less and an implanted defibrillator to ablation or medical therapy. Over a median 37.8 months the composite of death or hospitalisation for worsening heart failure occurred in 28.5 percent with ablation against 44.6 percent with medical therapy [140]. Earlier work had shown ablation feasible and beneficial in congestive heart failure [141], ablation beat amiodarone in persistent atrial fibrillation with heart failure [142], and left ventricular dysfunction can reverse after ablation [143] — the tachycardia-induced component of the cardiomyopathy resolving once the rhythm does. Set against this, AF-CHF had earlier found no mortality benefit from a drug-based rhythm-control strategy in atrial fibrillation with congestive heart failure (death from any cause 32 against 33 percent) [144]. Ablation and antiarrhythmic drugs are not interchangeable ways of achieving the same thing.

Antiarrhythmic drugs themselves remain in use for termination and maintenance — amiodarone for persistent atrial fibrillation [145] and acute rate control in the critically ill [146], ibutilide for flutter [147] — with the perennial constraint that the agents that suppress atrial fibrillation best also carry the most proarrhythmic and organ toxicity. Digoxin's safety in this population has been examined and no increased mortality found in one large analysis [148]. Upstream therapy targeting the substrate rather than the rhythm has been tried, for example with angiotensin receptor blockade to maintain sinus rhythm [149].

Treating the drivers, which is neither decision and belongs to both

Obesity. In Framingham, 5,282 participants followed a mean 13.7 years produced 526 incident cases, with body mass index associated with new-onset atrial fibrillation in part through left atrial dimension [150]; obesity also predicts progression from paroxysmal to permanent atrial fibrillation [151] [152]. And it is reversible: in 150 overweight or obese patients with symptomatic atrial fibrillation randomised to structured weight management against general lifestyle advice, with intensive cardiometabolic risk-factor management in both arms, the intervention reduced symptom burden and severity, atrial fibrillation episodes and cumulative duration on 7-day Holter, and left atrial area [153].

Alcohol is a risk factor for incident atrial fibrillation [154], and abstinence is a treatment: 140 patients drinking 10 or more standard drinks weekly with paroxysmal or persistent atrial fibrillation were randomised to abstain or continue. The abstinence group cut intake from 16.8 to 2.1 drinks weekly (a 87.5 percent reduction) and had greater freedom from recurrence and lower atrial fibrillation burden over 6 months [155].

Obstructive sleep apnoea. Recurrence of atrial fibrillation 12 months after cardioversion was 82 percent in untreated or inadequately treated sleep apnoea, against 42 percent in the CPAP-treated group and 53 percent in controls [156]. Treated apnoea improves outcomes after ablation too [157] [158], and the mechanism has been probed experimentally [159].

And hypertension, diabetes and coronary disease are the shared cardiometabolic substrate; guidelines fold their control into atrial fibrillation management [36].

Pillar 3: progress

Early rhythm control, and ablation earlier in the pathway. EAST-AFNET 4 is the shift [117] [118], reinforced in heart failure by CASTLE-AF [140] and the heart-failure analysis of early rhythm control [119]. The open question is which patients gain enough to accept a 4.9-versus-1.4 percent rate of therapy-related serious adverse events [117].

Screening at population scale. Consumer wearables make identification cheap and continuous where it was episodic and expensive [44] [45] [46], on top of the established yield from organised primary-care and community screening [40] [41] [42]. The unresolved question is what to do with what they find: device-detected subclinical atrial fibrillation raises stroke risk [47] [48] [49], but the episodes do not reliably precede the events [50], so the burden threshold at which anticoagulation becomes worthwhile is genuinely unknown. Atrial fibrillation also associates with cognitive decline and dementia, partly through subclinical cerebral infarcts [160] [161] — an outcome that screening advocates increasingly cite.

Risk-factor and lifestyle management as therapy rather than as advice, following the weight-reduction [153], alcohol-abstinence [155] and sleep-apnoea [156] [158] results. This is the least glamorous and possibly the most transferable finding in the field: three randomised or controlled interventions that change atrial fibrillation burden without touching the heart directly.

A gap this substrate cannot cover. The task specification lists factor XI inhibitors — abelacimab, asundexian, milvexian — as a progress area. This substrate contains no work on them at all: not a trial, not a review, not a mechanism paper. The rationale is easy to state and is not cited here because nothing in the substrate supports it. Readers should go elsewhere for that literature; this review will not assert results it cannot ground.

Two further directions the substrate does support: improved risk prediction, where the ATRIA score already outperformed CHADS2 and CHA2DS2-VASc on discrimination and was designed as a foundation for adding biomarkers [28] [32]; and the atrial substrate itself as a treatment target, with fibrosis on MRI predicting ablation outcome [127], extracellular matrix remodelling maintaining the arrhythmia [54], structural correlates identified in human atria [53], abnormal calcium handling from the sarcoplasmic reticulum implicated in the cellular mechanism [162], and the rapid-atrial-pacing model showing that atrial fibrillation remodels the atrium to sustain itself [163]. The cardiac ion channels through which several antiarrhythmics act and in which atrial-fibrillation-associated variants sit — SCN5A and KCNQ1 — are the molecular end of that story.

Dig deeper in lmmol

Stroke is the reason atrial fibrillation is treated aggressively, and the two reviews should be read together: this arrhythmia accounted for 14.7 percent of first strokes overall in Framingham and 36.2 percent at ages 80–89 [6], its stroke effect uniquely fails to weaken with age [5], and cardioembolic stroke is preventable by anticoagulation in a way that most stroke is not [61]. Heart failure has a two-way relationship with atrial fibrillation — each causes and worsens the other, rhythm control by drugs failed to help in AF-CHF [144], and ablation produced the largest effect seen anywhere in this literature in CASTLE-AF [140], partly by reversing tachycardia-induced ventricular dysfunction [143]. Coronary artery disease shares the risk-score logic exactly — a list of clinical factors converted into an annual probability that decides treatment — and intersects clinically when a patient on anticoagulation also needs antiplatelet therapy after stenting [164] [165]. Hypertension is the commonest driver, present in 70.8 percent of AFFIRM participants [107]. Obstructive sleep apnea and obesity are the two modifiable drivers with randomised or controlled evidence that treating them reduces the arrhythmia itself [156] [153]. Diabetes and glycemic control contributes a point to both risk scores [19] [20]. On the molecular side, the three anticoagulant classes map to three distinct targets: VKORC1 for warfarin [69], prothrombin for dabigatran [75], and coagulation factor X for the Xa inhibitors [77] [78] [79]; SCN5A and KCNQ1 are the cardiac ion channels behind the rhythm itself. The full collection is at health.

Key papers

  1. W2156390580: Left atrial appendage: structure, function, and role in thromboembolism (cited 691×)
  2. W1967841125: Left Atrial Appendage (cited 644×)
  3. W2067296986: Anatomy of the Normal Left Atrial Appendage (cited 419×)
  4. W2029000698: Assessment of left atrial appendage function by transesophageal echocardiography. Implications for the development of thrombus. (cited 484×)
  5. W2114775432: Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. (cited 7,523×)
  6. W4233664165: Atrial fibrillation: a major contributor to stroke in the elderly. The Framingham Study (cited 1,213×)
  7. W2132505348: Prevalence of Diagnosed Atrial Fibrillation in Adults (cited 6,246×)
  8. W2159975225: Projections on the number of individuals with atrial fibrillation in the European Union, from 2000 to 2060 (cited 1,445×)
  9. W2147018083: Worldwide Epidemiology of Atrial Fibrillation (cited 4,574×)
  10. W3036116092: Epidemiology of Atrial Fibrillation in the 21st Century (cited 1,524×)
  11. W1984364985: Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study (cited 2,929×)
  12. W2067232731: Echocardiographic predictors of nonrheumatic atrial fibrillation. The Framingham Heart Study. (cited 1,076×)
  13. W2148056560: Prevalence, incidence and lifetime risk of atrial fibrillation: the Rotterdam study (cited 1,954×)
  14. W2141889414: Secular Trends in Incidence of Atrial Fibrillation in Olmsted County, Minnesota, 1980 to 2000, and Implications on the Projections for Future Prevalence (cited 2,586×)
  15. W2139693734: Population prevalence, incidence, and predictors of atrial fibrillation in the Renfrew/Paisley study (cited 663×)
  16. W2067026685: Incidence of and Risk Factors for Atrial Fibrillation in Older Adults (cited 1,470×)
  17. W2324628668: Epidemiologic Features of Chronic Atrial Fibrillation (cited 2,344×)
  18. W2137434301: Lifetime Risk for Development of Atrial Fibrillation (cited 2,055×)
  19. W2062786076: Validation of Clinical Classification Schemes for Predicting Stroke (cited 4,861×)
  20. W2162586165: Refining Clinical Risk Stratification for Predicting Stroke and Thromboembolism in Atrial Fibrillation Using a Novel Risk Factor-Based Approach (cited 6,798×)
  21. W2070601915: The value of the CHA2DS2-VASc score for refining stroke risk stratification in patients with atrial fibrillation with a CHADS2 score 0–1: A nationwide cohort study (cited 493×)
  22. W2022880742: Should Atrial Fibrillation Patients With 1 Additional Risk Factor of the CHA2DS2-VASc Score (Beyond Sex) Receive Oral Anticoagulation? (cited 260×)
  23. W1985993462: Using the CHA2DS2-VASc Score for Refining Stroke Risk Stratification in ‘Low-Risk’ Asian Patients With Atrial Fibrillation (cited 175×)
  24. W2036575618: CHADS2 and CHA2DS2-VASc Scores in the Prediction of Clinical Outcomes in Patients With Atrial Fibrillation After Catheter Ablation (cited 162×)
  25. W1969247937: Chronic atrial fibrillation: Incidence, prevalence, and prediction of stroke using the Congestive heart failure, Hypertension, Age >75, Diabetes mellitus, and prior Stroke or transient ischemic attack (CHADS2) risk stratification scheme (cited 211×)
  26. W1968635578: Impact of the CHA2DS2-VASc Score on Anticoagulation Recommendations for Atrial Fibrillation (cited 130×)
  27. W4244894279: Comparison of 12 Risk Stratification Schemes to Predict Stroke in Patients With Nonvalvular Atrial Fibrillation (cited 241×)
  28. W2154328340: A New Risk Scheme to Predict Ischemic Stroke and Other Thromboembolism in Atrial Fibrillation: The ATRIA Study Stroke Risk Score (cited 412×)
  29. W4243106582: Predictors of Thromboembolism in Atrial Fibrillation: II. Echocardiographic Features of Patients at Risk (cited 544×)
  30. W4246465322: Predictors of Thromboembolism in Atrial Fibrillation: I. Clinical Features of Patients at Risk (cited 479×)
  31. W4236773584: Transesophageal Echocardiographic Correlates of Thromboembolism in High-Risk Patients with Nonvalvular Atrial Fibrillation (cited 516×)
  32. W2019140613: Additive Role of Plasma von Willebrand Factor Levels to Clinical Factors for Risk Stratification of Patients With Atrial Fibrillation (cited 156×)
  33. W2096867256: Stroke and thromboembolism in atrial fibrillation: A systematic review of stroke risk factors, risk stratification schema and cost effectiveness data (cited 371×)
  34. W1992944164: A Risk Score for Predicting Stroke or Death in Individuals With New-Onset Atrial Fibrillation in the Community (cited 782×)
  35. W2097173178: Simple Risk Model Predicts Incidence of Atrial Fibrillation in a Racially and Geographically Diverse Population: the CHARGE‐AF Consortium (cited 896×)
  36. W3004050106: 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS (cited 6,484×)
  37. W2164861810: Guidelines for the management of atrial fibrillation (cited 5,887×)
  38. W2023143117: 2012 focused update of the ESC Guidelines for the management of atrial fibrillation (cited 3,834×)
  39. W1485439346: Assessing Arrhythmia Burden After Catheter Ablation of Atrial Fibrillation Using an Implantable Loop Recorder: The ABACUS Study (cited 107×)
  40. W2020061893: Screening versus routine practice in detection of atrial fibrillation in patients aged 65 or over: cluster randomised controlled trial (cited 410×)
  41. W2055181085: Stepwise Screening of Atrial Fibrillation in a 75-Year-Old Population (cited 234×)
  42. W1488841895: Screening to identify unknown atrial fibrillation. A systematic review. (cited 192×)
  43. W2021104859: Screening for atrial fibrillation in primary care (cited 105×)
  44. W2987985090: Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation (cited 1,927×)
  45. W2506102335: Diagnostic Performance of a Smartphone‐Based Photoplethysmographic Application for Atrial Fibrillation Screening in a Primary Care Setting (cited 245×)
  46. W3144735448: Smart Wearables for Cardiac Monitoring—Real-World Use beyond Atrial Fibrillation (cited 129×)
  47. W2154658516: Subclinical Atrial Fibrillation and the Risk of Stroke (cited 2,107×)
  48. W2145610874: Device-detected atrial fibrillation and risk for stroke: an analysis of >10 000 patients from the SOS AF project (Stroke preventiOn Strategies based on Atrial Fibrillation information from implanted devices) (cited 428×)
  49. W2782760865: Device-detected subclinical atrial tachyarrhythmias: definition, implications and management—an European Heart Rhythm Association (EHRA) consensus document, endorsed by Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS) and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología (SOLEACE) (cited 246×)
  50. W2024497977: Temporal Relationship Between Subclinical Atrial Fibrillation and Embolic Events (cited 675×)
  51. W4388609933: Apixaban for Stroke Prevention in Subclinical Atrial Fibrillation (cited 435×)
  52. W2609045359: Rationale and design of the Apixaban for the Reduction of Thrombo-Embolism in Patients With Device-Detected Sub-Clinical Atrial Fibrillation (ARTESiA) trial (cited 293×)
  53. W2060379181: Structural correlate of atrial fibrillation in human patients (cited 508×)
  54. W2026978136: Atrial Extracellular Matrix Remodeling and the Maintenance of Atrial Fibrillation (cited 355×)
  55. W2112505092: Atrial fibrillation in acute myocardial infarction: a systematic review of the incidence, clinical features and prognostic implications (cited 616×)
  56. W1982821721: Effects of Biatrial Pacing in Prevention of Postoperative Atrial Fibrillation After Coronary Artery Bypass Surgery (cited 165×)
  57. W2154224888: Assessment of results of surgical treatment for persistent atrial fibrillation during coronary artery bypass grafting using implantable loop recorders (cited 56×)
  58. W2127985968: A Novel User-Friendly Score (HAS-BLED) To Assess 1-Year Risk of Major Bleeding in Patients With Atrial Fibrillation (cited 4,534×)
  59. W2152584205: Comparative Validation of a Novel Risk Score for Predicting Bleeding Risk in Anticoagulated Patients With Atrial Fibrillation (cited 1,088×)
  60. W2004603572: Factors affecting bleeding risk during anticoagulant therapy in patients with atrial fibrillation: Observations from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Study (cited 219×)
  61. W1999468381: Meta-analysis: Antithrombotic Therapy to Prevent Stroke in Patients Who Have Nonvalvular Atrial Fibrillation (cited 4,828×)
  62. W2313681591: Warfarin in the Prevention of Stroke Associated with Nonrheumatic Atrial Fibrillation (cited 1,169×)
  63. W2465663631: The Effect of Low-Dose Warfarin on the Risk of Stroke in Patients with Nonrheumatic Atrial Fibrillation (cited 1,543×)
  64. W2040452288: PLACEBO-CONTROLLED, RANDOMISED TRIAL OF WARFARIN AND ASPIRIN FOR PREVENTION OF THROMBOEMBOLIC COMPLICATIONS IN CHRONIC ATRIAL FIBRILLATION (cited 1,833×)
  65. W2550633476: Design of a multicenter randomized trial for the Stroke Prevention in Atrial Fibrillation Study. The Stroke Prevention in Atrial Fibrillation Investigators. (cited 62×)
  66. W2159730165: Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial (cited 1,574×)
  67. W2159193915: Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the Atrial fibrillation Clopidogrel Trial with Irbesartan for prevention of Vascular Events (ACTIVE W): a randomised controlled trial (cited 1,967×)
  68. W1979862336: Mutations in VKORC1 cause warfarin resistance and multiple coagulation factor deficiency type 2 (cited 1,133×)
  69. W2043674997: Effect of VKORC1 Haplotypes on Transcriptional Regulation and Warfarin Dose (cited 1,398×)
  70. W2151774137: Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications (cited 1,248×)
  71. W1515782956: Estimation of the Warfarin Dose with Clinical and Pharmacogenetic Data (cited 1,587×)
  72. W2156687558: Systematic Overview of Warfarin and Its Drug and Food Interactions (cited 1,168×)
  73. W2006587134: Bleeding in outpatients treated with warfarin: relation to the prothrombin time and important remediable lesions (cited 209×)
  74. W2007303097: Major bleeding in outpatients treated with warfarin: incidence and prediction by factors known at the start of outpatient therapy (cited 767×)
  75. W2113446992: Dabigatran etexilate – a novel, reversible, oral direct thrombin inhibitor: Interpretation of coagulation assays and reversal of anticoagulant activity (cited 1,381×)
  76. W2136489990: Dabigatran versus Warfarin in Patients with Atrial Fibrillation (cited 11,257×)
  77. W2097854437: Rivaroxaban versus Warfarin in Nonvalvular Atrial Fibrillation (cited 9,459×)
  78. W2162508946: Apixaban versus Warfarin in Patients with Atrial Fibrillation (cited 8,988×)
  79. W2096246254: Edoxaban versus Warfarin in Patients with Atrial Fibrillation (cited 5,175×)
  80. W2152623033: Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials (cited 5,048×)
  81. W2113021651: Risks for Stroke, Bleeding, and Death in Patients With Atrial Fibrillation Receiving Dabigatran or Warfarin in Relation to the CHADS2 Score: A Subgroup Analysis of the RE-LY Trial (cited 201×)
  82. W2792857103: The 2018 European Heart Rhythm Association Practical Guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation (cited 2,152×)
  83. W2101863341: Management and Outcomes of Major Bleeding During Treatment With Dabigatran or Warfarin (cited 278×)
  84. W1513839038: Idarucizumab for Dabigatran Reversal (cited 1,465×)
  85. W1641613989: Idarucizumab for dabigatran reversal (cited 1,169×)
  86. W2481569543: Reversal of Rivaroxaban and Dabigatran by Prothrombin Complex Concentrate (cited 1,363×)
  87. W2130340163: Effect of extremes of body weight on the pharmacokinetics, pharmacodynamics, safety and tolerability of apixaban in healthy subjects (cited 272×)
  88. W2074947904: Body Weight Has Limited Influence on the Safety, Tolerability, Pharmacokinetics, or Pharmacodynamics of Rivaroxaban (BAY 59‐7939) in Healthy Subjects (cited 334×)
  89. W2016875529: Dabigatran versus Warfarin in Patients with Mechanical Heart Valves (cited 1,511×)
  90. W2149388786: Apixaban in Patients with Atrial Fibrillation (cited 2,431×)
  91. W2117244124: Percutaneous Left Atrial Appendage Transcatheter Occlusion to Prevent Stroke in High-Risk Patients With Atrial Fibrillation (cited 435×)
  92. W2112970567: Percutaneous Left Atrial Appendage Closure vs Warfarin for Atrial Fibrillation (cited 1,005×)
  93. W50472604: Prospective Randomized Evaluation of the Watchman Left Atrial Appendage Closure Device in Patients With Atrial Fibrillation Versus Long-Term Warfarin Therapy (cited 1,841×)
  94. W2127884497: Safety of Percutaneous Left Atrial Appendage Closure (cited 862×)
  95. W2278350753: Implant success and safety of left atrial appendage closure with the WATCHMAN device: peri-procedural outcomes from the EWOLUTION registry (cited 525×)
  96. W2115996784: Left Atrial Appendage Closure With the Watchman Device in Patients With a Contraindication for Oral Anticoagulation (cited 725×)
  97. W2621231439: Efficacy and safety of left atrial appendage closure with WATCHMAN in patients with or without contraindication to oral anticoagulation: 1-Year follow-up outcome data of the EWOLUTION trial (cited 439×)
  98. W3035828091: Left Atrial Appendage Closure Versus Direct Oral Anticoagulants in High-Risk Patients With Atrial Fibrillation (cited 638×)
  99. W620992480: Left Atrial Appendage Closure as an Alternative to Warfarin for Stroke Prevention in Atrial Fibrillation (cited 548×)
  100. W132631509: Left atrial appendage occlusion for stroke prevention in atrial fibrillation: multicentre experience with the AMPLATZER Cardiac Plug (cited 547×)
  101. W2102464096: Percutaneous Left Atrial Appendage Closure for Stroke Prophylaxis in Patients With Atrial Fibrillation (cited 688×)
  102. W2165920611: Left atrial appendage closure with amplatzer cardiac plug in atrial fibrillation: Initial european experience (cited 425×)
  103. W95963230: Comparison of imaging techniques to assess appendage anatomy and measurements for left atrial appendage closure device selection. (cited 38×)
  104. W1964115470: Left Atrial Appendage Studied by Computed Tomography to Help Planning for Appendage Closure Device Placement (cited 388×)
  105. W2160613225: Detection of Left Atrial Appendage Thrombus by Cardiac Computed Tomography in Patients With Atrial Fibrillation (cited 378×)
  106. W3161524032: Left Atrial Appendage Occlusion during Cardiac Surgery to Prevent Stroke (cited 718×)
  107. W2110253: A Comparison of Rate Control and Rhythm Control in Patients with Atrial Fibrillation (cited 4,432×)
  108. W1994231358: A Comparison of Rate Control and Rhythm Control in Patients with Recurrent Persistent Atrial Fibrillation (cited 2,372×)
  109. W2149647480: A randomised, controlled study of rate versus rhythm control in patients with chronic atrial fibrillation and heart failure: (CAFÉ-II Study) (cited 178×)
  110. W2123571173: Rate control versus rhythm control for patients with persistent atrial fibrillation with mild to moderate heart failure: Results from the RAte Control versus Electrical cardioversion (RACE) study (cited 190×)
  111. W2129290969: Impact of rate versus rhythm control on quality of life in patients with persistent atrial fibrillationResults from a prospective randomized study (cited 186×)
  112. W1989801282: Rate vs Rhythm Control in Patients With Atrial Fibrillation (cited 239×)
  113. W2115580917: Analysis of Cause-Specific Mortality in the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) Study (cited 230×)
  114. W2128106558: Occurrence and Characteristics of Stroke Events in the Atrial Fibrillation Follow-up Investigation of Sinus Rhythm Management (AFFIRM) Study (cited 234×)
  115. W1207268963: Relationships Between Sinus Rhythm, Treatment, and Survival in the Atrial Fibrillation Follow-Up Investigation of Rhythm Management (AFFIRM) Study (cited 1,223×)
  116. W2102129403: Lenient versus Strict Rate Control in Patients with Atrial Fibrillation (cited 1,034×)
  117. W3082794714: Early Rhythm-Control Therapy in Patients with Atrial Fibrillation (cited 2,153×)
  118. W3194283352: Systematic, early rhythm control strategy for atrial fibrillation in patients with or without symptoms: the EAST-AFNET 4 trial (cited 189×)
  119. W3187003271: Early Rhythm Control Therapy in Patients With Atrial Fibrillation and Heart Failure (cited 245×)
  120. W2313203550: Spontaneous Initiation of Atrial Fibrillation by Ectopic Beats Originating in the Pulmonary Veins (cited 8,034×)
  121. W2088825357: Electrophysiological End Point for Catheter Ablation of Atrial Fibrillation Initiated From Multiple Pulmonary Venous Foci (cited 929×)
  122. W2122843638: Pulmonary Vein Isolation for Paroxysmal and Persistent Atrial Fibrillation (cited 1,123×)
  123. W2172234567: Catheter Ablation for Paroxysmal Atrial Fibrillation (cited 901×)
  124. W1583406651: Approaches to Catheter Ablation for Persistent Atrial Fibrillation (cited 2,330×)
  125. W2033170044: Catheter Ablation of Long‐Lasting Persistent Atrial Fibrillation: Clinical Outcome and Mechanisms of Subsequent Arrhythmias (cited 759×)
  126. W2109650013: Catheter Ablation of Paroxysmal Atrial Fibrillation Initiated by Non–Pulmonary Vein Ectopy (cited 756×)
  127. W2150866578: Association of Atrial Tissue Fibrosis Identified by Delayed Enhancement MRI and Atrial Fibrillation Catheter Ablation (cited 1,436×)
  128. W2130527455: Spatial Distribution of Fibrosis Governs Fibrillation Wave Dynamics in the Posterior Left Atrium During Heart Failure (cited 321×)
  129. W4234891227: Worldwide Survey on the Methods, Efficacy, and Safety of Catheter Ablation for Human Atrial Fibrillation (cited 1,322×)
  130. W2146673314: Updated Worldwide Survey on the Methods, Efficacy, and Safety of Catheter Ablation for Human Atrial Fibrillation (cited 2,872×)
  131. W2154624542: Atrio-Esophageal Fistula as a Complication of Percutaneous Transcatheter Ablation of Atrial Fibrillation (cited 880×)
  132. W2080068325: Right Phrenic Nerve Injury Following Electrical Disconnection of the Right Superior Pulmonary Vein (cited 55×)
  133. W1481100471: Recurrent dyspnea following multiple ablations for atrial fibrillation explained by the “stiff left atrial syndrome” (cited 24×)
  134. W2038456369: Comparison of Antiarrhythmic Drug Therapy and Radiofrequency Catheter Ablation in Patients With Paroxysmal Atrial Fibrillation (cited 1,177×)
  135. W2133730204: Catheter Ablation Versus Antiarrhythmic Drugs for Atrial Fibrillation (cited 896×)
  136. W2154269490: Radiofrequency Ablation vs Antiarrhythmic Drugs as First-line Treatment of Symptomatic Atrial Fibrillation (cited 1,129×)
  137. W2165779304: Long‐term Outcomes of Catheter Ablation of Atrial Fibrillation: A Systematic Review and Meta‐analysis (cited 797×)
  138. W2164489173: Radiofrequency Ablation as Initial Therapy in Paroxysmal Atrial Fibrillation (cited 725×)
  139. W2920804236: Effect of Catheter Ablation vs Antiarrhythmic Drug Therapy on Mortality, Stroke, Bleeding, and Cardiac Arrest Among Patients With Atrial Fibrillation (cited 1,501×)
  140. W2786781710: Catheter Ablation for Atrial Fibrillation with Heart Failure (cited 2,494×)
  141. W2348720246: Catheter Ablation for Atrial Fibrillation in Congestive Heart Failure (cited 894×)
  142. W2333399234: Ablation Versus Amiodarone for Treatment of Persistent Atrial Fibrillation in Patients With Congestive Heart Failure and an Implanted Device (cited 867×)
  143. W1995961708: Reversal of Left Ventricular Dysfunction Following Ablation of Atrial Fibrillation (cited 276×)
  144. W2131260805: Rhythm Control versus Rate Control for Atrial Fibrillation and Heart Failure (cited 1,587×)
  145. W1979232262: Efficacy of amiodarone for the termination of persistent atrial fibrillation (cited 98×)
  146. W2069805683: Intravenous Amiodarone for Acute Heart Rate Control in the Critically Ill Patient With Atrial Tachyarrhythmias (cited 215×)
  147. W2060841225: Antiarrhythmic Actions of Intravenous Ibutilide Compared With Procainamide During Human Atrial Flutter and Fibrillation (cited 159×)
  148. W2128605957: Lack of evidence of increased mortality among patients with atrial fibrillation taking digoxin: findings from post hoc propensity-matched analysis of the AFFIRM trial (cited 192×)
  149. W2614763070: Use of Irbesartan to Maintain Sinus Rhythm in Patients With Long-Lasting Persistent Atrial Fibrillation (cited 690×)
  150. W2070350109: Obesity and the Risk of New-Onset Atrial Fibrillation (cited 1,410×)
  151. W2116664693: Obesity as a risk factor for the progression of paroxysmal to permanent atrial fibrillation: a longitudinal cohort study of 21 years (cited 306×)
  152. W2176626491: The role of obesity in atrial fibrillation (cited 237×)
  153. W1963940525: Effect of Weight Reduction and Cardiometabolic Risk Factor Management on Symptom Burden and Severity in Patients With Atrial Fibrillation (cited 775×)
  154. W2086418097: Alcohol Consumption and Risk of Atrial Fibrillation in Men and Women (cited 268×)
  155. W2997250629: Alcohol Abstinence in Drinkers with Atrial Fibrillation (cited 435×)
  156. W2166034725: Obstructive Sleep Apnea and the Recurrence of Atrial Fibrillation (cited 984×)
  157. W2118251699: Safety and Efficacy of Pulmonary Vein Antral Isolation in Patients With Obstructive Sleep Apnea (cited 236×)
  158. W2065996377: Impact of obstructive sleep apnea and continuous positive airway pressure therapy on outcomes in patients with atrial fibrillation—Results from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) (cited 247×)
  159. W2100509598: Renal Sympathetic Denervation Suppresses Postapneic Blood Pressure Rises and Atrial Fibrillation in a Model for Sleep Apnea (cited 227×)
  160. W2004798125: Atrial Fibrillation and Cognitive Decline–The Role of Subclinical Cerebral Infarcts (cited 129×)
  161. W4210409146: Atrial Fibrillation and Dementia: A Report From the AF-SCREEN International Collaboration (cited 203×)
  162. W2165211571: Atrial Fibrillation Is Associated With Increased Spontaneous Calcium Release From the Sarcoplasmic Reticulum in Human Atrial Myocytes (cited 355×)
  163. W2034282563: Chronic Rapid Atrial Pacing (cited 1,045×)
  164. W2749929702: Dual Antithrombotic Therapy with Dabigatran after PCI in Atrial Fibrillation (cited 1,375×)
  165. W2159370853: Use of clopidogrel with or without aspirin in patients taking oral anticoagulant therapy and undergoing percutaneous coronary intervention: an open-label, randomised, controlled trial (cited 1,718×)