Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation

INTRODUCTION — Spontaneous or intended conversion of atrial fibrillation (AF) to sinus rhythm (SR) is associated with a short-term increase from the baseline risk of clinical thromboembolism due to atrial mechanical stunning. This topic will discuss management strategies that attempt to decrease this thromboembolic risk, based on the duration of the AF episode, prior anticoagulant therapy, and the patient’s individualized risk of stroke (CHA₂DS₂-VASc score (table 1)).

The modalities used to perform cardioversion, long-term anticoagulation in patients with AF, and an overview of the management of AF are presented separately. (See "Atrial fibrillation: Cardioversion" and "Atrial fibrillation in adults: Selection of candidates for anticoagulation" and "Atrial fibrillation in adults: Use of oral anticoagulants" and "Atrial fibrillation: Overview and management of new-onset atrial fibrillation".)

EXTREMELY HIGH-RISK PATIENTS — Patients with AF and certain types of valvular heart disease (rheumatic mitral stenosis or a mechanical valve) are at extremely high risk of thromboembolic complications at all times, not only at the time of cardioversion. The approach to antithrombotic therapy in such patients is discussed in other UpToDate topics. (See "Rheumatic mitral stenosis: Overview of management", section on 'Prevention of thromboembolism' and "Antithrombotic therapy for mechanical heart valves".)

RATIONALE FOR ANTICOAGULATION — All patients with AF, whether paroxysmal, persistent, or permanent, have an increased risk of thromboembolism compared with those without AF. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

At the time of reversion to SR, whether pharmaceutical, electrical, or spontaneous, there is a transient incremental increase from the baseline risk due to more depressed atrial mechanical function. Most embolic events occur within 10 days of reversion to SR [1-5]. Patients undergoing cardioversion of AF of more than 48 hours duration represent a particularly high-risk group (compared with AF of less than 48 hours duration), with an embolic risk from as low as 1 to as high as 5 percent in the first month after reversion to SR in the absence of anticoagulation [2-4,6-8]. This rate is substantially higher than the rate that would be calculated for the general population of patients with AF, in whom the yearly rate is between 1.3 and 5.1 (or higher) percent, depending on age and additional comorbidities.

The most common source of stroke associated with cardioversion in these patients is migration of a thrombus from the left atrial appendage during or in the first 10 days after the procedure. Possible causes include embolism of a left atrial thrombus that was already present at the time of conversion to SR, embolism of a thrombus that formed after conversion due to depressed left atrial appendage ejection velocity postconversion, or delay in recovery of left atrial mechanical function after conversion, and thrombus formation during subsequent episodes of AF:

●Precardioversion left atrial thrombus. Embolization after return of synchronous atrial contraction is due to the dislodgement of left atrial thrombi present at the time of cardioversion. This is felt to be the dominant cause of postcardioversion thromboembolism and the rationale for performing transesophageal echocardiogram (TEE) prior to cardioversion to exclude an appendage thrombus prior to attempted cardioversion.

The prevalence of left atrial appendage thrombus in nonanticoagulated patients with AF of less than 72 hours undergoing TEE is 12 and 14 percent [9,10]. This value is similar to that found among AF patients with a duration of unknown or more than two days duration [11,12].

The prevalence of left atrial appendage thrombus is increased in high-risk patients with prior or acute thromboembolism, severe left ventricular systolic dysfunction, left atrial enlargement, depressed left atrial appendage ejection velocity, or left atrial appendage spontaneous echo contrast (a marker of blood stasis).

●Postcardioversion atrial mechanical dysfunction creates a milieu that promotes new (postcardioversion) thrombus formation. The transient atrial contractile dysfunction after cardioversion is referred to as atrial "stunning" and can occur whether SR is restored spontaneously, by external or internal direct current cardioversion, or by antiarrhythmic medications. The duration of the left atrial contractile dysfunction appears to be related in part to the duration of AF prior to cardioversion. Recovery of atrial mechanical function may be delayed for several weeks [13] for those who have been in AF for a few months prior to cardioversion. In comparison, for those with AF for only a few days, left atrial mechanical recovery occurs within a day (but may still be associated with more pronounced but transient dysfunction immediately after cardioversion). (See "Hemodynamic consequences of atrial fibrillation and cardioversion to sinus rhythm", section on 'Atrial stunning'.)

In support of the atrial stunning after cardioversion hypothesis, there have been case reports and small series of patients developing TEE evidence for de novo left atrial appendage thrombi (primarily in the setting of no anticoagulation) immediately following cardioversion, when the precardioversion TEE showed no left atrial appendage thrombus [9,14-16]. (See "Role of echocardiography in atrial fibrillation", section on 'Spontaneous echo contrast' and "Mechanisms of thrombogenesis in atrial fibrillation".)

●Recurrent AF is common during the first month after conversion [17]. Up to 90 percent of these episodes are asymptomatic [18], and asymptomatic episodes lasting more than 48 hours are not uncommon, occurring in 17 percent of patients in a report using continuous monitoring [17]. Anticoagulation during the four weeks postcardioversion (independent of CHA2DS2-VAs score used for guiding long-term anticoagulation in nonvalvular AF) thereby provides prophylaxis against new thrombus formation and facilitates early cardioversion without a screening TEE should recurrent AF occur.

The rationale and indications for chronic anticoagulation after the period of postconversion anticoagulation are similar to those for the broad population of patients with AF and are discussed separately. (See "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Summary and recommendations'.)

PATIENTS WITH SPONTANEOUS CONVERSION — Some patients with AF have spontaneous conversion prior to planned cardioversion. The risk of thromboembolism after spontaneous conversion or electrical cardioversion is relatively low, but the risk during this time is likely higher than the ambient rate of thromboembolic events associated with AF. There is no evidence that risk of embolization in the first few weeks after spontaneous conversion differs from that for patients with AF undergoing electrical or chemical cardioversion. In a study of 1041 patients who were anticoagulated prior to and after cardioversion, 16 percent experienced spontaneous conversion (prior to planned electrical cardioversion) [19]. The rate of thromboembolism was similar in patients with spontaneous conversion compared with patients who underwent electrical cardioversion (<1 percent in both groups) although this comparison is limited by the small number of events).

Though of unproven efficacy, some of our contributors recommend anticoagulation for four weeks after reversion to SR (either spontaneous or via cardioversion) for patients with AF of less than 48 hours duration, even for those with a low CHA₂DS₂-VASc score (table 1). The rationale for this approach is concern regarding the high likelihood of AF recurrence in the first month after reversion to SR, as well as transient postcardioversion atrial stunning in the immediate pericardioversion period. This approach may be modified in patients at very high bleeding risk. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

Management of long-term anticoagulation (after the initial four weeks) including the role of CHA₂DS₂-VASc score is discussed separately. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation" and "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Approach to anticoagulation'.)

URGENT CARDIOVERSION — Patients with new onset AF in whom the ventricular rate is rapid may require urgent (or emergent) cardioversion to prevent adverse clinical consequences such as hemodynamic decompensation. (See "Atrial fibrillation: Overview and management of new-onset atrial fibrillation", section on 'Symptom and hemodynamic management'.)

The indications for urgent cardioversion of AF are uncommon, but in the setting of hemodynamic instability due to rapid AF that is refractory to pharmacologic support, such as in patients with Wolff-Parkinson-White syndrome, the need for restoration of SR may take precedence over the need for protection from thromboembolism. When possible, the patient should receive precardioversion anticoagulation (eg, bolus of unfractionated heparin) as soon as possible due to the risk of postcardioversion left atrial appendage stunning. Anticoagulation should be considered for four weeks postcardioversion, unless it is contraindicated [20] (see 'AF duration less than 48 hours' below).

Management of long-term anticoagulation (after the initial four weeks) is discussed separately. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation" and "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Approach to anticoagulation'.)

AF DURATION LESS THAN 48 HOURS

Anticoagulation prior to cardioversion — If conversion to SR (either spontaneous or via cardioversion) occurs within 48 hours of the onset of AF, the thromboembolic risk appears to be very low [21-23]. However, many, and perhaps most, patients cannot accurately define the onset of AF. As a result, we categorize a patient as having AF of less than 48 hours duration only if we have a high level of confidence in the patient’s history. Otherwise, we approach the patient as if AF has been present for more than 48 hours. (See 'AF duration uncertain or 48 or more hours' below.)

For most patients in whom cardioversion will take place less than 48 hours after the onset of AF, we start a DOAC prior to cardioversion rather than no anticoagulant. Intravenous heparin is a reasonable alternative for hospitalized patients. When a DOAC is used, the specific choice of DOAC should be individualized for each patient. We generally choose the agent that will be given at the time of discharge. Of note, the approach presented here is in contrast to the historical approach of some cardiologists proceeding to early cardioversion without anticoagulation if the duration was less than 24 hours. I(See "Atrial fibrillation in adults: Use of oral anticoagulants".)

We generally wait at least two to four hours after the first dose of a DOAC to cardiovert. For patients at very high bleeding risk, some of our experts suggest cardioversion without anticoagulation if normal SR can be restored within 48 hours of documented onset. Other experts recommend short-term anticoagulation prior to cardioversion even in these high-bleeding-risk patients.

If cardioversion needs to take place within three hours, whether for patient instability or convenience (see 'Urgent cardioversion' above), we start intravenous unfractionated heparin (bolus and continuous drip goal partial thromboplastin time 1.5 to 2.0 times control) or a low molecular weight heparin (1 mg/kg subcutaneously every 12 hours); we do not give DOAC and heparin together. However, if warfarin is the agent selected for longer term anticoagulation, warfarin is started while heparin therapy is continued until the international normalized ratio exceeds 2.0.

For extremely high-risk patients (eg, those with rheumatic mitral stenosis, mechanical valves, prior thromboembolism, severe left ventricular dysfunction, heart failure, or diabetes), we anticoagulate for at least three weeks or initiate therapeutic anticoagulation (with heparin or DOAC) in combination with TEE prior to an attempt at cardioversion as described above for AF of more than 48 hours duration. (See 'AF duration uncertain or 48 or more hours' below.)

The 48 hour cut point is based on limited evidence and is somewhat arbitrary. For example, the prevalence of left atrial thrombus on TEE is substantially lower when the duration of AF is less than 48 hours (1.4 percent) [24]. Among patients with a history of AF of less than 48 hours in duration, there is likely a range of risk based on CHA₂DS₂-VASc score (table 1). A retrospective study of 3143 patients with AF of less than 48 hours duration demonstrated that patients with heart failure and diabetes were at high risk for clinical thromboembolism (up to 10 percent if both risk factors were present). The absence of both risk factors and age <60 years conveyed a very low risk of 0.2 percent [23]. (See 'AF duration uncertain or 48 or more hours' below and 'Rationale for anticoagulation' above.)

No randomized trial has evaluated anticoagulation compared with no anticoagulation in AF patients undergoing cardioversion with a definite duration of AF <48 hours. Observational data suggest that the risk of stroke/thromboembolism is very low (0  to 0.2 percent) in patients with a definite AF duration of <12 hours and a very low stroke risk (CHA₂DS₂-VASc 0 in men, 1 in women), in whom the benefit of four-week anticoagulation after cardioversion is undefined. The 2020 European Society of Cardiology guidelines for the diagnosis and management of AF suggest that prescription of anticoagulants can be optional, based on an individualized approach [25].

With regard to the question as to whether to anticoagulate these patients or not, there are no studies comparing heparin with no heparin in patients with AF of less than 48 hours duration. However, data regarding the rate of clinical thromboembolization after cardioversion in patients with AF of less than 48 hours duration have raised a concern about the safety of cardioversion without anticoagulation in this population. In an observational study of 2481 such individuals (5116 successful cardioversions) who were not treated with peri- or postprocedural anticoagulant, definite thromboembolic events occurred in 38 (0.7 percent) within 30 days (median of two days); of these, 31 were strokes [23]. Four additional patients suffered a transient ischemic attack. Age greater than 60 years, female sex, heart failure, and diabetes were the strongest predictors of embolization, with nearly 10 percent of those with both heart failure and diabetes experiencing a stroke. The risk of stroke in those without heart failure and age less than 60 years was 0.2 percent. An observational study of 16,274 patients undergoing electrical cardioversion with and without oral anticoagulant therapy also demonstrated that the absence of postcardioversion anticoagulation was associated with a high risk of thromboembolism, regardless of CHA₂DS₂-VASc scores [26]. There was a greater-than-twofold increased risk of thromboembolism in those not treated with postcardioversion anticoagulation (hazard ratio 2.21; 95% CI 0.79-6.77 and 2.40; 95% CI 1.46-3.95 with CHA₂DS₂-VASc score 0 to 1 and CHA₂DS₂-VASc score 2 or more, respectively). The rationale for lack of postcardioversion anticoagulation could not be exactly discerned in this trial but was deemed to be multifactorial, including presumed short-duration AF, perceived low thromboembolic risk, and lack of guideline adherence.

With regard to the question of which anticoagulant to use, there are no studies comparing differing forms of heparin in patients with AF of short duration nor are there studies comparing a DOAC with heparin. Indirect evidence comparing the two heparins comes from a trial of 496 patients with AF of more than 48 hours duration who were randomly assigned to either low molecular weight heparin or unfractionated heparin followed by oral anticoagulation [27]. Patients were cardioverted after either 21 days of anticoagulation or after a TEE that was negative for thrombus; anticoagulation continued for 28 days after cardioversion. Low molecular weight heparin was noninferior to unfractionated heparin followed by oral anticoagulation in terms of the combined primary end point of ischemic neurologic events, major hemorrhage, or death by the end of study treatment (2.8 versus 4.8 percent). Low molecular weight heparin also has a safety and efficacy profile similar to unfractionated heparin when used as a bridge to oral anticoagulation in patients undergoing TEE-based therapy [28].

Anticoagulation after reversion to sinus rhythm — Though of unproven in efficacy, some of our contributors recommend anticoagulation for four weeks after reversion to SR (either spontaneous or intended) for patients with AF of less than 48 hours duration, even for those with a low CHA₂DS₂-VASc score. The rationale for this approach is a concern regarding the high likelihood of AF recurrence in the first month after reversion to SR, as well as transient postcardioversion atrial stunning in the immediate pericardioversion period. This decision may be modified in patients at very high bleeding risk.

Some of our contributors do not anticoagulate patients with a low CHA₂DS₂-VASc score (0 in men or 1 in women) after restoration of SR if AF was less than 48 hours duration [23,29].

Management of long-term anticoagulation (after the initial four weeks) is discussed separately. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation" and "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Approach to anticoagulation'.)

AF DURATION UNCERTAIN OR 48 OR MORE HOURS — Patients with AF of more than 48 hours or of unknown duration should receive at least four weeks of therapeutic anticoagulation prior to cardioversion and four weeks of anticoagulation after cardioversion. In this setting, this treatment regimen can reduce the risk of thromboembolism during the four weeks after cardioversion from 6 percent to less than 1 percent [2-4,6,7,30-32].

For patients in whom there is a reason to not wait four weeks, an option for management is precardioversion therapeutic anticoagulation in conjunction with a screening TEE to guide early cardioversion. This strategy can be used for patients in whom cardioversion needs to be performed before at least four weeks of therapeutic anticoagulation have been completed [12]. While the TEE approach shortens the precardioversion duration of anticoagulation, it does not change our recommendation for four weeks of anticoagulation after cardioversion or the need to be therapeutically anticoagulated at the time of the cardioversion due to the risk associated with postcardioversion atrial appendage stunning. (See 'Transesophageal echocardiography-based approach' below.)

Prospective studies have shown that the risk of clinical stroke or systemic embolism ranges from 0 to 0.9 percent if preceded by four weeks of therapeutic anticoagulation with warfarin (target international normalized ratio [INR] 2.0 to 3.0) or one of the DOACs [2-4,12], or shorter-term anticoagulation with TEE-guided approach discussed directly above. Retrospective data demonstrated that the thromboembolism risk is 4 to 7 percent in nonanticoagulated patients [7,8,33].

Anticoagulant approach — Since many patients will require long-term anticoagulation, we prefer the DOACs to warfarin before and after cardioversion (see "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Choice of anticoagulant'). While there has been longer experience with use of warfarin than DOACs prior to cardioversion, we believe there is sufficient evidence that DOACs are as effective and as safe as warfarin in this setting.

Advantages of DOAC therapy include convenience (no INR testing required) and the possibility of a shorter duration of precardioversion anticoagulation in reliably adherent patients, since it often takes five or more weeks for a patient to have at least three continuous weeks of therapeutic anticoagulation with warfarin (INR 2.0 to 3.0).

In patients in whom adherence to DOAC therapy is questionable, with possible missed doses leading up to the cardioversion, we often obtain precardioversion TEE to exclude an atrial (appendage) thrombus. Routine precardioversion TEE is not recommended for patients who have been therapeutically anticoagulated (INR 2 or greater) with warfarin for three weeks or who have been compliant with their daily DOAC, though data suggest that patients with heart failure and diabetes may be at increased risk for postcardioversion thromboembolism despite adequate-duration anticoagulation. (See 'Transesophageal echocardiography-based approach' below.)

Compliance with warfarin can be ascertained with INR monitoring. For patients started on warfarin, the target INR should be 2.5 (range 2.0 to 3.0), and cardioversion should not take place until an INR of 2.0 or greater has been documented for at least three consecutive weeks (figure 1 and figure 2) [34,35].

The following data are available for the DOACs:

●Dabigatran – In a post-hoc analysis of the RE-LY trial, which compared dabigatran with warfarin, in which there were 1983 cardioversions in 1270 participants, there was no significant difference in the rate of thromboembolism and stroke within 30 days between those who received at least three weeks of dabigatran 110 or 150 mg twice daily or warfarin (0.8, 0.3, and 0.6 percent, respectively) [2]. Little data are available on the use of dabigatran in an early cardioversion strategy, and this agent has a large accumulation over time, so reaching steady-state concentrations (administered at least 48 to 72 hours) prior to cardioversion is recommended.

●Apixaban – In a post-hoc analysis of the ARISTOTLE trial, which compared apixaban with warfarin, 743 cardioversions were performed in 540 patients. No strokes or systemic embolism occurred during the 30-day follow-up period of both warfarin and apixaban groups [3]. Efficacy and safety with cardioversion using apixaban were reconfirmed in the randomized, prospective, open-label EMANATE trial of 1500 anticoagulation-naïve patients. This trial also utilized a loading-dose (10 or 5 mg; a single double-dose, depending on baseline dosing recommendation) strategy in nearly half of the apixaban-treated group (342 out of 753 patients) followed by early cardioversion two hours after the loading dose [36]. (See "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Choice of anticoagulant'.)

●Rivaroxaban – Ina post-hoc analysis of the ROCKET-AF trial, which compared rivaroxaban with warfarin, 143 patients underwent 181 electric cardioversions, 142 patients underwent 194 pharmacologic cardioversions, and 79 patients underwent 85 catheter ablations. There was no significant difference in the long-term rate of stroke or systemic embolism (hazard ratio 1.38; 95% CI 0.61-3.11) [37]. (See "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Choice of anticoagulant'.)

●In the X-VeRT study, 1504 patients with AF of unknown or longer than 48 hours duration were randomly assigned in a 2:1 manner to cardioversion after at least three weeks of rivaroxaban or a vitamin K antagonist. There was no significant difference in the rate of the primary efficacy outcome (a composite of stroke, transient ischemic attack, peripheral embolism, myocardial infarction, and cardiovascular death) or the safety outcome of major bleeding (0.51 versus 1.02 percent and 0.6 percent versus 0.8 percent, respectively) [4]. Rivaroxaban was given at least four hours prior to cardioversion.

●Edoxaban – In the ENSURE-AF trial, 2199 patients were randomly assigned to receive edoxaban or enoxaparin and warfarin with discontinuation of enoxaparin when the INR was >2.0 [5]. There was no significant difference in the primary efficacy end point (0.5 percent in the edoxaban group versus 1 percent in the enoxaparin–warfarin group; odds ratio [OR] 0.46, 95% CI 0.12-1.43). The primary safety end point occurred in 1.6 percent of the edoxaban group versus 1.1 percent in the enoxaparin–warfarin group (OR 1.48, 95% CI 0.64-3.55). The results were independent of the TEE-guided strategy and anticoagulation status. Edoxaban was administered at least two hours prior to cardioversion.

Therapeutic anticoagulation prior to cardioversion appears to be effective largely due to thrombus resolution, rather than organization and adherence of left atrial thrombi [38,39]. (See 'Rationale for anticoagulation' above.)

Transesophageal echocardiography-based approach — We suggest a TEE-based approach (table 2A-B) for symptomatic patients and for patients for whom there is a concern about a three-week (or more) delay to cardioversion. Such a concern might arise from a preference to not have ongoing symptoms of AF or a possible lower likelihood of successful cardioversion with a longer period of AF. Other individuals for whom this strategy may be reasonable include those at high bleeding risk, as the TEE-guided approach shortens the total precardioversion anticoagulation time for those without thrombus; and those at highest risk for a cardioversion-related thromboembolic event, including prior thromboembolism and elderly women with diabetes and heart failure. Patients who require hospitalization are also candidates for this approach [40,41]. This recommendation for a focused use of the TEE-based approach is based on our concerns about cost, the small potential for complications, and the possibility of worse outcomes.

We also recommend precardioversion TEE for all patients with a percutaneous left atrial appendage occlusion device in place (eg, Watchman, Lariat, Amulet,) or who have undergone surgical LAA exclusion (eg, by stapling, suture or approved device closure). Following LAA occlusion, adjacent thrombus may occur (with or without incomplete closure) with associated risk of thromboembolism. (See "Atrial fibrillation: Left atrial appendage occlusion".)

In a multicenter retrospective study of 148 patients with an LAA occlusion device undergoing direct current cardioversion (DCCV), 100 percent of patients received precardioversion TEE. Device-related thrombus was noted in 2.7 percent, and those patients were treated with OAC for six to eight weeks prior to successful thrombus resolution and DCCV. Twenty-two percent of the entire cohort were newly started on OAC postcardioversion. There were no differences in DCCV-related complications between the patients treated with or without OAC post-DCCV [42]. A prospective multicenter registry evaluated 93 patients undergoing DCCV post-LAA occlusion. Two hundred and eighty-four DCCVs were performed with a wide variety of treatment strategies observed. There were no thromboembolic events noted during 30-day follow-up. Precardioversion TEE was only performed in 34 DCCV procedures [43]. Though these small observational data suggest safety, we recommend awaiting further validation with larger randomized prospective trial data. We continue to recommend precardioversion TEE imaging followed by four-week postcardioversion anticoagulation.

In a TEE-based approach, the imaging study is performed after therapeutic anticoagulation (of short duration) and prior to anticipated cardioversion. Patients without evidence of left and right atrial (specifically the left atrial appendage, which is the site for the vast majority of thrombi) thrombus proceed to cardioversion. If thrombus is found (or cannot be confidently excluded) on TEE, cardioversion should not be performed, and therapeutic anticoagulation should be continued for at least four weeks after which time we recommend that a TEE be repeated (to screen for residual thrombus, which would be a contraindication to cardioversion) if cardioversion is desired. Among those with a thrombus on TEE, lifeline anticoagulation should be considered independent of CHA2DS2-VASc score.

The TEE approach should include the following sequential steps before cardioversion:

●For inpatients, the options include using heparin plus warfarin or using an DOAC. With the former, we administer either low molecular weight or unfractionated heparin (bolus and continuous drip with a goal partial thromboplastin time 1.5 to 2 times control) and simultaneously initiate oral warfarin (target INR 2.0 to 3.0). With the latter, we give at least two doses of a DOAC. As the pharmacokinetics of the DOACs are different than warfarin, the combination of a heparin plus DOAC may lead to supratherapeutic anticoagulation. We do not recommend overlap of or combined use of heparin and a DOAC.

●For most outpatients, we prefer DOAC to warfarin. There are multiple factors that determine whether a DOAC or warfarin would be used, including cost and patient preference, but DOACs have the advantage of faster onset of action and ease of dosing. A strategy of at least two days of DOAC prior to TEE-guided cardioversion can be used. As an alternative, oral warfarin can be started five days before TEE with the target INR 2.0 to 3.0 [12]. A minimal precardioversion INR of 2.0 is acceptable, though 2.5 may be preferred.

●Obtain a TEE to assess for the presence of atrial thrombi. The use of an endocardial border definition echo contrast agent may help in cases where there is uncertainty about the presence or absence of thrombus [44].

If no thrombus is seen, proceed with cardioversion. Continue therapeutic anticoagulation from the time of TEE through cardioversion and extend for another four weeks.

If a thrombus is seen on TEE, the patient should receive a minimum of four weeks of therapeutic anticoagulation and a repeat TEE to document thrombus resolution if cardioversion is desired [38]. If no cardioversion is desired, a follow-up TEE is not needed. We recommend the patient receive lifelong antithrombotic therapy regardless of CHA₂DS₂-VASc. If thrombus is absent on repeat TEE, cardioversion may be performed. If thrombus is still evident, the rhythm control strategy may be changed to a rate control strategy, especially when AF-related symptoms are controlled, since there is a high risk of thromboembolism if cardioversion is performed. However, the evidence supporting this latter recommendation of avoidance of cardioversion with a residual thrombus is minimal.

It is best to be conservative with at least four weeks of precardioversion oral anticoagulant if an atrial thrombus cannot be confidently excluded on TEE.

●Continuous oral anticoagulation (warfarin INR 2.0 to 3.0 or full-dose DOAC) for at least four weeks after cardioversion in all eligible patients, regardless of the cardioversion method, CHA₂DS₂-VASc score, or apparent maintenance of SR. In patients who have not achieved therapeutic anticoagulation with warfarin at the time of cardioversion, unfractionated or low molecular weight heparin should be continued until the INR is therapeutic.

Observational studies have suggested that patients with AF of more than 48 hours duration can be acutely anticoagulated with heparin/oral anticoagulant and proceed directly to cardioversion without prolonged anticoagulation if no atrial thrombus is seen on precardioversion TEE (table 2A-B) [11,45-47]. The ACUTE trial compared a TEE-guided strategy with a conventional strategy (including therapeutic warfarin [INR 2.0 to 3.0] anticoagulation for at least three weeks prior to electrical cardioversion) in 1222 patients with AF of more than two days duration (median duration 13 days) who were undergoing electrical cardioversion [12,48]. Patients assigned to the TEE-guided strategy were anticoagulated with heparin before TEE if they were inpatients or with oral warfarin for five days (target INR 2.0 to 3.0) before TEE if they were outpatients. TEE was then followed by cardioversion if no atrial thrombi were identified. With both approaches, warfarin therapy was continued for four weeks after cardioversion. If the initial TEE demonstrated thrombus (which was present in 12 percent), cardioversion was postponed and patients received therapeutic (INR 2.0 to 3.0) anticoagulation for three weeks, at which time a repeat TEE was performed. Patients assigned to conventional strategy received three weeks of therapeutic anticoagulation before cardioversion.

The following findings were noted:

●Within the eight weeks after study enrollment, there was no significant difference between the TEE and conventional groups in the incidence of ischemic stroke (0.6 versus 0.3 percent, respectively; relative risk [RR] 1.95, 95% CI 0.36-10.60) or all embolic events, including stroke, transient ischemic attack, and peripheral embolism (0.8 versus 0.5 percent, respectively; RR 1.62, 95% CI 0.39-6.76). One important difference is that the majority of thromboembolic events in the TEE arm occurred in patients who had reverted back to AF and/or had a subtherapeutic INR at the time of the event, while the thromboembolic events in the warfarin arm occurred in patients with SR with a therapeutic INR.

●There were significantly fewer hemorrhagic events with the TEE strategy (2.9 versus 5.5 percent), but no significant difference in the incidence of major bleeding (0.8 versus 1.5 percent) [12,49]; in addition, there was no significant difference in all-cause mortality (2.4 versus 1 percent) or cardiac deaths (1.3 versus 0.7).

●The TEE strategy led to a shorter mean time to cardioversion (3 versus 31 days) and a greater incidence of successful restoration of SR (71 versus 65 percent).

Thromboembolism has been reported after a negative precardioversion TEE in some patients who were not therapeutically anticoagulated at the time of TEE with anticoagulation continuing for one month after cardioversion [9,14,15]. These adverse events may be related to the limited sensitivity of TEE for small thrombi, or to new thrombus formation that has been reported during the period between TEE and cardioversion or after cardioversion [9,14,15]. Thus, we recommend therapeutic anticoagulation for all patients undergoing a TEE-based approach to cardioversion.

The development of impaired left atrial mechanical function and of new thrombi after cardioversion provides the rationale for four weeks of therapeutic anticoagulation after cardioversion (INR 2.0 to 3.0 or daily DOAC), even when the precardioversion TEE shows no thrombus [15,16]. There is suggestive evidence that such an approach reduces the incidence of embolic events [16]. (See 'Rationale for anticoagulation' above.)

Although the results of the ACUTE study discussed above raise concerns about possible worse outcomes in patients treated with this strategy [40], some experts have suggested that the TEE strategy is a reasonable alternative to a conventional approach in some patients, such as those with a strong preference for early cardioversion, those with AF of less than three to four weeks duration who would benefit most from left atrial mechanical recovery, and those at increased risk of hemorrhagic complications (as the duration of precardioversion anticoagulation may be shortened). Another potential reason to consider this strategy is that a shorter period of AF may increase the likelihood of successful cardioversion and long-term maintenance of SR. (See "Atrial fibrillation: Cardioversion", section on 'Electrical cardioversion'.)

RECOMMENDATIONS OF OTHERS — Our recommendations are in broad agreement with those from the American Heart Association/American College of Cardiology/Heart Rhythm Society, the European Society of Cardiology, and the European Heart Rhythm Association [20,25,50,51].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Atrial fibrillation" and "Society guideline links: Arrhythmias in adults".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)

●Beyond the Basics topic (see "Patient education: Atrial fibrillation (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Rationale for anticoagulation – Conversion of atrial fibrillation (AF) to sinus rhythm (SR), either spontaneously or intended, is associated with a clinically important transient increase in the risk of thromboembolism, particularly stroke. This risk increases significantly after 48 hours of AF and can be lowered by therapeutic anticoagulation before cardioversion. (See 'Rationale for anticoagulation' above.)

●AF duration less than 48 hours – The following recommendations apply to patients with AF of clearly less than 48 hours duration (See 'AF duration less than 48 hours' above.):

•For patients with one or more high risk factors for thromboembolism (eg, prior thromboembolism, heart failure, or diabetes mellitus), we suggest deferral of cardioversion to allow for four weeks of effective therapeutic precardioversion anticoagulation rather than early cardioversion (Grade 2C). Anticoagulation with heparin or a direct-acting oral anticoagulant (DOAC) before, during, and after cardioversion along with precardioversion transesophageal echocardiography (TEE) is an alternative approach for these high-risk patients.

•For patients not at high risk of thromboembolism (listed in the above bulleted recommendation), we anticoagulate most patients with a CHA₂DS₂-VASc score ≥1 (Grade 2C). We prefer initiating a DOAC at least two to four hours prior to cardioversion, with continuation of the same DOAC for one month postcardioversion

•For patients with low risk of thromboembolism (CHA₂DS₂-VASc score 0 in men, 1 in women), our experts have differing approaches regarding postcardioversion anticoagulation, with some using four weeks of postcardioversion warfarin or DOAC anticoagulation and others not.

●AF duration uncertain or 48 hours or more – The following recommendations apply to patients with AF of more than 48 hours duration or when the duration is unknown (see 'AF duration uncertain or 48 or more hours' above):

•We recommend a minimum of four consecutive weeks of therapeutic anticoagulation (warfarin with an international normalized ratio [INR] greater than 2 or DOAC) prior to cardioversion, rather than proceeding directly to cardioversion (Grade 1B).

•We recommend a DOAC prior to elective cardioversion rather than warfarin irrespective of whether the anticoagulant will be given long term (Grade 1B). (See 'Anticoagulant approach' above.)

•For symptomatic patients in whom there is a strong preference to not delay cardioversion, or in whom there is a concern about bleeding with prolonged oral anticoagulation, or who are not likely to tolerate AF despite adequate rate slowing, a TEE strategy is a reasonable approach using therapeutic anticoagulation with heparin/warfarin or DOAC throughout the pericardioversion period. (See 'Transesophageal echocardiography-based approach' above.)

•We recommend therapeutic oral anticoagulation (with a DOAC or warfarin with target INR of 2 to 3) for four weeks after cardioversion in all patients, rather than discontinuing anticoagulation after cardioversion (Grade 1B).