2014 aats guidelines for the prevention and management of perioperative atrial fibrillation and flutter for thoracic surgical procedures

Clinical Guidelines 2014 AATS guidelines for the prevention and management ofperioperative atrial fibrillation and flutter for thoracicsurgical procedures Gyorgy Frendl, MD, Alissa C. Sodickson, Mina K. Chung, Albert L. Waldo, MD, Bernard J. Gersh, MB, ChB, James E. Tisdale, PharmD,Hugh Calkins, MD,Sary Aranki, Tsuyoshi Kaneko, Stephen Cassivi, MD,Sidney C. Smith, Jr, Dawood Darbar, Jon O. Wee, Thomas K. Waddell, MD, MSc, PhD,David Amar, MD,and Dale Adler, M guidelines within the framework of the new IOM Supplemental material is available online.
recommendations. In order to meet these standards, mostsocieties (American Heart Association and AATS included)initiated the revisioof existing guidelines.
Task force subgroups were formed and tasked with pre- Our mission was to develop evidence-based guidelines for paring a summary of the available literature for each sub- the prevention and treatment of perioperative/postoperative topic. Literature searches were conducted using PubMed, atrial fibrillation and flutter (POAF) for thoracic surgical focused on articles published since 2000 except in rare cir- procedures. Sixteen experts were invited by the American cumstances. Both the summaries and original articles were Association for Thoracic Surgery (AATS) leadership: 7 car- made available to each task force member via a shared diologists and electrophysiology specialists, 3 intensivists/ electronic folder. The subgroup summaries as well as the anesthesiologists, 1 clinical pharmacist, joined by 5 original literature were presented and discussed at 9 sched- thoracic and cardiac surgeons who represented AATS (see uled teleconferences. The conferences were recorded. Arti- for the list of members and cles were selected for inclusion based on consensus opinion for the conflict of interest decla- by task force members. Writing groups were formed to ration online).
develop the draft guidelines for each subtopic, with 3 to 7members and a leader for each group. Group recommenda- Methods of Review tions were submitted before being presented for discussion Members were tasked with making recommendations based and voting at a 1-day face-to-face conference.
on a review of the literature, with grading the quality of the ev- Members were specifically asked to assess the applica- idence supporting the recommendations, and with assessing bility of the available evidence to patients undergoing the risk-benefit profile for each recommendation. The level thoracic surgery. All recommendations were subjected to of evidence was graded by the task force panel according to a vote. Acceptance for the final document required greater standards published by the Institute of Medicine than 75% approval of each of the recommendations.
For the development of the guidelines we followed the recom- A final draft was prepared by the chairman of the task mendations of The Institute of Medicine (IOM) 2011 Clinical force and made available in a written form to each member Practice Guidelines We Can Trust: Standards for Developing for final comments. Subsequently, the recommendations Trustworthy Clinical Practice Guidelines; were posted for public comments for AATS members (via Efforts were made to minimize repetition of ex- REDCap), and then peer reviewed by outside experts isting ; rather we focused on new information and selected by the AATS Council.
advances in diagnosis and therapy, and present these current From the Department of Anesthesiology,a Perioperative Critical Care and Pain Med- of Cardiovascular Medicine,k Department of Medicine, Arrhythmia Service, Van- icine, Brigham and Women's Hospital and Harvard Medical School, Boston, Mass; derbilt University School of Medicine, Nashville, Tenn; Division of Thoracic Sur- Department of Cardiovascular Medicine,b Heart and Vascular Institute, Depart- gery,l Department of Surgery, Brigham and Women's Hospital and Harvard ment of Molecular Cardiology, Lerner Research Institute Cleveland Clinic, Lerner Medical School, Boston, Mass; Division of Thoracic Surgery,m Department of Sur- College of Medicine of Case Western Reserve University Cleveland Clinic, Cleve- gery, University of Toronto, Toronto, Ontario, Canada; Memorial Sloan-Kettering land, Ohio; Division of Cardiovascular Medicine,c Department of Medicine, Case Cancer Center,n Department of Anesthesiology and Critical Care Medicine, New Western Reserve University, Cleveland, Ohio; Harrington Heart & Vascular Insti- York, NY; Division of Cardiovascular Medicine,o Department of Medicine, Brig- tute,d University Hospitals Case Medical Center, Cleveland, Ohio; Division of Car- ham and Women's Hospital and Harvard Medical School, Boston, Mass.
diovascular Diseases and Internal Medicine,e Department of Medicine, Mayo Disclosures: See Clinic College of Medicine, Rochester, Minn; Department of Pharmacy Practice,f Received for publication June 9, 2014; accepted for publication June 10, 2014.
College of Pharmacy, Purdue University and Indiana University School of Medi- Address for reprints: Gyorgy Frendl, MD, PhD, Department of Anesthesiology, Peri- cine, Indianapolis, Ind; Department of Medicine,g Cardiac Arrhythmia Service, operative and Pain Medicine, Brigham and Women's Hospital, CWN-L1, 75 Fran- Johns Hopkins University, Baltimore, Md; Division of Cardiac Surgery,h Depart- cis St, Boston, MA 02115 (E-mail: ment of Surgery, Brigham and Women's Hospital and Harvard Medical School, J Thorac Cardiovasc Surg 2014;148:e153-93 Boston, Mass; Division of Thoracic Surgery,i Department of Surgery, Mayo Clinic College of Medicine, Rochester, Minn; Center for Heart and Vascular Care,j Copyright Ó 2014 by The American Association for Thoracic Surgery Department of Medicine, University of North Carolina, Chapel Hill, NC; Division The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines The following recommendations are based on the best lead to hemodynamic instability, necessitating prompt inter- available evidence from thoracic surgery. When evidence vention. A sustained increased heart rate can result in heart fail- specific to thoracic surgery was not available, we extrapo- ure, a less common but clinically devastating situation, the lated from the cardiac surgical literature. In the absence incidence of which is not reported in the literature.
of direct evidence, we present the best expert opinion based The incidence of POAF varies widely based on the intensity on cardiology/cardiac electrophysiology experience and of surgical stress and patient characteristics best practices.
). Some of the risk factors for AF such An executive summary was prepared for publication in a as hypertension, obesity, and smoking, are modifiable, printed format; this more extensive guideline was prepared whereas others, such as older age, Caucasian ancestry, and for online publication with additional comments, data, and a male sex, are not.
comprehensive list of references.
Thromboembolic events such as stroke or acute limb ischemia are the most serious and feared consequences of AATS Member Survey AF. Studies have reported a wide range of the incidence Our survey of the AATS members (results presented in of stroke related to POAF, although the risk for cardiac indicated the need for a guide- and thoracic surgery seems to be 50% to 200% higher line update and identified opportunities for improvement in than for general surgery.
the areas of prevention, standards for postoperative electro- Many studies show an increase in mortality in patients with cardiography (ECG) monitoring, and for the possible use of POalthough some studies have not shown novel oral anticoagulants. When asked how the AATS could such an effectGiven that patients with other significant help members improve their practices; 29% of respondents comorbidities or who are undergoing more complex recommended ‘‘initiating studies,'' whereas 58% recom- operations are more likely to experience POAF, it is unclear mended that the AATS ‘‘issue guidelines'' and promote uni- to what extent the arrhythmia itself contributes to mortality. It form practices.
is feasible that the contribution of POAF to mortality is moresignificant for those patients with fewer other comorbidities, Target Audience and the Patient Population however this independent effect is more difficult to measure These guidelines are intended for all noncardiac intratho- and has not been well reported in the literature.
racic surgeries and esophagectomies, as well as for patients POAF is associated with longer intensive care unit and whose risk factors and comorbidities place them at interme- hospital stays, increased morbidity (including strokes/new diate to high risk for POAF, independent of the procedure.
central neurologic events) with incidence of 1.3%- In assessing the patient's risk for POAF, it must be noted 1.7; and mortality (up to 5.6%-7.5%; RR 1.7- that the risks posed by the procedure and by patient fac- 3.as well as higher resource utilization.
tors/comorbidities will likely be additive, if not synergistic.
Multiple studies have consistently demonstrated an in- Therefore, these factors should be evaluated in combination crease in length of hospital stay in patients who develop during the preoperative assessment.
POAF, generally by a mean of 2 to 4 days The target audience includes not only thoracic surgeons An analysis of the Society of Thoracic Surgeons (STS) and anesthesiologists but all providers who participate in database by Onatis and colleaguedemonstrated that, in the care of thoracic surgical patients.
patients undergoing lobectomy or greater resection for lung The following novel information is included in this 2014 cancer, the presence of POAF lengthened hospital stay by a document: (1) standardized definitions for atrial fibrillation median of 3 days. The cost of hospitalization is likewise (AF) and (2) recommendations for: (a) ECG monitoring, (b) increased for patients who develop POAF, with an increase postdischarge management, (c) use of the new class of novel reported in the literature anywhere from 30% to 68% oral anticoagulants (NOAC); and (d) obtaining cardiology To some extent, this increase reflects comorbid conditions consultation. In addition, flow diagrams summarize the that occur along with POAF, but POAF itself is associated strategies for acute and chronic management. Specific with an increase in cost. Vaporciyan and colleagufound drug recommendations and dosing tables are also included.
that for patients who developed POAF without any othercomplications, the cost of care increased by more than EPIDEMIOLOGY OF POAF, ITS IMPACT ON US$6000, representing a greater than 30% increase.
OUTCOMES, COST, AND MORBIDITY AF, the most common sustained arrhythmia after pulmonary THE POSSIBLE MECHANISMS OF POAF AFTER and esophageal surgery, is a major, potentially preventable, adverse outcome. POAF peaks on postoperative days 2 to 4, The mechanisms that initiate and sustain AF, including and 90% to 98% of new-onset POAF resolves within 4 to 6 POAF, are complex and require both a vulnerable atrial weeks. Postoperative atrial fibrillation has multiple negative substrateand a trigger to initiate AF ). Today implications. In the acute setting, the tachyarrhythmia can they remain incompletely understood. The role of triggers The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines TABLE 1. Size of treatment effect and level of evidence for its impact Schema used to guide the grading of available published evidence and the expected effect of the interventions for their impact on patient outcomes (the arrow indicates the di-rection of increased effect size). COR, Class of recommendation.
from the pulmonary veins and other atrial sites initiating can promote triggers that initiate AFIn addition, surgical AFis well appreciated. However, it remains to be under- procedures are associated with local or systemic inflamma- stood why they occur and what exact mechanisms are essen- tion (such as pericarditis), an important risk factor affecting tial for their propagation. The identified risk factors for the the vulnerability of the atrial substrate to POAFThe extent development of sustained POAF are mostly identical to of pulmonary resection is another important risk factor for those known to make the atrium vulnerable to development development of POAF.The development of POAF is likely of AF in the nonsurgical setting. They include several risk to involve some or all of these mechanisms.
factors that are associated with atrial fibrosis, such as Insight into the mechanism of POAF can be gained by increasing age, atrial dilatation, myocardial ischemia, vol- examining what prophylactic therapies decrease the rate of ume overload, and a history of heart They also POAF occurrence after thoracic surgery. Higher norepineph- include risk factors such as increased norepinephrine rine levels were found in patients on preoperative b-blockers levels and increased vagal tone, both of which shorten who had their b-blocker therapy interrupted than in patients atrial wavelength, the latter known to increase atrial not receiving a b-blocker at all. This was associated with a vulnerability to AF.Both adrenergic and vagal stimulation significantly higher incidence of POAF.
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines TABLE 2, A. Risk stratification of thoracic surgical procedures for their risk of POAF Risk of POAF by surgical procedures Low risk procedures Intermediate risk procedures High risk procedures Type of procedures (<5% incidence) (5%-15% incidence) (>15% incidence) Flexible bronchoscopy with and Photodynamic therapy Tracheal stenting Placement of thoracostomy tube or PleurX catheter (CareFusion Corporation, San Diego, Calif) Rigid bronchoscopy Thoracoscopic wedge Bronchoscopic laser surgery Resection of anterior mediastinal Thoracoscopic lobectomy Open thoracotomy for lobectomy Tracheal resection and Bronchopleural fistula repair Lung transplantation Esophageal procedures Laparoscopic Nissen fundoplication/ dilation and/or stenting Zenker diverticulectomy Pericardial window Thoracic surgical procedures were divided into low (<5%), moderate (5%-15%) and high (>15%) risk groups based on their expected incidence of POAF in order to facilitatethe preoperative risk stratification of patients. POAF, Postoperative atrial fibrillation; PEG, percutaneous endoscopic gastrostomy.
Diltiazem therapy initiated in the early postoperative RECOMMENDATIONS AND REASONING period has been found to significantly reduce the rate ofPOAF.This is believed to be related to its effects of 1. Recommend the Use of the Following Definitions decreasing pulmonary vascular resistance. It is known that for the Diagnosis of POAF pulmonary hypertension and dilatation of the right side of the heart are associated with an increased incidence of 1.1. Electro-physiologic definition/diagnosis: ECG record- POAF.There is also the possibility that as a systemic ings (1 or more ECG leads) that demonstrate the pres- vasodilator, diltiazem could reduce preload and left atrial ence of characteristic ECG features of AF lasting at pressures.The data on use of verapamil have been incon- least for 30 seconds or for the duration of the ECG sistent with regard to decreasing the incidence of POAF recording (if shorter than 30 seconds)(level of Magnesium has been consistently shown to decrease evidence LOE C). Clinical symptoms may include the incidence of POAF after cardiac surgery, and the only hypotension, dizziness, decreased urinary output, prospective, randomized study on patients undergoing fatigue, and so on.
thoracic surgery also showed a significant decrease in the 1.2. Clinical definition/diagnosis: clinically significant incidence of POAFThe reason for its effectiveness is POAF is AF in the (intra- and) postoperative setting that requires treatment with rate or rhythm In the presence of a vulnerable substrate, additional elec- control agents, or requires anticoagulation, and/or ex- trophysiologic abnormalities (drivers) will sustain AF.
tends the duration of hospitalization (LOE C). Clinical The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines TABLE 2, B. Known patient risk factors for and comorbidities that TABLE 3. Probable mechanisms contributing to POAF increase the risk of POAF Clinically meaningful AF requires the presence of both a trigger and a Risk factors and comorbidities Thoracic surgery references vulnerable atrial substrate Modifiable risk factors Atrial substrate changes that facilitate AF Sympathetic or parasympathetic stimulation Atrial dilation or acute atrial stretch Inhomogeneous dispersion of conduction abnormalities Obstructive sleep apnea Short wavelength (conduction velocity 3 ERP) Other (like inflammation and oxidative stress) In addition, a driver(s) is thought to be needed to sustain AF in the vulnerable substrate Rapidly firing ectopic focus (atrial or other) Increased pulse pressure Reentrant circuit(s) of short cycle length (ordered reentry) Mitral regurgitation Potential role, if any, of multiple reentrant wavelets (random reentry) AF, Atrial fibrillation; ERP, effective refractory period.
Increased LV wall thickness should also receive ECG monitoring in the im- Nonmodifiable risk factors mediate preoperative period if procedures (eg, African American (protective factor) epidural catheter or other regional anesthesia blocks) are performed (LOE C).
2.2. Not using routine ECG telemetry is reasonable for pa- History of arrythmias tients who undergo low-risk (<5% expected incidence Derived from the 2014 American Heart Association Atrial Fibrillation Guidelines of AF) procedures, and have neither a previous history and relevant literature for thoracic surgery. Patient risk factors and comorbiditiesthat were shown to increase the risk of atrial fibrillation (AF) are listed. Much of of AF nor significant risk for stroke (based on this information was extracted from the general population, thoracic surgery–specific references are listed when available. These risk factors/comorbidities should be as- comorbidities (such as heart failure or previous stroke) sessed in conjunction with the procedure-related risks of AF in order to determinethe true risk of POAF. MI, Myocardial infarction; VHD, valvular heart disease; LV, left ventricle; LVH, left ventricular hypertrophy.
2.2.1. If patients exhibit clinical signs of possible AF decreased urinary output, fatigue, and so on.
while not monitored with telemetry, ECG re- We recommend that both electrophysiologically docu- cordings to diagnose POAF and ongoing telem- mented AF and clinically diagnosed AF be included in etry to monitor the period of AF should be the clinical documentation and reported in the clinical immediately implemented (LOE C).
2. Physiologic (ECG) Monitoring of Patients at Risk 3. Rate Control and Antiarrhythmic Drugs, Mechanism of Action, Side Effects, and Limitations Recommendations for ECG monitoring of patients at risk A detailed description of the drugs used for the manage- for POAF are presented in .
ment of rate or rhythm control (), theirmechanism of action, side effects, and limitations are discussed here. Dosing information is also presented in 2.1. Patients should be monitored with continuous ECG telemetry postoperatively for 48 to 72 hours (or lessif their hospitalization is shorter) if: 2.1.1. They are undergoing procedures that pose inter- mediate (5%-15% expected incidence of AF) or 3.1. To optimize the efficacy and safety of amiodarone, it is high (>15%) risk for the development of postop- reasonable to exercise caution when selecting its doses erative AF or have significant additional risk fac- or intravenous versus oral route, because cases of acute tors (CHA2DS2-VASc
2) for stroke (LOE C).
respiratory distress syndrome (ARDS) have been re- 2.1.2. They have a history of preexisting or periodic ported following pneumonectomy with cumulative recurrent AF before their surgery. These patients intravenous doses more than 2150 mg(LOE C).
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines TABLE 4. Recommended definitions for the diagnosis of POAF repolarization, thus slowing the discharge from the sinus noThis antiadrenergic activity in-hibits the renin-angiotensin-aldosterone system, ECG recordings (1 or more ECG leads) with ECG features of AF inhibits apoptosis, and reduces hyperphos- lasting at least for 30 seconds or phorylation of calcium-releasing chann for the duration of the ECG Metoprolol and atenolol are relatively selective recording (if <30 seconds) b-1 receptor antagonists (primarily affecting cardiac tissue) and in moderate doses have less ef- Clinical definition/ Clinically significant POAF: fect on the b-2 receptors in smooth muscle cells in intra- and postoperative AF the vasculature and bronchial tree. Propranolol requiring treatment, or and esmolol are nonselective, and carvedilol is anticoagulation, and/or nonselective and possesses a-receptor blocking extending the duration of hospitalization (LOE C) Intravenous administration of metoprolol, pro- These measures should be included in the clinical documentation and reported in theclinical trials/studies. POAF, Postoperative atrial fibrillation; ECG, electrocardiog- pranolol, and esmolol reduces ventricular raphy; COR, class of recommendation; LOE, level of evidence; AF, atrial fibrillation.
response in patients with AF within 5 minutes of administration,and both intravenous andoral regimens attain resting and exercise rate 3.2. Rate control agents: their mechanisms of action control, variably defined, in 68% to 75% of pa- tienRate-lowering efficacy varies with 3.2.1. b-Blockers acuity and cardiac function and is enhanced b-Blockers are Vaughan Williams class II antiar- with digoxin.
rhythmic agents that inhibit sympathetic nervous The major adverse effects of b-blockers are bron- system activity and slow the rate of phase IV chospasm in patients with asthma, particularly if TABLE 5. Recommendations for physiologic (ECG) monitoring the asthma is not well controlled; worsening of Recommendations for monitoring symptoms in patients with severe peripheralarterial disease; hypotension; and worsening of Patients should be monitored with continuous ECG telemetry postoperatively for 48-72 h (or less if their TABLE 6. Commonly used rate control agents hospitalization is shorter) if:  they are undergoing procedures that pose high Significant limitations (>15% expected incidence of AF) or intermediate Recommended doses and known side effects (5%-15%) risk for POAF or 0.25 mg/kg IV loading dose  they have significant additional risk factors over 2 min, then 5-15 mg/ (CHA2DS2-VASc >2) for stroke (LOE C) h IV continuous infusion Heart failure exacerbation  they have a history of preexisting or periodic 0.25 mg IV repeated every Nausea, vomiting, anorexia recurrent AF before their surgery 2-4 h to a maximum dose These patients should also receive ECG monitoring of 1.5 mg over 24 h in the immediate preoperative period if procedures Ventricular arrhythmias (epidural catheter, regional anesthesia blocks, and Accumulates in acute so forth) are performed (LOE C) kidney injury/chronic Not using routine ECG telemetry is reasonable for 500 mg/kg IV bolus over 1 min, then 50-300 mg/kg/  undergo low risk surgery (<5% expected incidence min IV continuous Heart failure exacerbation  had no previous history of AF, or 2.5-5.0 mg IV bolus over 2  have no significant risk for stroke and min; maximum 3 doses  have no relevant comorbidities (eg, heart failure or previous stroke) (LOE C) Heart failure exacerbation If patients exhibit clinical signs of possible AF while 150-300 mg IV over 1 h, not monitored with ECG telemetry, ECG followed by 10-50 mg/h QT interval prolongation recordings to diagnose POAF and continuous IV continuous infusion Pulmonary toxicity has not telemetry to monitor the period of AF should be been demonstrated at this immediately implemented (LOE C) ECG, Electrocardiography; COR, class of recommendation; AF, atrial fibrillation; Detailed information in section and in references IV, Intravenous; AV, POAF, postoperative atrial fibrillation; LOE, level of evidence.
The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines TABLE 7. Commonly used antiarrhythmic agents Significant limitations and Recommended doses known side effects Conversion to sinus rhythm: 20-50 mg/min IV continuous infusion until AF terminated, QT interval prolongation hypotension occurs, or QRS duration prolonged by Torsades de pointes 50%, or cumulative total dose of 15 mg/kg reached Contraindicated in patients with heart failure with Alternative dose: 100 mg IV every 5 min until AF reduced left ventricular ejection fraction terminated or other conditions as listed above are Contraindicated in patients with pretreatment QTc interval >470 ms (men) or 480 ms (women) If available orally, could be used for maintenance Conversion to sinus rhythm: 200-300 mg single Maintenance of sinus rhythm: 50-150 orally once Sinus bradycardia Contraindicated in patients with heart failure with reduced left ventricular ejection fraction Contraindicated in patients with coronary artery disease/structural heart disease Conversion to sinus rhythm: 450-600 mg single oral Maintenance of sinus rhythm: 150-300 mg orally every Sinus bradycardia 8 h (immediate release); 225-425 mg orally every 12 h (extended release) Contraindicated in patients with heart failure with reduced left ventricular ejection fraction Contraindicated in patients with coronary artery disease/structural heart disease Prophylaxis: 300 mg IV bolus, then 600 mg orally twice daily for 3-5 d QT interval prolongation Treatment: 150 mg IV over 10 min; then 1 mg/min IV Pulmonary toxicity has not been demonstrated at this continuous infusion for 6 h; the 0.5 mg/min IV continuous infusion for 18 h or change to oral administration at 100-400 mg daily QT interval prolongation Pulmonary toxicity has occurred at cumulative IV doses >2150 mg in patients undergoingpneumonectomy Not ideal for conversion to sinus rhythm in the QT interval prolongation postoperative setting; may take 2-3 d to convert to Torsades de pointes normal sinus rhythm, which would require Risk of torsades de pointes is greater in patients with commitment to anticoagulation Maintenance of sinus rhythm: calculated CrCl 20-40 Dose adjustment is important in patients with acute mL/min: 125 mg orally once every 12 h kidney injury or chronic kidney disease Calculated CrCl 40-60 mL/min: 250 mg orally once Contraindicated in patients with calculated CrCl <20 Calculated CrCl>60 mL/min: 500 mg orally every 12 h Contraindicated in patients with pretreatment QTc interval >470 ms (men) or 480 ms (women) Monitor ECGs 2 h after doses, telemetry for at least 3 d Conversion to sinus rhythm: QT interval prolongation Corvert prescribing Weight
60 kg: 1 mg IV administered over 10 min Torsades de pointes Weight <60 kg: 0.01 mg/kg IV administered over Risk of torsades de pointes greater in patients with heart 2006; Pfizer, Inc The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines TABLE 7. Continued Significant limitations and Recommended doses known side effects If the AF does not terminate within 10 min of Nonsustained ventricular tachycardia completion of the first infusion, a second dose of Sinus pauses after AF conversion equal strength may be administered IV over 10 min Contraindicated in patients with pretreatment QTc Not indicated for maintenance of sinus rhythm interval >470 ms (men) or 480 ms (women) Maintenance of sinus rhythm: 40-160 mg orally every Sinus bradycardia Dosing interval should be adjusted in patients with QT interval prolongation acute kidney injury or chronic kidney disease: Torsades de pointes If the calculated CrCl is 30-59 mL/min: administer Heart failure exacerbation Risk of torsades de pointes greater in patients with heart If the calculated CrCl is 10-29 mL/min: administer Dose adjustment is important in patients with acute kidney injury or chronic kidney disease Use with extreme caution in patients with calculated CrCl <10 mL/min and in patients undergoinghemodialysis Contraindicated in patients with pretreatment QTc interval >470 ms (men) or 480 ms (women) IV, Intravenous; AF, atrial fibrillation; AV, atrioventricular; CrCl, creatinine clearance; ECG, electrocardiography.
heart failure symptoms in patients with decom- Diltiazem must be used cautiously, especially pensated heart failure with reduced ejection frac- acutely, in patients concomitantly receiving tion. Intravenous b-blockers should not be used b-blockers, and is contraindicated in patients in patients with suspected accessory conduction with an accessory pathway result from acute concomitant administration of Digoxin inhibits sodium potassium adenosine b-blockers and diltiazem or verapamil.
triphosphatase (ATPase), thereby increasing intracellular sodium concentration leading to Diltiazem is a nondihydropyridine calcium chan- increased intracellular calcium concentrations.
nel antagonist and class IV Vaughan Williams In addition, digoxin administration is associated agent. Diltiazem inhibits L-type calcium channels with an increase in baroreceptor sensitivity in vascular and conduction tissue, and especially in disproportionate to hemodynamic improve- nodal tissue.In addition, diltiazem affects ment, and imparts vagomimetic (parasympa- the transient outward and ultrarapid delayed recti- thetic) effects. The vagomimetic effects of fier potassium currents in atrial myocytes. Intrave- digoxin occur at low serum concentrations and nous diltiazem administered as a bolus and contribute to decreasing sinus and atrioventric- ular (AV) nodal conduction. At higher serum response in 70% to 90% of patients with the concentrations, the parasympathetic effects recent-onset of AF. The onset of action of diltiazem actually shorten the refractory period of nonno- dal specialized conduction tissue.
Oral treatment with diltiazem in the Atrial Fibril- The onset of action of digoxin after intravenous lation Follow-Up Investigation of Rhythm Man- administration of 0.5 to 0.75 mg bolus doses is 30 agement (AFFIRM) trial was efficacious in minutes to 2 hoursWith additional intra- controlling rest and exercise heart rate in approx- venous bolus doses of 0.25 mg every 2 to 6 hours imately 60% of patients, and in 66% and 79% of after the first dose, up to a total dose within 24 patients, respectively, when combined with hours of 1.25 to 1.5 mg, 75% of patients with AF can achieve rate control at rest.Exercise rate Diltiazem can worsen heart failure in patients with control is achieved much less frequently, except reduced ejection fraction, and can cause important when digoxin is administered con-comitantly with gastrointestinal adverse effects including ileus.
a b-blocker or calcium channel blocker.
The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines Digoxin should not be administered to patients 3.3. Antiarrhythmic medications (mechanisms of ac- with suspected accessory pathways or obstructive tion, side effects) hypertrophic cardiomyopathy. The potential for 3.3.1. Amiodarone (see section 3.2.4) digoxin toxicity, including accelerated junctional 3.3.2. Flecainide rhythm, accelerated ventricular escape rhythms Flecainide is a Vaughan Williams class IC antiar- (sometimes heralded by regularization of the rhythmic agent that is a potent inhibitor of fast longest R-R intervals), nausea, and visual symp- sodium conductionConsequently, flecainide toms is increased in the presence of hypokalemia, decreases the maximum upstroke velocity and hypomagnesemia, hypercalcemia, and concomi- amplitude of atrial, ventricular, and Purkinje fiber tant therapy with amiodarone, dronedarone or action potentialsFlecainide may also inhibit Propensity matched comparisons IKr current, and prolongs the duration of atrial in the AFFIRM trial do not suggest an increase and ventricular action potential. In patients in mortality associated with chronic digoxin use.
without structural heart disease, oral flecainide 3.2.4. Amiodarone is relatively well tolerated; adverse effects Amiodarone is a Vaughan Williams class III agent include dizziness (15%-20%) and visual abnor- that inhibits inward potassium current, prolonging malities, including blurred vision and difficulty the action potential. However, amiodarone also in focusing (up to 15%), which can usually occur has properties that could place it in the other 3 during dose uptitrationHowever, in patients Vaughan Williams classes. It has antisympathetic with structural heart disease, flecainide is associ- and calcium-blocking activity that leads to impor- ated with more severe adverse effects. Flecainide tant effects on the sinoatrial (SA) and AV nodes, is associated with ventricular proarrhythmia in and the drug also has sodium channel–inhibiting this population; this proarrhythmia is not torsades properties that increases the threshold for depolar- de pointes (TdP), but rather monomorphic ven- tricular tachycardia. This proarrhythmia was thelikely cause of death associated with flecainide Intravenous amiodarone, administered as a bolus (and encainide) in the Cardiac Arrhythmia Sup- and continuous infusion, has an effect on heart pression Trial (CASTin which patients with rate within 4 hours that is similar to intravenous dil- a history of myocardial infarction and symptom- tiazem and intravenous digoxin, and improves ven- atic or asymptomatic ventricular ectopy (
6 ven- tricular rate in 74% of patients with AF by 24 tricular premature depolarizations VPDs per hourOral amiodarone can require days for hour) were randomized to receive flecainide, effective rate control to occur. Chronic oral amio- another Vaughan Williams class IC agent encai- darone therapy for rate control can have effects nide, or placebo for VPD suppression. Patients similar to those of digoxi randomized to receive therapy with flecainide Amiodarone is highly lipophilic, and intravenous or encainide had an increased risk of total mortal- administration may exert effects that are different ity and an increased risk of nonfatal cardiac arrest from those following oral administration. Intrave- and death from arrhythmia. The risk of proar- nous amiodarone can be associated with AV rhythmia associated with Vaughan Williams block, vasodilation, and hypotension. Intravenous class IC antiarrhythmic agents seems to be high- amiodarone should not be used in patients est in patients with ventricular conduction delays who have a suspected accessory pathway (QRS duration >120 milliseconds), structural Pulmonary toxicity associated with high-dose heart disease, ventricular scar tissue, or left intravenous amiodarone is discussed in section ventricular (LV) dysfunctionConsequently, flecainide should be avoided in these patients.
Chronic administration of oral amiodarone can In addition to the risk of proarrhythmia, flecai- be associated with pulmonary, hepatic, thyroid, nide has potent negative inotropic activity, and neurologic, cutaneous, and ocular toxicities.
has been associated with worsening heart failure in patients with coronary artery disease or pre- warfarin and inhibits elimination of the new existing heart failure (New York Heart Associa- oral anticoagulants. Amiodarone administration tion NYHA class II to IV and/or LV ejection can restore sinus rhythm so patients who fraction <30%Therefore, flecainide is con- receive it after 24 to 48 hours of AF require traindicated in patients with heart failure and reduced ejection fraction.
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines Intravenous flecainide is not available in the darone (10 to 40 days).Dronedarone's primary United States, but is available in other countries.
adverse effects include gastrointestinal distress In addition to the potential for ventricular proar- (16%), dizziness (9%), and bradycardia (3% rhythmia in patients with structural heart disease Dronedarone was associated with an increased and worsening of heart failure in patients with incidence of mortality in a randomized, double- LV dysfunction, intravenous flecainide may be blind, placebo-controlled study.and therefore is associated with hypotension.
contraindicated in patients with NYHA class III to IV heart failure, and in those patients with unsta- Magnesium administered intravenously is often ble NYHA class II heart failure.
referred to as a physiologic calcium channel Dronedarone has been shown to be effective for blocker, due to its antagonism of L- and T-type cal- maintenance of sinus rhythm in patients with cium channelsIntravenous magnesium dimin- nonsurgical paroxysmal AF. Dronedarone is con- ishes atrial automaticit) and inhibits AV node traindicated in patients with permanent AF, due to conduction.Intravenous magnesium is well toler- increased mortality associated with dronedarone ated; sinus bradycardia or AV block have been re- in that patient populatioThe efficacy of drone- ported with an incidence of approximately 3% darone for maintenance of sinus rhythm in patients Intravenous magnesium may also cause hypoten- with nonsurgical AF has not been investigated.
sion (approximate incidence 4%Transient adverse effects including flushing, tingling, and Ibutilide is a Vaughan Williams class III antiar- dizziness may occur in up to 17% of patients rhythmic agent that exerts its antiarrhythmic 3.3.4. Dofetilide activity via activation of slow inward sodium cur- Dofetilide is a Vaughan Williams class III antiar- reand inhibition of IKr.Ibutilide is effective rhythmic agent that inhibits I for conversion of atrial flutter and fibrillation to sinus Kr current,and pro- longs atrial and ventricular action potential rhythmIbutilide is not available in an oral dosage duration.Although dofetilide has been shown form, and therefore is not used for maintenance of si- to be effective for converting nonsurgical AF to si- nus rhythm. Ibutilide has been shown to be effective nus rhythmand for maintenance of sinus rhythm for conversion to sinus rhythm of AF occurring after in patients with nonoperative AF,it has not been coronary artery bypass graft surgery.The efficacy studied specifically for prevention or management of ibutilide for conversion to sinus rhythm of AF af- of AF after noncardiac thoracic surgery. As a ter noncardiac surgery has not been investigated.
result of its propensity to inhibit I The primary adverse effect associated with ibutilide ventricular repolarization, dofetilide may cause is TdP, which occurs in 1% to 3% of patients. The TdP, with an incidence of approximately 1% in incidence of TdP is 2- to 3-fold higher in patients patients with normal LV function.However, with heart failure as a result of reduced ejection the incidence increases to 3.3% in patients with fraction, which is a known risk factor for TdP. Ibu- heart failure with reduced LV ejection tilide may also cause nonsustained monomorphic To minimize the risk of TdP, dofetilide doses ventricular tachycardia in up to 8% of patients.
must be appropriately adjusted for kidney dis- 3.3.7. Procainamide Procainamide is a Vaughan Williams class IA anti- 3.3.5. Dronedarone arrhythmic agent that exerts its antiarrhythmic ef- Dronedarone is a Vaughan Williams class III fects through inhibition of fast sodium current as antiarrhythmic agent that was developed as a well as inhibition of IKr. In addition, a primary potentially safer congener of amiodarone. Drone- metabolite of procainamide, N-acetylprocaina- darone is similar to amiodarone in that it inhibits mide, inhibits IKr current and contributes to the multiple ion currents, including fast Naþ current, overall antiarrhythmic activity of procainamide.
Procainamide is effective for conversion of Kr, acetylcholine-activated Kþ current, and L- type calcium current.Dronedarone is also a nonoperative AF to sinus rhythm.The efficacy noncompetitive b-adrenergic inhibitor. Unlike of procainamide for conversion to sinus rhythm amiodarone, however, which possesses 2 iodine of AF after noncardiac thoracic surgery has not atoms that compose 37% of its molecular weight, been investigated. Procainamide is no longer dronedarone's structure does not include iodine available in an oral dosage form, and therefore is atoms. In addition, the half-life of dronedarone no longer indicated for maintenance of sinus (13 to 31 hours) is much shorter than that of amio- rhythm in patients with nonsurgical AF.
The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines The primary adverse effects associated with intra- may prolong the QT interval and cause TdP. The venous procainamide are hypotension, QT interval prolongation and TdP, and lengthening of the QRS management of AF after noncardiac thoracic surgery has not been evaluated.
3.3.8. Propafenone 3.4. Serum drug concentration monitoring Propafenone is a Vaughan Williams class IC anti- arrhythmic agent that is a potent inhibitor of Serum drug concentration monitoring may be sodium conductance.Propafenone may also warranted only if toxicity is of concern.
inhibit the transient outward potassium current Digoxin has a narrow therapeutic index, meaning (Ito) and the ultrarapid delayed rectifier potassium that serum concentrations required for efficacy (Ikur) current in atrial myocytes.Propafenone is are similar to those that may cause toxicity.
effective for maintenance of sinus rhythm in When used for heart failure, the desired therapeu- patients with nonoperative AF.In addition, tic range is 0.5 to 0.9 ng/mL.The optimal single-oral dose propafenone is effective for con- therapeutic range for digoxin for the management version of nonsurgical AF to sinus rhytThe of AF has not been established. The incidence efficacy of propafenone for prophylaxis or man- of adverse effects associated with digoxin agement of AF after noncardiac thoracic surgery increases with serum concentrations greater than has not been investigated.
Oral propafenone is well tolerated overall. Adverse During the management of AF after noncardiac effects include dizziness and blurred vision. How- thoracic surgery, monitoring of serum digoxin con- ever, propafenone possesses negative inotropic centrations for assessment of efficacy is not neces- activity, and is contraindicated in patients with sary, as a strong relationship between rate control heart failure due to reduced ejection efficacy and serum digoxin concentration has not In addition, propafenone is contraindicated in been established. Determination of serum digoxin patients with coronary artery disease or a history concentration may be warranted if patients exhibit of myocardial infarction. Although propafenone symptoms of digoxin toxicity, including nausea, was not studied in the CAST trial, the effects of vomiting, anorexia, or ventricular arrhythmias. If flecainide and encainide in that study are believe a serum concentration is believed to be necessary, to be to the result of potent sodium channel the blood sample should be obtained at least inhibition, and contraindications in patients with 12 hours, and preferably 24 hours, after the previ- structural heart disease have been applied to ous digoxin dose, as a result of the prolonged tissue distribution phas0-439); if the blood sample is obtained less than 12 hours after the dose, the Sotalol is an adrenergic b-receptor blocking serum concentration may be falsely increased, as agentthat also prolongs atrial and ventricular ac- a result of incomplete distribution of digoxin tion potential duration via inhibition of IKrSota- from serum to tissue.
lol is effective for reducing the incidence of To reduce the risk of digoxin toxicity in patients recurrent AF in patients with paroxysmal AF receiving the drug for AF after noncardiac and after conversion to sinus rhythm.Sotalol thoracic surgery, serum digoxin concentration has not been shown to be effective for conversion monitoring may be warranted if digoxin therapy of AF to sinus rhythm. Sotalol has been used to must be continued for longer than 1 week, for reduce the risk of AF after coronary artery bypass those patients who remain in AF after hospital graft (CABG) surgery.However, the efficacy of discharge. For patients with normal kidney func- sotalol for prophylaxis of AF after noncardiac tion, the half-life of digoxin is approximately 36 thoracic surgery has not been investigated.
hours; therefore, steady state serum concentra- 3.3.10. Quinidine tions require approximately 1 week. Routine Quinidine is a Vaughan Williams class IA antiar- determination of a steady state serum digoxin rhythmic agent that inhibits sodium conductio concentration after 1 week of therapy is not as well as conductance of a variety of potassium required in all patients. However, determination currents, including IKr, IKI, and ItoThe use of of a serum digoxin concentration after 1 week oral quinidine for management of AF has largely of therapy may be warranted in patients with been discontinued, because of evidence that chronic kidney disease or acute kidney injury, or quinidine may increase mortalityQuinidine in patients who are treated concomitantly with a The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines drug that inhibits digoxin elimination, such as thoracic surgery, has not been established. Serum amiodarone, dronedarone, propafenone, quini- flecainide concentration monitoring for prophy- dine, and verapam(pp410-439) laxis or treatment of AF after noncardiac thoracic 3.4.2. Procainamide: surgery is not warranted.
monitoring is not warranted The suggested therapeutic range for procainamide 3.5. Key limitations of drugs 3.5.1. Pulmonary toxicity efficacy is 4 to 10 mg/L.(pp440-462) However, this A primary concern regarding the administration of therapeutic range was determined using suppres- intravenous amiodarone following lung resection sion of ventricular premature depolarizations and is pulmonary toxicity, specifically ARDS. This prevention of episodes of ventricular tachycardia.
concern was prominently identified by Van Mie- Serum procainamide concentrations have not been ghem and colleawho initiated a study to correlated with efficacy in AF, and therefore, determine the comparative effectiveness of amio- desired serum procainamide concentrations for ef- darone, verapamil, or placebo for prevention of ficacy in AF are unknown. Serum concentration AF after pulmonary resection. The study was monitoring for intravenous procainamide for man- terminated prematurely due a high incidence of agement of AF after noncardiac thoracic surgery is ARDS in amiodarone-treated patients, specif- not warranted. The risk of adverse effects associ- ically in patients who had undergone pneumonec- ated with intravenous procainamide can be mini- tomy. At the time of discontinuation of the mized by terminating the loading dose of 20 to amiodarone arm, the drug had been administered 50 mg/min continuous infusion if hypotension oc- to 32 patients, of whom 21 had undergone lobec- curs, QRS duration is prolonged by 50%, or a cu- tomy and 11 had undergone pneumonectomy.
mulative intravenous dose of 17 mg/kg has been No patients who underwent lobectomy developed amiodarone-associated ARDS. In contrast, 3 of 11 3.4.3. Amiodarone: patients (27%) in the amiodarone group who un- monitoring is not warranted derwent pneumonectomy developed ARDS. The Serum amiodarone concentration monitoring has investigators recommended avoiding amiodarone been performed during therapy for ventricular ar- administration for patients undergoing pulmonary rhythmias. However, a relationship between serum amiodarone concentrations and efficacy Other investigators have administered intrave- for prevention or management of AF has not nous amiodarone to patients undergoing lung been established. Similarly, a relationship be- surgery without adverse effects. In a prospective, tween serum amiodarone concentrations and most of the adverse effects of amiodarone, partic- laxis,the incidence of ARDS among the 65 ularly those that occur during short-term therapy, amiodarone-treated patients (of whom 40 under- has not been established. Therefore, monitoring went lobectomy, 8 underwent bilobectomy, and of serum amiodarone concentrations during pro- 17 underwent pneumonectomy) was 0%. Barbe- phylaxis or management of AF after noncardiac takis and colleaguesadministered intravenous thoracic surgery is not warranted. However, to amiodarone to 43 patients for treatment of AF minimize the risk of pulmonary toxicity, it is rec- after lung resection. No patients developed ommended to keep total cumulative intravenous ARDS; 21 of these patients underwent pneumo- amiodarone doses to less than 2150 mg.
nectomy. Riber and colleaguesconducted a 3.4.4. Flecainide: serum drug concentration moni- randomized, prospective, double-blind, placebo- toring is not warranted controlled study of amiodarone for prevention The therapeutic range for serum flecainide of AF after lung resection. Only 2 patients of the 122 who received amiodarone underwent 2.5 mg/L(pp440-462) However, this therapeutic range was developed using suppression of ventric- right-side lobectomy or bilobectomy. No patients ular premature depolarizations as an end point, in this study developed ARDS or any pulmonary rather than efficacy for the management of AF. A relationship between serum flecainide concentra- One potential difference in patients undergoing tions and efficacy for prophylaxis or management pneumonectomy in the Van Miegham stu of AF, particularly that occurring after noncardiac compared with these more recent trials The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines include the cumulative intravenous amiodarone calcium concentrations should be maintained in dose administered. In the Van Mieghem study, the normal range. Drug interactions leading to intravenous amiodarone was administered as a increased concentrations of a QT interval– bolus of 150 mg over 2 minutes, followed by a prolonging drug should be avoided. Doses of continuous infusion of 1200 mg over 24 renally eliminated QT interval–prolonging drugs hours for 3 consecutive days, for a possible (dofetilide, procainamide, sotalol) should be cumulative intravenous amiodarone dose of 3750 appropriately adjusted for declining kidney mg. The 3 patients who developed amiodarone- function. In addition, concomitant therapy with induced ARDS received cumulative intravenous other QT interval-prolonging drugs, particularly amiodarone doses of 2150, 3750, and 3350 mg noncardiovascular QT interval-prolonging drugs before discontinuation of therapy. In the more (fluoroquinolone and macrolide antibiotics, azole recent studies, patients received a cumulative intravenous amiodarone dose of 1050 mg, after chotics, many others)should be avoided or per- which oral amiodarone was initiated,or a formed cautiously.
loading dose of 300 mg intravenous amiodarone 3.5.3. Hypotension before switching to oral amiodaroIn the Bar- Several drugs administered intravenously for betakis study,intravenous amiodarone was prophylaxis or management of postoperative administered as a loading dose of 5 mg/kg over AF may cause hypotension, including diltiazem, 5 minutes, followed by 15 mg/kg for an undefined esmolol, metoprolol, procainamide, and amio- time period. In addition, in the Van Miegham darone. Drug-associated hypotension is more study, the 3 patients who developed amiodarone- likely to occur when patients are volume associated ARDS underwent right-sided pneumo- depleted, which is often the case after thoracic nectomy, which is associated with a higher risk of surgery. In the population with AF after coronary postoperative ARDS than other types of lung artery bypass graft, hypotension associated with intravenous diltiazem was more likely when Overall, administration of amiodarone at the dose the pretreatment systolic blood pressure was shown to be effective by Riber and colelagues less than 115 mm (300 mg intravenous loading dose followed by 3.5.4. Bradycardia 600 mg orally twice daily for 5 days) seems to Drugs used for ventricular rate control can also be safe and effective for prevention of AF after result in bradycardia through inhibition of sinus node function or AV nodal conduction. These 3.5.2. QT interval prolongation/torsades de pointes drugs include amiodarone, propafenone, flecai- Several of the drugs that may be used for prophy- nide, esmolol, metoprolol, sotalol, and diltia- laxis or management of postoperative AF may The risk is higher when combinations of cause QT interval prolongation, and therefore sinus node or AV node-inhibiting drugs are used.
pose a risk for the life-threatening polymorphic 3.5.5. Exacerbation of heart failure with reduced LV ventricular arrhythmia known as TdPDrugs ejection fraction that prolong the QT interval are generally those Several drugs used for prophylaxis or treatment of that inhibit IKr, and include amiodarone, procaina- postoperative AF possess negative inotropic ac- mide, dofetilide, dronedarone, ibutilide, sotalol, tivity and are contraindicated in patients with and quinidine. A Bazett-corrected QT (QTc) in- heart failure with reduced LV ejection fraction.
terval greater than 500 ms markedly increases These drugs include diltiazem, procainamide, propafenone, and flecainide.
receiving a drug that prolongs the QTc intervalshould have the QTc interval measured from arhythm strip or 12-lead ECG at least once daily 4. Prevention Strategies and Their Efficacy during therapy. In addition, because the occur- Recent evidence suggest that some prevention strategies rence of TdP is highly dependent on the presence (avoiding b blockade withdrawal for those chronically on of other risk factors (female sex, hypokalemia, those medications, correction of serum magnesium when hypomagnesemia, hypocalcemia, bradycardia, abnormal) may be effective for all patients for reducing heart failure, increased serum drug concentra- the incidence of POAF. By surveying the AATS member- modifiable risk factors should be ship, we also found that many of these strategies are corrected. Serum potassium, magnesium, and currently underused ).
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines 4.1. Recommended prevention strategies for all patients 4.3.1. Left atrial appendage excision may be considered at the time of extensive left lung surgery for pa-tients with preexisting AF who are considered 4.1.1. Patients taking b-blockers before thoracic surgery too high a risk for anticoagulation in the perioper- should continue them in the postoperative period ative period(LOE C).
to avoid b-blockade withdrawal.(LOEA).
4.1.2. Intravenous magnesium supplementation may 4.4. Prevention of postoperative AF be considered to prevent postoperative AF AF, the most common sustained arrhythmia after pul- when serum magnesium level is low or it is monary and esophageal surgery, is associated with longer intensive care unit and hospital stays, increased depleted.(LOE C).
morbidity and mortality, and higher utilization of health care resourcePOAF also represents a major 4.1.3. Digoxin should not be used for prophylaxis potentially preventable adverse outcome. Several against AF.(LOE A).
randomized controlled studies and meta-analyses 4.1.4. Catheter or surgical pulmonary vein isolation have examined the efficacy of a variety of agents (at the time of surgery) is not recommended for including antiarrhythmics, b-blockers, and novel prevention of POAF for patients who have no agents such as magnesium and statins to prevent the previous history of A(LOE C).
development of POAF in patients undergoing thoracic 4.1.5. Complete or partial pulmonary vein isolation at surgery. However, it should be appreciated that there the time of (even bilateral) lung surgery should is a dearth of data indicating that prophylactic therapy not be considered for prevention of POAF, as it for AF improves outcomes after thoracic surgery (eg, is unlikely to be effecti(LOE B).
stroke) and reduces length of hospital stay, and many For those patients at increased risk for the develop- of the recommendations are extrapolated from the car- ment of POAF, preventive administration of medi- diac surgery arena.
The recommendation to avoid withdrawal of b- reasonable. However, these strategies may not be blockers in all patients undergoing thoracic surgery is useful for all thoracic surgical patients.
mainly derived from the cardiac surgery literature. Nat-tel and colleaguesshowed that abrupt propranolol 4.2. Recommended prevention strategies for intermedi- withdrawal was associated with increased sensitivity ate to high-risk patients to isoproterenol, and a large meta-analysis of random- ized studies confirmed that acute withdrawal of b- 4.2.1. It is reasonable to administer diltiazem to those blockers before cardiac surgery increases the risk of patients with preserved cardiac function who are developing POAF.There are only limited data sup- not taking b-blockers preoperatively in order to porting the role of prophylactic b-blockers in patients prevent POAF(LOE B).
undergoing thoracic surgery.Although two of 4.2.2. It is reasonable to consider the postoperative admin- these randomized studies showed a reduction in istration of amiodarone to reduce the incidence of POAF, there was a high incidence of hypotension and POAF for intermediate and high risk patients under- bradycardia that limited the use of b-blockers in the going pulmonary resectio(LOE A).
perioperative setThere remains controversy in the recent literature as to whether to initiate periopera-tive b-blockers in patients who are not already taking 4.2.3. Postoperative them. At recommended doses aimed at achieving a amiodarone may be considered to prevent target heart rate, b-blockers may cause significant post- operative hypotension and stroke-related mortality.
In randomized controlled trials, diltiazem has not 4.2.4. Atorvastatin may be considered to prevent POAF been associated with perioperative hypotension. The for statin-naive patients scheduled for intermedi- ability of diltiazem to reduce AF after thoracic surgery ate- and high-risk thoracic is moderate.
dure(LOE C).
4.3. Recommended prevention strategies for the highest- To date, the best evidence for efficacy of AF prevention in general thoracic surgery patients has been with The Journal of Thoracic and Cardiovascular Surgery c September 2014

Clinical Guidelines FIGURE 1. Prevention strategies and their efficacy for postoperative atrial fibrillation (POAF). LOE, Level of evidence; PVI, pulmonary vein isolation; i.v., intravenous; LVEF, left ventricular ejection fraction; AF, atrial fibrillation.
amiodarone. An important issue with any prevention ef- 5 days. They showed that amiodarone-treated pa- forts is the acceptance of a recommended medication tients had a rate of POAF (lasting>5 min) of 9% by the responsible surgical team, particularly with a (11 of 122), compared with placebo controls who drug like amiodarone that has potential for side effects.
had a rate of 32% (38 of 120). A final study of The antiarrhythmic mechanism of amiodarone com- patients undergoing pulmonary resection random- bines varying degrees of class III antiarrhythmic activ- ized 2 groups of patients in a prospective, double- ity, b-blockade, and calcium channel antagonism.
blind design to either amiodarone (postoperative Slower postoperative heart rates with short-term use intravenous loading 5 mg/kg, then 15 mg/kg for and greater than moderate efficacy in reducing AF 48 hours intravenous infusion) or magnesium sul- may result in wider physician acceptance of amiodar- fate (preoperative loading of 80 mg/kg and then 8 one, although concerns regarding rare reports of pulmo- mg/kg/h for 48 hours intravenous infusion after nary toxicity with right lung resection or lung surgery)This study showed that the incidence transplantation may moderate its use (discussed in of POAF (lasting >30 seconds) was 10% (21 of more detail in section 3).
219) with amiodarone and 13% (27 of 219) withmagnesium. None of these studies reported any 4.5. Pharmacologic therapies to prevent POAF serious adverse effects caused by amiodarone 4.5.1. Amiodarone except occasional bradycardia.
Efficacy of amiodarone: Tisdale and colleagues Safety of amiodarone: In the nonsurgical popula- showed that amiodarone 1.05 g given by contin- tion, it is commonly accepted that amiodarone- uous intravenous infusion over the first 24 hours related pulmonary toxicity does not occur with after pulmonary resection and then 400 mg orally short-term (<1 month) exposure. Concerns about twice daily for up to 6 days, reduced the rate of POAF requiring treatment, 9 of 65 (14%) in toxicity were raised 2 decades ago in a small ran- comparison with 21 of 65 (32%), in an untreated domized study that was interrupted early because control group. In a similar study, the same inves- 3 right-sided pneumonectomy patients of a total showed that continuous infusion of of 11 patients who received amiodarone for pre- amiodarone 43.75 mg/h for 96 hours (total dose vention of POAF developed ARDS, whereas 4200 mg) was associated with a lower POAF none of the 21 patients undergoing lobectomy rate of 6 of 40 (15%) in patients undergoing and exposed to amiodarone developed this compli- esophagectomy compared with 16 of 40 (40%) cation.The investigators acknowledged that in an untreated control group. The largest trial right-sided pneumonectomy in itself was a well- to date by Riber and colleaguesused a random- established risk for ARDS, but nevertheless ized, double-blind, placebo-controlled design of cautioned on the use of amiodarone for AF preven- amiodarone given by loading 300 mg intrave- tion after pulmonary resection. Since then, several nously immediately when stable after surgery observational and more recent prospective ran- followed by 600 mg orally twice daily for up to domized trials failed to find a link between use of The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines amiodarone for AF prevention and ARDS immedi- diltiazem-treated patients, especially early after surgery, with resumption of diltiazem therapy Several other studies used amiodarone for acute soon thereafter.
treatment of AF after general thoracic surgery, 4.5.3. Novel therapies to prevent postoperative AF and none of these reported amiodarone-related pul- Inflammation and oxidative stress play an im- monary toxicity. Of 3 retrospective studies portant role in the pathogenesis of AFSeveral describing risk and treatment of AF after lung trans- studies have examined the role of statin therapy in plant, only 1 studreported an association be- preventing POAF. One of the largest (n ¼ 200) tween pulmonary toxicity and amiodarone use randomized studies evaluated the role of atorvas- and cautioned on the routine use of amiodarone af- tatin, given 7 days before and 7 days after cardiac ter lung and heart-lung transplants surgery. Those patients who received the statin demonstrated a 22% reduction in incidence of POAF. Amar and colleaguesconducted a randomized controlled trials that evaluated calcium prospective study of 131 patients undergoing ma- channel blockers given immediately before, dur- jor lung or esophageal surgery to evaluate the ing, or after CABG surgery or valve surgery relationship between C-reactive protein and showed that these drugs reduced rates of myocar- POAF. A secondary analysis in this study showed dial injury and supraventricular tachycardia that in the subset of patients receiving preopera- (SVTIn patients undergoing thoracic surgery, tive statins, the risk of developing AF was almost there have been 3 prospective randomized trials 3-fold lower than in those not taking them. Two of nondihydropyridine calcium channel antago- meta-analyses of randomized studies examining nists for the prevention of AF. Verapamil prophy- prophylactic statin therapy was performed that laxis was used in a large, randomized, open-label involved more than 2200 patients.These study of patients undergoing lobectomy or pneu- monectomy using somewhat aggressive loading preventing POAF in statin-naive patients under- (started within 1 hour of arrival in recovery, 10 going high-risk cardiac and noncardiac surgery mg over 10 minutes followed by 0.375 mg/min or after acute coronary syndromes. Additional over 30 minutes) and then by continuous infusion randomized placebo-controlled studies will be (0.125 mg/min for 3 days).This medical regimen required before statin therapy can be recommen- was associated with a nonsignificant reduction of ded as a Class I or IIa indication to prevent AF from 15% (15 of 99) in placebo patients to POAF in statin-naive patients undergoing moder- 8% (8 of 100), respectively. In a small, random- ate to high-risk lung surgery. As in other patients ized, open-label study of patients undergoing stan- undergoing noncardiac surgery, most physicians dard or intrapericardial pneumonectomy, diltiazem continue statin therapy preoperatively to avoid prophylaxis was associated with reduced overall incidence of SVT in comparison with digoxin- There is strong evidence supporting the use of treated patients (0 of 21 vs 8 of 25, P<.005, respec- magnesium supplementation to prevent POAF in tively)In a larger follow-up, randomized, patients undergoing cardiac surgery.In the double-blind, placebo-controlled study of patients undergoing lobectomy or pneumonectomy, diltia- controlled trial (n ¼ 200) in patients undergoing zem started within 1 hour of arrival in recovery, thoracic surgery, Terzi and colleaguesdemon- 0.15 mg/kg (loading while the patient was fasting strated that the incidence of postoperative atrial then 120 mg orally twice a day for 14 days) reduced tachyarrhythmias, mainly AF, was reduced from the rates of postoperative atrial arrhythmias in com- 23% to 11% in those patients treated with an parison with placebo (25 of 167 [15%] vs 40 of 163 intravenous infusion of magnesium during the [25%], respectively, P ¼ .03) Safety of diltiazem: With the doses described in a Three large randomized clinical trials have clearly randomized open-label study of verapamil given demonstrated that prophylaxis with digoxin does early after lobectomy or pneumonectomy, 14% not prevent and may in fact increase the incidence of the patients experienced hypotension and 9% of POAF in patients undergoing all types of had bradycardia requiring temporary interruption thoracic surgery.Although, acute digoxin of the drug infusion.In contrast, mild transient loading may be beneficial in controlling rapid hypotension was reported in 4% (6 of 163) of ventricular rates during AF in patients with The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines hypotension, there is no place for digoxin namic stability. Although some interventions are likely to prophylaxis in patients undergoing thoracic benefit all patients (see section 5.1), hemodynamically un- stable patients will require urgent efforts for the restitution 4.5.4. Surgical prevention strategies of sinus rhythm (section 5.2). However, for stable patients Cardiac surgery patients with preexisting AF who with POAF, the emphasis shifts to rate control strategies undergo surgical pulmonary vein isolation (PVI) (see details in section and receive additional biatral linear lines of blockmay achieve a 75% to 85% freedom from AF at 6 to 12 months, and the procedure adds on average 5.1. Management strategies recommended for all pa- an additional 9 minutes to the surgery, without tients with new-onset POAF ) perceptible safety risks, although possibly the pa- tients have a slightly higher risk of needing a pace- 5.1.1. Reduce or stop catecholaminergic inotropic maker during the early postoperative period.
agents if hemodynamics allow (LOE C).
Patients who undergo the equivalent of near- 5.1.2. Optimize fluid balance and maintain normal complete PVI associated with bilateral lung trans- electrolyte levels (LOE C).
plantation have a low incidence of AF 3 to 6 months 5.1.3. Evaluate the presence of and treat all possible after the procedure.For patients with preexisting correctable triggering factors. These may include AF who are known to tolerate the arrhythmia bleeding, pulmonary embolism, pneumothorax, poorly or who have an increased bleeding risk on pericardial processes, airway issues, myocardial anticoagulants, 2 questions arise: would PVI, bilat- ischemia, or infection/sepsis (LOE C).
eral or unilateral, help prevent perioperative AF, and would operative left atrial appendage exclusionlower perioperative thrombotic risk? 5.1.4. Cardiology consultation may be useful for those patients (LOE C) who: Multiple studhave shown that theincidence of POAF is in the 20% to 40% range 5.1.4.1. Develop recurrent or refractory POAF.
even after double lung transplant, confirming 5.1.4.2. Develop a hemodynamically unstable condi- that this form of AF is related to inflammatory, mechanical, and autonomic factors, in addition 5.1.4.3. Are at high risk for stroke based on to pulmonary vein triggers. Unilateral PVI is not CHA2DS2-VASc score and will likely likely to be any more beneficial, and has a require longer-term anticoagulation.
distinct disadvantage beyond the perioperative 5.1.4.4. Require a second-line antiarrhythmic medi- cation for stabilization.
Excision of the left atrial appendage can be per- 5.1.4.5. Also develop acute kidney injury.
formed safely, with efficacy rates approaching 87%, after a learning curve,but there are no For all patients with new-onset POAF after thoracic sur- studies that show a reduction in perioperative gery, consideration should be given to triggering causes.
thrombotic events. The Left Atrial Appendage Although inflammation provoked by surgical procedures, pa- Occlusion Study is ongoing. Data from the tient risk factors for AF, and mechanical proximity of Watchmaand Prevail trials cannot necessarily thoracic surgery procedures to cardiac structures are often be extrapolated to operative left atrial appendage sufficient to explain the occurrence of POAF, other triggers exclusion, and neither of these studies involved may need to be identified in patients with recurring, symp- patients undergoing thoracotomy. Although the tomatic, or refractory AF. These include bleeding, pulmonary alteration in left atrial compliance and filling pres- embolism, pneumothorax, pericardial processes, airway is- sures after appendage exclusion may be small, the sues, myocardial ischemia, or infection/sepsis. Minimiza- effects may be different in a lung population that tion, weaning or discontinuation of catecholaminergic may have a smaller atrial size than in a cardiac inotropic agents, if possible, optimization of fluid status, valve surgery population.
and correction of any electrolyte/metabolic disturbancesmay also facilitate restoration and maintenance of sinusrhythm.
5. Treatment Strategies for POAF and Their Efficacy As a general rule, although much, if not most, POAF is The management of patients presenting with POAF transient and largely limited to the postoperative period requires different strategies depending on their hemody- (2-6 weeks), consultation with a cardiologist or cardiac The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines FIGURE 2. Management algorithm for postoperative atrial fibrillation (POAF). AF, Atrial fibrillation; MI, myocardial infarction; HF, heart failure; WPW, Wolff-Parkinson-White syndrome; DC, direct current; i.v./IV, intravenous; HR, heart rate; LV, left ventricular; COPD, chronic obstructive pulmonary dis- ease; LA/LAA, left atrial/left atrial appendage; TEE, transesophageal echocardiography.
electrophysiologist may be useful, especially for patients with recurrent or refractory POAF. This is usually for 3 5.2. Recommendations for the management of the he- issues: management of rate control during AF; consider- modynamically unstable patient with new-onset ation of whether, when, and how to restore sinus rhythm; and consideration of anticoagulation. The first issue is usually not difficult to accomplish, and standard prophy-lactic use of diltiazem or b blockers usually ensures 5.2.1. Emergency R-wave synchronized direct current that should POAF occur, ventricular rate control may be (DC) electrical cardioversion is recommended as simple as maintaining this therapy. However, tachy- for hemodynamically unstable patients and for brady syndrome may complicate efforts at rate control patients with evidence of acute myocardial that may require alternative medical options, rhythm con- ischemia or infarction. Signs of hemodynamic trol strategies, or antibradycardic pacing. The second instability include severe symptomatic hypoten- issue may be more complex, but decisions about if, sion, shock, or pulmonary edem(LOE C).
when, and how to restore sinus rhythm often benefits 5.2.1.1. For unstable patients with new-onset POAF from direct, nuanced cardiology and/or cardiac electro- of less than 48-hours duration, emergency physiology involvement. Such consultation can be useful DC cardioversion is indicated and is accept- in the selection and management of antiarrhythmic med- able before initiation of anticoa ications or in determining a need for permanent pacing.
The third issue is probably the most important and a chal- 5.2.1.2. For unstable patients who undergo cardio- lenge particularly for patients at high risk for bleeding.
version more than 48 hours after the onset Cardiologists may assist with management of patients at of AF, and who do not have an excessive high risk for stroke needing longer-term anticoagulation, bleeding risk or other contraindication, anti- unstable patients, or patients with acute kidney injury, coagulation should be initiated as soon as which can worsen outcomes, including stroke, and limit possible and continued for at least 4 antiarrhythmic and anticoagulant choices. Close interac- tion between the surgical team and the cardiology team should provide excellent, well-considered anticoagulationdecisions. In the end, the patient and thoracic surgical 5.2.2. If initial DC cardioversion is unsuccessful or he- team will be well served by a close consultative relation- modynamically unstable AF recurs, the following ship with the cardiologist.
steps can be useful: The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines FIGURE 3. Management of the hemodynamically unstable patient with new onset postoperative atrial fibrillation (POAF). AF, Atrial fibrillation; MI, myocardial infarction; HF, heart failure; DC, direct current; i.v., intravenous; TEE, transesophageal echocardiography.
5.2.2.1. Initiate rate and possible rhythm control there is a higher risk of left atrial or atrial appendage therapy with intravenous esmolol, diltiazem, thrombus that could dislodge at the time of or in the days digoxin, or amiodarone while preparing for following cardioversion in the absence of anticoagulation.
repeat DC cardioversion (LOE C).
Thus, it is recommended that for these patients, in the 5.2.2.2. Repeat DC cardioversion (more likely to be absence of contraindications (such as excessive bleeding successful after initiating a rhythm control risk or known heparin sensitivity), heparin be administered agent) (LOE C).
concurrently with the cardioversion, and used duringtransition to an oral anticoagulant. The oral anticoagulant should be provided for at least 4 weeks after cardioversion, Some patients with new-onset AF are hemodynamically as for patients undergoing elective cardioversi unstable, defined as AF associated with symptomatic severe If the initial cardioversion is unsuccessful or hemody- hypotension, evidence of acute myocardial ischemia or namically unstable AF recurs, repeat cardioversion can be infarction, or pulmonary edema/heart failure. For such attempted. To facilitate this, and while preparing for repeat patients, immediate electrical DC cardioversion is recom- cardioversion, attempts at pharmacologic rate or rhythm menElectrical cardioversion should be performed control may be considered with such drugs as intravenous under deep conscious sedation with R-wave synchronized amiodarone, esmolol, diltiazem, or digoxin. When hypoten- shocks. Cardioversion can be performed with biphasic or sion is a problem, intravenous digoxin may be considered.
monophasic waveforms. However, biphasic waveform However, should pharmacologic management fail, repeat shocks are preferred over monophasic waveforms, as the electrical cardioversion is recommended.
latter can require higher defibrillation energies forsuccess. Anterior-posterior electrode patch positioning (eg, R parasternal to L posterior or midlow sternal to poste- 5.3. Recommendations for the management of the he- rior) may produce a more successful defibrillation modynamically stable patient with new-onset AF vector for cardioversion of AF than anterior only (eg, R par- asternal to anterior or anterolateral apex) positions. If DC Primary treatment goal is rate control with rhythm cardioversion using 1 defibrillator patch location fails, the control as a secondary option.
alternate patch position should be used.
If AF duration in the unstable patient is less than 48 hours, cardioversion can be performed before initiation of anticoa- 5.3.1. It is reasonable to manage stable, well-tolerated gulation.However, for patients with AF of more than 48 new-onset POAF with a rate control strat- hours duration who become hemodynamically unstable, The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines 5.3.2. Rhythm control with antiarrhythmic drugs and/or the rhythm control arm and longer 6-minute walk test dis- DC cardioversion can be useful for patients with tances in patients in sinus rhythm.
hemodynamically stable new-onset POAF who However, there are no randomized trials studying rate have recurrent or refractory POAF, continued versus rhythm control strategies for POAF after thoracic symptoms, intolerance to rate control medica- surgery, and there have been only small, randomized, pilot tions, or ventricular rates that cannot be trials performed after cardiac surgery. In a randomized pilot adequately controlled(LOE C).
study by Lee and colleaguesof 50 patients with POAF 5.3.3. A rhythm control approach with pharmacologic or after cardiac surgery, 27 were randomized to antiarrhythmic DC cardioversion is reasonable for patients with drug therapy  electrical cardioversion and 23 to a rate new-onset POAF nearing 48 hours in duration, control approach. The end points were length of stay and who are at high risk for bleeding, in order to incidence of recurrent AF. There was no significant differ- avoid anticoagulation that would be otherwise ence in time to conversion to sinus rhythm. With multivari- indicated for AF persisting longer than 48 hours able Cox analyses, adjusting for other covariates, there was a trend toward a reduction in time from treatment to sinusrhythm in the antiarrhythmic arm (P ¼ .08), as well as a shorter length of stay (P ¼ .05). At termination, 91% Similar to AF occurring after cardiac surgery, new-onset were in sinus rhythm in the rate control arm and 96% POAF after thoracic surgery is often self-limited with were in sinus rhythm in the antiarrhythmic arm. Most patients returning to sinus rhythm within 4 to 6 weeks after were in sinus rhythm after 2 months. In a randomized pilot surgery regardless of a rate or rhythm control strategy. An study by Soucier and colleaguesof stable patients with observational study of 30 patients with new-onset AF after new AF after cardiac surgery, 42 patients were randomized lung resection and no history of heart rhythm disease to propafenone 600 mg (n ¼ 20) versus ibutilide 1 mg intra- reported that sinus rhythm was restored within the first venously up to 2 doses (n ¼ 10) versus rate control (n ¼ 12).
24 hours in 70% of patients treated with diltiazem, and in At 24 hours, 0%, 65%, and 34% of patients in the ibutilide 67% of patients treated with amiodarone; after 48 hours, (P ¼ .01), propafenone (P > .05), and rate control arms 80% in both groups were in sinus rhythm.AF recurred remained in AF. Ibutilide decreased AF duration, but recur- in 11 (37%), but 10 converted after intravenous treatment.
rence rates were 90%, 41%, and 58% in the 3 arms (P > In a retrospective review of 41 patients who developed .05). The 3 patients who did not convert all received prop- POAF after lung resection, 98% of AF disappeared within afenone. There were no differences in length of stay or a day of discharge and 85.4% required pharmacologic man- rhythm at discharge. These 2 small prospective randomized agement, but none required electrical cardioversiSinus pilot studies thus showed few differences between rate and rhythm was restored after loading with digoxin in 80%, rhythm control strategies.
11.5% after amiodarone, and 8.5% with both. All patients The absence of significant differences in the small rate except 1 were discharged in sinus rhythm. In another study versus rhythm control studies of AF after cardiac surgery of aortic surgery in 211 patients, 22 developed POAF; 16 justifies the use of either rate or rhythm control strategies spontaneously converted to sinus rhythm, 2 converted chem- in patients with new-onset POAF who are hemodynamical- ically and 1 electrically, and 3 continued in AF at discharge, ly stable. However, the high rate of spontaneous conversion but all were in sinus rhythm documented with an ECG a to sinus rhythm in the first 24 hours after onset of POAF mean of 14 months after discharThus, most patients makes it reasonable to opt for an initial rate control with new POAF after thoracic surgery can be expected to re- approach in stable patients, especially over the first 24 hours.
turn to sinus rhythm regardless of a rate or rhythm control Because anticoagulation is generally recommended in pa- tients with AF lasting longer than 48 hours, the higher risks Rate versus rhythm control strategies have been studied in of postoperative bleeding with anticoagulation can also randomized trials for non-POAFThe largest of these, justify a rhythm control approach in patients with new the Atrial Fibrillation Follow-Up Investigation of Rhythm postoperative AF that persists longer than 24 hours despite Management (AFFIRM) study,was powered to detect a a rate control approach. A rhythm control approach with difference in overall mortality, but showed no difference be- pharmacologic or electrical cardioversion may also be tween a strategy of rhythm versus rate control in the primary reasonable in patients whose ventricular rates cannot be end point of all-cause mortality, with a slight trend toward adequately controlled, or in patients who either do not better survival in the rate control arm. Secondary analyses tolerate AV nodal blockers to control ventricular rate or demonstrated no differences in quality of although who remain symptomatic or hemodynamically compro- other subanalyses demonstrated better mortality in patients mised despite control of the ventricular rate.
in sinus rhythm or on warfarin,and a functional status For the patient with stable hemodynamics and minimal substudy demonstrated better NYHA functional class in symptoms, a trial of rate control for the first 24 hours is The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines FIGURE 4. Management of the hemodynamically stable patient with new onset postoperative atrial fibrillation (POAF) of less than 48 hours duration.
WPW, Wolff-Parkinson-White syndrome; HR, heart rate; i.v., intravenous; HF, heart failure; LV, left ventricular; COPD, chronic obstructive pulmonary dis- ease; AF, atrial fibrillation; DC, direct current; TEE, transesophageal echocardiography.
generally recommended, as a high proportion will convert considered within 48 hours of onset; anticoagulation is to sinus rhythm within 24 hours using rate control or rhythm indicated for AF persisting for more than 48 hours. Alterna- control agents. Inotropes should be stopped or reduced, if tively, if the AF is well tolerated, the patient could be started clinically acceptable, fluid balance optimized, and normal on anticoagulation and rate control with plans for elective electrolyte balance maintained. Rate control may be cardioversion in 4 to 6 weeks. If AF is recurrent after cardio- achieved with intravenous esmolol or metoprolol, intrave- version, antiarrhythmic therapy with repeat DC cardiover- nous diltiazem, intravenous verapamil (although this carries sion can be continued with maintenance oral therapy for more risk for hypotension than diltiazem), digoxin (espe- 4 to 6 weeks, or a rate control approach can be adopted cially if there is hypotension or heart failure), or intravenous with anticoagulation and plans for elective cardioversion amiodarone. If AF persists, DC cardioversion may be if AF persists after 4 to 6 weeks (see ).
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines FIGURE 5. Management of the hemodynamically stable patient with new onset postoperative atrial fibrillation (POAF) of more than 48 hours duration.
WPW, Wolff-Parkinson-White syndrome; HR, heart rate; i.v., intravenous; HF, heart failure; LV, left ventricular; COPD, chronic obstructive pulmonary dis- ease; TEE, transesophageal echocardiography; LA/LAA, left atrial/left atrial appendage; DC, direct current; AF, atrial fibrillation.
bpm) for patients who develop POAFwith 5.4. Medical management of patients with new-onset 5.4.1.1.1. Caution should be used with patients 5.4.1. Rate control recommendations with hypotension, LV dysfunction, orheart failure(LOE B).
5.4.1.1. Intravenous administration of b-blockers (eg, esmolol or metoprolol) or non–dihy- 5.4.1.2. Combination use of AV nodal blocking agents, such as b-blockers (eg, esmolol (diltiazem or verapamil) is recommended or metoprolol), non–dihydropyridine cal- to achieve rate control (heart rate 110 cium channel antagonists (eg, diltiazem The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines or verapamil), or digoxin, can be useful The use of digoxin is generally less effective in the acute to control heart rates when a single postoperative high catecholaminergic state, and it has a agent fails to control rates of POAF.
slower onset of action. But in the face of hypotension, The choice should be individualized and digoxin may be the treatment of choice. b-Blockers and doses modified to avoid bradycardia calcium channel blockers have been shown to be more effective at controlling ventricular rates with shorter times 5.4.1.3. For patients with hypotension, heart failure, to effect than digoxin. Tisdale and colleaguescompared or LV dysfunction, or when other measures intravenous diltiazem (n ¼ 20) versus digoxin (n ¼ 20).
are unsuccessful or contraindicated, intrave- The end points included ventricular rate control, defined nous amiodarone can be useful for control as a 20% or greater decrease in pretreatment ventricular of heart rate. Amiodarone could result in rate and postoperative length of stay. Intravenous diltiazem conversion to sinus rhythm, and if it is achieved rate control within a mean of 10 minutes initiated after 48 hours of AF, both tran- compared with 352 minutes with digoxin (P <.0001). At sesophageal echocardiography (TEE) when 2 and 6 hours, successful rate control was higher in the dil- possible, to rule out left atrial (LA)/LA tiazem group, but by 24 hours there was no difference, as conversion to sinus rhythm occurred in 55% on diltiazem and 65% on digoxin. There was no difference in postoper- ative length of stay. However, digoxin may be particularly useful in patients with heart failure, LV dysfunction, orhypotension, or in combination with other agents. The 5.4.1.4. For patients with heart failure, LV dysfunc- addition of digoxin might also facilitate a lower dose of tion, or hypotension, intravenous digoxin b-blockers or calcium channel blockers in patients with hy- may be considered for rate control of potension. Combination use of b-blockers, calcium channel POAF(LOE B).
blockers, or digoxin can be attempted in patients with rapid rates refractory to monotherapy, but caution should be exer- 5.4.1.5. For patients with ventricular preexcitation cised with dosage modification to avoid hypotension and bradycardia, including pauses on termination of AF.
and POAF, use of AV nodal blocking It should be noted that in the presence of ventricular agents, such as b-blockers (eg, esmolol or metoprolol), intravenous amiodarone, nodal blocking agents, such as calcium channel blockers, b-blockers, digoxin, and intravenous amiodarone may antagonists (eg, diltiazem or verapamil), potentiate rapid conduction through the accessory atrioven- tricular pathway due to removal of concealed conduction from the AV node. Digoxin may also shorten the AV nodeeffective refractory period within the accessory pathway.
For these patients, AV nodal blocking agents should be Achieving control of ventricular rates in AF is a first-line avoided, and antiarrhythmic medication (intravenous ibuti- approach to patients with POAF after thoracic surgery.
lide, amiodarone, or procainamide) considered.
This may be achieved using intravenous or oral AV nodal Amiodarone has also been used for ventricular rate blocking agents, but intravenous b-blockers or non–dihy- control. However, as its antiarrhythmic properties could dropyridine calcium channel blockers (eg, diltiazem, lead to conversion of AF to sinus rhythm, caution should verapamil) can often achieve more rapid rate control than be exercised if amiodarone is initiated after 24 to 48 hours oral agents. Choice of agents is usually based on comorbid- after the onset of AF, as there is a possibility that the AF ities. b-Blockers have often been first-line therapy for ven- could convert to sinus rhythm with the attendant risk of tricular rate control after cardiac surgery, and may be thromboembolism. In these circumstances, TEE should be preferred over calcium channel blockers in patients with considered to exclude left atrial or left atrial appendage coronary disease. Calcium channel blockers are preferred thrombi before initiation of amiodarone.
in patients with bronchospasm limiting consideration of Parameters for optimal control of ventricular rates dur- b-blockers, but should be avoided in patients with heart ing AF remain controversial. The RACE II studevalu- failure or severe LV dysfunction. Diltiazem is often as ated a lenient (resting heart rate <110 bpm) versus strict effective as b-blockers with less hypotension, can be (resting heart rate <80 bpm) rate control strategy in 614 titrated as a continuous infusion, and has a greater margin patients with permanent AF. There was no difference in of safety than verapamil, which may be limited by cardiovascular death, hospitalization for heart failure, stroke, systemic embolism, bleeding, and life-threatening The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines arrhythmic events. The mean ventricular rate in the lenient and new-onset POAF, but no hypotension or control group was 85 bpm and 76 bpm in the strict control manifestations of congestive heart failure.
group at the end of the follow-up period. Although this Serum electrolytes and QTc interval must be population is different from patients with new-onset within a normal range and patients must be POAF, more lenient rate control (to heart rate 110 closely monitored during and for at least 6 hours bpm) may be preferable to strict rate control in the postop- after the infusion if either ibutilide or procaina- erative setting when patients are prone to hemodynamic mide(LOE B).
instability or hypotension. The normal metabolic response 5.4.2.5. Intravenous ibutilide or procainamide may be to surgery is associated with an increase in catechol- considered for patients with POAF and an amines, often manifested in sinus tachycardia in the early accessory pathwa(LOE B).
perioperative period and reflected in higher ventricular rates in AF.
5.4.2.6. Flecainide and propafenone should not be used There are no data to suggest efficacy for adding magnesium to treat POAF in patients with a history of a pre- or potassium to facilitate conversion to sinus rhythm or to vious myocardial infarction, coronary artery improve rate control after thoracic surgery. However, it seems disease, and/or severe structural heart disease, reasonable to recommend maintaining normal levels.
including severe left ventricular hypertrophy,or significantly reduced left ventricular ejection 5.4.2.7. Dronedarone should not be used for treatment 5.4.2. Recommendations for the use of antiarrhythmic of POAF in patients with heart failure 5.4.2.1. Restoration of sinus rhythm with pharmaco- logic cardioversion is reasonable in patients For patients with symptomatic but hemodynamically with symptomatic, hemodynamically stable stable AF after thoracic surgery, consideration should be given to restoring sinus rhythm with pharmacologiccardiovAlthough 1 study demonstrated a 5.4.2.1.1. Intravenous amiodarone can be useful cardioversion rate of 86% with intravenous amiodarone in patients undergoing pulmonary resection for lung carcinoma,a meta-analysis that included both medi- 5.4.2.2. It is reasonable to administer antiarrhythmic cal and POAF suggested a slightly lower rate of conver- medications in an attempt to maintain sinus rhythm for patients with recurrent or refractory antiarrhythmic drugs (flecainide, propafenone) may also be considered to restore and maintain sinus rhythm.
5.4.2.2.1. Amiodarone, sotalol, flecainide, propafe- Reisinger and colleaguescompared the efficacy and none, or dofetilide can be useful to maintain safety of intravenous flecainide versus intravenous ibuti- sinus rhythm in patients with POAF, de- lide in patients with recent-onset AF and showed that pending on underlying heart disease, renal the rate of cardioversion was similar (56% vs 50%, P > status and other comorbidities (see .05). However, it should be appreciated that the intrave- nous form of flecainide is not available in the United States and an oral loading dose of flecainide (and propafe- 5.4.2.3. Flecainide or propafenone may be considered for none) would be required to restore sinus rhythm pharmacologic cardioversion of POAF and It is usually customary to combine flecainide and maintenance of sinus rhythm if the patient propafenone with AV nodal blocking agents to prevent has had no previous history of myocardial 1:1 atrial flutter and rapid ventricular conduction.
infarction, coronary artery disease, impaired LV function, significant LV hypertrophy, or moderate efficacy at restoring sinus rhythm.However, it is only available in an intravenous form and it necessi- moderate or greater. These agents may need to tates close monitoring of serum electrolytes and QTc. Pa- be combined with an AV nodal blocking tients must be monitored during and after intravenous agen(LOE C).
ibutilide for at least 6 hours.
5.4.2.4. Intravenous ibutilide or procainamide may be For patients with recurrent symptomatic POAF, it is considered for pharmacologic conversion of reasonable to not only restore sinus rhythm but also POAF for patients with structural heart disease consider maintaining sinus rhythm with antiarrhythmic The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines FIGURE 6. A, Antiarrhythmic drugs recommended for pharmacologic cardioversion of postoperative atrial fibrillation (POAF). B, Antiarrhythmic drugs recommended for maintenance of sinus rhythm after cardioversion of POAF. MI, Myocardial infarction; CAD, coronary artery disease; LV, left ventricular; AV, atrioventricular; CHF, congestive heart failure; HF, heart failure.
drugs. Although many membrane active drugs (amio- function. Overall, the selection of antiarrhythmic drugs darone, sotalol, flecainide, propafenone, dofetilide, or to maintain sinus rhythm after thoracic surgery is dronedarone) have been shown to prevent recurrences similar to that outlined in the recently published AF of AF in both POAF and in the nonoperative setting guidelines for the management of nonoperative AF.
with variable efficacy, the choice of antiarrhythmic A review of antiarrhythmic drugs, their side effects, drug is very much governed by associated comordibi- and interactions are outlined in section and in ties, such as structural heart disease and impaired renal The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines 5.5. Nonpharmacologic management of POAF 5.6.2. During the first 48 hours after the onset of POAF, 5.5.1. Recommendations for DC cardioversion for the need for anticoagulation before and after DC stable patients with POAF cardioversion may be based on the patient's risk of thromboembolism (CHA2DS2-VASc score; 5.5.1.1. DC cardioversion is recommended for symp- ) balanced by the risk of postoperative bleeding(LOE C).
compromised patients with POAF if they 5.6.3. For POAF lasting longer than 48 hours, as an do not respond promptly to pharmacologic alternative to 3 weeks of therapeutic anticoagu- lation before cardioversion of POAF, it is rates(LOE C).
5.5.1.2. DC cardioversion is recommended for patients thrombus in the LA or LA appendage, preferably without hemodynamic instability when symp- with full anticoagulation at the time of TEE in toms of AF are unacceptable to the patient or anticipation of DC cardioversion after the when rapid ventricular rates do not respond to pharmacologic measures(LOE C).
5.6.3.1. For patients with no identifiable thrombus, 5.5.1.3. DC cardioversion can be a reasonable ediately after the TEE examination if thera- Anticoagulation should continue for at least 5.5.1.4. Pretreatment with an antiarrhythmic drug can 4 additional weeks although the benefits be useful to enhance the success of DC cardio- must be weighed against the risk of version (as described in section 5.2.2.1.1) and to prevent recurrent A(LOE B).
5.6.4. For POAF lasting longer than 48 hours in patients 5.5.1.5. Caution is advised for patients with preopera- who are not candidates for TEE (eg, after esoph- tive or unknown sinus node dysfunction or ageal surgery), an initial rate control strategy with patients receiving significant doses of combined with therapeutic anticoagulation using rate controlling medications, as significant warfarin (aiming for INR 2.0-3.0), a direct pauses can occur after DC cardioversion. For thrombin inhibitor (eg, dabigatran), factor Xa in- those patients, external pacing may be required hibitor (eg, rivaroxaban, apixaban), or LMWH is and should be readily available (LOE C).
recommended for at least 3 weeks before and 4 5.5.1.6. It is reasonable to repeat DC cardiover- weeks after cardioversion (LOE C).
sion, after administration of an antiar- 5.6.5. Anticoagulation recommendations for cardiover- rhythmic medication, for patients who sion of atrial flutter are similar to those for atrial relapse to AF after successful cardiover- 5.6.6. For patients with an identified thrombus, cardio- reasonable considerations for selecting DC version should not be performed until a longer period of anticoagulation is achieved (usually at 5.6. Recommendations for prevention of thromboembo- least 3 weeks) and in accordance with established lism for patients with stable atrial fibrillation/flutter AF g(LOE B).
undergoing DC cardioversion Electrical DC cardioversion is recommended for new- 5.6.1. For stable patients with POAF of 48 hours dura- onset POAF that is associated with unstable hemodynamics.
tion or longer, anticoagulation (with warfarin DC cardioversion should be performed under deep for internationalized normalized ratio [INR] 2.0- conscious sedation with R-wave synchronized shocks.
3.0, an NOAC or low molecular weight heparin Biphasic waveform shocks are preferred over monophasic [LMWH]) is recommended for at least 3 weeks waveforms, which can require higher defibrillation energies before and 4 weeks after cardioversion, regardless for success.
of the method (electrical or pharmacologic) used Commonly, rate control is attempted for at least the first to restore sinus rhythm(LOE B).
24 hours, because up to 80% of patients may spontaneously The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines convert with rate control alone (see rate control agents in an identified thrombus, cardioversion should be deferred For patients with persistent AF and significant until a longer period of anticoagulation is achieved and in symptoms despite attempts to control ventricular response, accordance with established AF guidelines. The TEE- pharmacologic or electrical cardioversion can be consid- guided cardioversion approach is supported by results of ered. When AF nears 48 hours in duration, such pharmaco- the ACUTE trial,which enrolled 1222 patients with logic or electrical cardioversion may be reasonable, AF of greater than 2 days duration and randomized them particularly in patients who are at high risk for bleeding, to TEE-guided cardioversion versus warfarin anticoagula- to avoid anticoagulation that would otherwise be indicated tion for at least 3 weeks before cardioversion. There was for AF persisting longer than 48 hours (see section 5.3.1).
no difference between the groups in the rate of embolic Pretreatment with an antiarrhythmic drug (see ) events, but the rate of hemorrhagic events was lower in can be useful to enhance the success of electrical cardiover- the TEE group. Exclusion of left atrial thrombus by TEE sion and prevent recurrent AF. However, this requires some does not preclude thromboembolism in the absence of ther- caution if the preoperative or current status of sinus node apeutic anticoagulation. Black and colleaguesreported function is unknown. If sinus node dysfunction is present, 17 patients with nonvalvular AF who had embolic events because most antiarrhythmic drugs suppress sinus node 2 hours to 7 days after cardioversion despite a TEE showing function, successful cardioversion can be associated with no LA thrombus. None of the patients were on therapeutic initial prolonged asystole and/or prolonged hypotension.
anticoagulation at the time of the embolism. Thus, the In such patients, readiness for external pacing should be TEE-guided cardioversion strategy should be coupled anticipated. If the status of sinus node function is unknown, with therapeutic anticoagulation at the time of and after car- proceeding to electrical cardioversion without administra- dioversion for patients in whom the duration of AF is greater than 48 hours.
It is recognized that in the thoracic surgery population, For recurrent AF after initial conversion to sinus rhythm, some patients will not be candidates for TEE because of cardioversion may be considered, often after initiation of an esophageal procedures, including esophagectomy (or antiarrhythmic drug to prevent further recurrences. For those with esophageal pathology). In these patients, if recurrent or refractory AF, evaluation for potential trig- AF duration is greater than 48 hours, an initial rate control gering causes should be investigated. These include approach is reasonable with therapeutic anticoagulation, bleeding, pulmonary embolism, pneumothorax, pericardial using warfarin (INR 2.0-3.0), a direct thrombin inhibitor processes, airway issues, myocardial ischemia, infection, (eg, dabigatran), factor Xa inhibitor (eg, rivaroxaban, sepsis, or use of catecholaminergic inotropes. Should AF apixaban), or LMWH, recommended for at least 3 weeks manifest as frequent paroxysms with intervening episodes before and 4 weeks after cardioversion. Cardiac computed of sinus rhythm, electrical cardioversion is not recommen- tomography has been used to assess for LAA thrombus, ded, unless AF becomes persistent, because AF is likely predominantly before AF catheter ablation. A recent to recur after cardioversion. In these situations, an antiar- rhythmic drug may be beneficial.
TEE, especially when delayed images were acquired, For AF longer than 48 hours in duration, anticoagulation with a sensitivity and specificity of 96% and 92% with warfarin (INR 2.0-3.0), a direct thrombin inhibitor (eg, compared with TEE, a positive predictive value of 41%, dabigatran), factor Xa inhibitor (eg, rivaroxaban, apixaban), and a negative predictive value of 99%. As clinical out- or LMWH is recommended for at least 3 weeks before and 4 comes studies are needed to assess its clinical usefulness weeks after cardioversion, as in other patients with AF. As for cardioversion, in the case of patients with esophageal an alternative to anticoagulation before cardioversion of surgical procedures precluding TEE, the rate control strat- AF, TEE may be performed in search of thrombus in the egy with deferred cardioversion until at least 3 weeks of LA or LAA. It is preferable to perform the TEE on full therapeutic anticoagulation is achieved seems reasonable.
anticoagulation with heparin or therapeutic levels of oral For selected patients, cardiac computed tomography may anticoagulants, and then to perform electrical or pharmaco- be of some value.
logic cardioversion immediately afterwards on therapeuticanticoagulation. This is preferred over performing a TEE while off anticoagulation or on subtherapeutic anticoagula- 5.7. Recommendation for electrophysiology catheter tion, because a thrombus may form between the time of TEE and full anticoagulation. In patients with no identifi- able thrombus, cardioversion is reasonable immediatelyafter TEE on therapeutic anticoagulation, with anticoagula- 5.7.1. Catheter or surgical ablation of AF is not recom- tion continued for at least 4 weeks afterwards, as for pa- mended for management of patients with postop- tients undergoing elective cardioversion. For patients with erative AF after thoracic surgery (LOE C).
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines model for patients likely to develop POAF. Procedures such Catheter ablation of AF is a well-established and as PVI and/or LAA resection or ligation are not routinely commonly used therapeutic option for managing patients practiced for the prevention of POAF in cardiac surgery, with symptomatic AFAt the present time, catheter where the exposure allows such procedures to be performed ablation of AF plays no role in the management of patients who develop AF in the early postoperative It is well known that the most POAF is self-limiting in 4 to 6 weeks. For persistent AF beyond that time or important considerations. First, all patients who undergo POAF requiring long-term anticoagulation, patients should catheter ablation of AF must be anticoagulated for a be referred to a cardiologist/cardiac electrophysiologist for minimum of 2 months after ablation. Patients who cannot future management according to general AF guidelines. If be anticoagulated continuously for 2 months are not such patients are intolerant of antiarrhythmic medications, considered to be candidates for ablation. Second, catheter a catheter-based ablation procedure may be offered accord- ablation of AF is a complex and lengthy procedure (3-6 ing to the existing guidelines for AF ablation. A surgical hours) that is most commonly performed under general ablation procedure can be offered in the rare instance of a anesthesia. Third, it is common for AF to recur in the 2- patient requiring a cardiac surgical procedure. A full PVI to 3-month postablation healing phase. This reflects the or LA maze and possible right atrial maze procedure may presence of considerable inflammation and lesion matura- be performed. In addition, a LAA exclusion procedure tion that occurs after ablation. The presence of these healing could be also performed. If such patients are intolerant of phase arrhythmias means that AF ablation is an inappro- long-term anticoagulation, LAA exclusion could be priate strategy for the control of acute, symptomatic AF, such as occurs in the postoperative setting. Fourth, the effi-cacy of catheter ablation is modest. In optimal candidates 6. Management of the Patient With Preexisting AF for the procedure with paroxysmal AF who are otherwise Patients with preexisting AF represent a high-risk healthy, the single procedure success rate at 12 months is population for stroke, heart failure, and other POAF- 60% to 80%. Late recurrences after 12 months of follow- related complications. Some may have valvular heart dis- up are common. Fifth, AF ablation is associated with a sig- ease. The management of their antiarrhythmic medications nificant risk of complications. For more information and their perioperative anticoagulation may pose a chal- regarding the technique, risks, indications, and outcomes of AF ablation, please refer to the 2012 Heart Rhythm So-ciety/European Society of Cardiology/European Cardiac Arrhythmia Society Expert Consensus Statement on Cath-eter and Surgical Ablation of Atrial Fibrilla 6.1. Criteria for obtaining cardiology consult for preop- 5.8. Surgical and interventional treatment options 6.1.1. Preoperative cardiology consult can be useful 5.8.1. Recommendations for preexisting AF for patients with preoperative AF that is either newly diagnosed or persistent and symptomatic 5.8.1.1. Preexisting AF should be managed accord- ing to existing guidelines for non–postopera- 6.2. Perioperative management of anticoagulation for tive AF (see section 6).
patients on long-term (warfarin or NOAC) anticoagu-lation.
Preexisting AF should be treated according to the exist- ing guidelines for nonsurgical AFIn the rare situation 6.2.1. Decisions regarding the duration of interruption where a patient cannot be treated with anticoagulation, of anticoagulation and/or the need for periopera- consideration could be given to intraoperative LAA tive heparin bridging should be based on the pa- resection or ligation. This could only be done if the tient's stroke risk profile (based on their patient is undergoing a left thoracotomy procedure.
CHA2DS2-VASc score) (LOE C).
Regarding PVI procedures, a complete bilateral procedure can only be performed in the rare situation when bilateral 6.2.2. For patients who have a high stroke risk (based on thoracotomies are performed or if a clam-shell incision is their CHA2DS2-VASc score history of stroke, or presence of a mechanical For new-onset AF after thoracic surgery, an intraopera- heart valve, perioperative bridging with a short- tive procedure is not indicated based solely on a prediction acting anticoagulant (ie, enoxaparin) is reasonable The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines FIGURE 7. Algorithm for the management of patients with preoperative atrial fibrillation (AF). Preop, Preoperative; CHF, congestive heart failure; SOB, shortness of breath; EF, ejection fraction.
for patients with estimated glomerular filtration preoperatively identified AF who meet 1 or rate greater than 50% when warfarin anticoagula- more of the following criteria: tion is withheld (LOE C).
6.4.1.1. Ejection fraction 45% or less or a diagnosis of systolic heart failure or cardiomyopathy 6.2.3. Short-term withdrawal of anticoagulation without 6.4.1.2. Discharged on a new rate control and/or bridging may be considered for those patients rhythm control agent(s) who are on anticoagulation preoperatively as part 6.4.1.3. Dose of a home rhythm control agent(s) was of their treatment for persistent AF but have a adjusted while an inpatient CHA2DS2-VASc score less than 2, have not had 6.4.1.4. Discharged on a new anticoagulant (paren- heart failure, have an ejection fraction greater teral and/or oral) (LOE C) than 35%, and/or for whom bridging anticoagula-tion would be burdensome or otherwise undesirable In the Rivaroxaban Once Daily Oral direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Pre- 6.3. Postoperative resumption of anticoagulation vention of Stroke and Embolism in Atrial Fibrillation (ROCKET AF) trial comparing warfarin and rivaroxabanfor the prevention of thromboembolism in nonvalvular 6.3.1. If anticoagulation is interrupted, the duration AF, 33% of the 14,236 patients had a temporary interrup- should be minimized. It is reasonable to base tion of anticoagulation of greater than or equal to 3 days decisions about the duration of interruption and (mean duration 5 days). Eighty-one percent of these pa- the time of resumption of anticoagulation on the tients had persistent AF and more than 99% had a patient's stroke risk profile (CHA2DS2-VASc score) weighed against the risk of postoperative 2 score greater than or equal to 2 (mean Fifty percent of the patients had a history of stroke or tran- bleeding (LOE C).
sient ischemic attack and 62% had a history of congestiveheart failure. Forty percent of the temporary interruptionswere for procedures, although only 14% were for abdom- 6.4. Postoperative follow-up inal, thoracic, orthopedic, or cardiac procedures. The 30- day stroke or systemic embolism rate of 0.36%, although 6.4.1. It is reasonable to consider postoperative follow- similar in both the rivaroxiban and warfarin groups up with a cardiology specialist for patients with (0.30% vs 0.41%; hazard ratio, 0.74; confidence interval, The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines 0.36-1.5; P ¼ .4) was higher than the overall rate of 2.2%/ The full role of perioperative bridging will be further year throughout the study.
elucidated in 2 ongoing randomized trials: BRIDGE (Effec- In the ROCKET AF studybridging of anticoagula- tiveness of Bridging Anticoagulation for Surgery; tion for temporary interruptions was tracked, and only and PERIOP-2 (a double-blind, random- 6% of the temporary interruptions were bridged, 98.6% ized, controlled trial of postoperative LMWH bridging ther- with LMWH and 1.4% with fondaparinux. The bridge apy vs placebo bridging therapy for patients who are at high group was slightly older (74 vs 73 years; P ¼ .019), had risk for arteriothromboembolism; ).
a fractionally higher CHADS2 score (3.52 vs 3.40;P ¼ .0094), more often had diabetes (48% vs 41%;P ¼ .0049), and had more temporary interruptions. The 7. Management of Anticoagulation for New-Onset rates of previous stroke, transient ischemic attack, and congestive heart failure in the bridge group were similar In order to minimize the risk of perioperative bleeding to the rates in the nonbridge group. The bridged patients whereas/although providing sufficient protection from the had only 1 stroke/embolic event, a 30-day rate of POAF-related strokes a careful evaluation of the patients' 0.17%, whereas the nonbridge group 30-day event rate stroke risk is essential. The recently approved novel oral was 0.37%. Major bleeding was similar in both groups, anticoagulants (NOAC; direct thrombin inhibitors and but the bridged group had a higher incidence of major/ anti-Factor Xa agents) offer alternatives to warfarin, and nonmajor clinically relevant bleeding (4.83% vs 3.02%).
are gaining popularity in the community for the long-term Because of the small numbers and variable reasons for management of AF-related anticoagulation.
bridging, statistical significance was not calculated. These data raise the concern that discontinuation of anticoagula-tion may expose patients to a small but significant risk of 7.1. For the prevention of strokes for patients who develop Additional data that suggest the importance of consid- POAF lasting longer than 48 hours, it is recommended ering the risk of stroke when contemplating the temporary that antithrombotic medications are administered interruption of anticoagulation were derived from data similarly to nonsurgical patients (). The deci- obtained after the closing of the ROCKET AF trial. After sion to initiate therapy should be based on the benefit cessation of the study medication, the patients on rivaroxi- of reducing stroke risk versus the risk of bleeding in ban required an average of 13 days to achieve a therapeutic the postoperative period(LOE A).
INR, whereas the patients receiving warfarin requiredonly 3 days. The rivaroxiban group had 22 embolic events 7.1.1. For effective anticoagulation, an INR range of and the warfarin group had 7 events over 31 days of 2 to 3 (with a target of 2.5) for warfarin is FIGURE 8. Management of anticoagulation for postoperative atrial fibrillation (POAF) lasting longer than 48 hours. NOACs, New oral anticoagulants; INR, international normalized ratio.
The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines 7.1.2. The INR should be determined at least weekly esophagectThese latter patients may during initiation of therapy and monthly when better be managed by LMWH and/or by a TEE-guided the doses of anticoagulant and the INR are fast-track strategy to rule out an LAA thrombus and then receive DC cardioversion.Because thepotential for thromboembolism with new-onset AF de- velops early, prompt attempts to restore sinus rhythm The overall risk of a perioperative stroke in all patients within this period should be made. If the arrhythmia per- undergoing anesthesia has been estimated at 0.5% to sists beyond 24 to 48 hours, anticoagulant therapy should 0.8% in large studies of patients who have had noncardiac be considered after weighing the risk of postoperative surgery.One of these studies use the Nationwide bleeding. In a prospective study of 330 patients undergo- Inpatient Sample of 131,067 patients and determined ing anatomic lung resection, 1 of 60 patients (1.7%) with postoperative AF developed a stroke within 24 hours of onset of AF and Holter monitoring later showed that the ischemic stroke was 0.6%, which rose to 0.8% for initial 12 hours of AF were asymptOthers have patients more than 65 years of age.Risk factors associ- questioned the 48-hour window and suggest that it might ated with perioperative stroke in that study were renal dis- be reasonable to start anticoagulation therapy in the first ease, atrial fibrillation, history of stroke, and cardiac 48 hour if multiple risk factors for stroke are present.
valvular disease.Mortality in patients who developed They further suggest that a TEE-guided strategy may a stroke after lung resection was 33% compared with prove useful in situations where both the risk of stroke 3.2% in those who did not.The reported incidence of and the risk of postoperative bleeding pose a dilemma stroke or transient neurologic injury of 1.6% to 3.3% after regardless of the fact that AF was not present for 48 cardiac operations is consistently greater for patients who develop persistent postoperative AF compared with 0.2% The goal of anticoagulation should balance the risk of to 1.4% for those without AFIt has been established stroke and the risk of bleeding. The range should be that oral anticoagulation with warfarin is associated with optimal for adequate stroke prevention but at the same 60% to 70% reduction from the 4% to 5% overall risk time should be at the minimal bleeding threshold. In AF, of ischemic stroke per year in patients with persistent or an INR range of 2.0 to 3.0 with a target of 2.5 should chronic nonvalvular AF not receiving warfari fulfill this requirement.Randomized control studies Depending of the type of surgery (total hip replacement, have shown that warfarin therapy with an INR of 2.0 to hemicolectomy, or lung resection), 12% to 33% of 3.0 was associated with improved outcome compared arterial thromboembolic events are fatal and more than with aspirin.Hylek and colleaguesretrospectively 40% result in serious permanent disabilityOn the studied 13,559 patients with nonvalvular AF and showed other hand, 3% of episodes of major postoperative that an INR less than 2 at admission was associated with bleeding are fatal but most patients make a full an increased number of strokes. Recently, in the Random- recovery. As many as 50% of bleeding episodes require ized Evaluation of Long-Term Anticoagulation Therapy a reoperation.A retrospective study of patients who (RE-LY) trial, a randomized controlled trial, compared developed new-onset AF after general thoracic surgery the outcomes of warfarin versus dabigatran treatment in compared patients that received some form of antithrom- AF patients. Warfarin was managed with a target INR of botic therapy for AF with those who did not receive anti- 2.0 to 3.0 and the maximum interval between INR tests coagulation and found that the stoke rate among patients was 4 weeks. They used an algorithm to manage the who were anticoagulated was 2.2% compared with INR; for example, þ15% dose/wk increase for INR less 0.6%, and these patients had a greater incidence of than 1.5, þ10% dose/wk increase for INR from 1.5 to bleeding episodesIn that study, patients who were anti- 1.99, 10% dose/wk decrease for INR from 3.01 to coagulated had more comorbidities and greater risk score 4.00. Implementation of this algorithm resulted in an in- for stroke. Whether individuals require short-term antico- crease in time in therapeutic range (TTR).
agulation must be individualized for each patient based on For in-hospital patients on warfarin, INR is measured the intrinsic risk for thromboembolism and the risk of every day until it is therapeutic. For outpatient follow-up, bleedingFor most types of surgery, initiation or INR is followed every few days until it reaches the stable resumption of warfarin can be undertaken 12 to 24 therapeutic target, then the interval can be prolonged as hours after surgery unless the patient is at special risk long as 4 to 6 weeks. The frequency of follow-up depends for bleeding, such as those with a low platelet count, on patient compliance, drug and food interactions, interrup- prolonged excessive chest drainage, or those who might tion for surgical procedures, and existence of other comor- require an invasive procedure within days or weeks of Its frequency should be increased when discharge (eg, developing an anastomotic leak after switching over to another type of anticoagulant, such as a The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines FIGURE 9. Considerations for the management of anticoagulation within the first 48 hours of postoperative atrial fibrillation (POAF).
heparin bridge. Pengo and colleaguesrandomized 124 ered for anticoagulation because the benefits patients to 4-week-interval and 6-week-interval follow-up for those patients are less of INR testing for patients with prosthetic mechanical valves and showed that there was no difference in their 7.2.3. If not precluded by concerns for bleeding, antico- TTR. In a randomized study of 250 patients receiving agulation is also recommended when conversion warfarin, followed at 4-week intervals versus 12-week in- to sinus rhythm is attempted by (DC or chemical) tervals with phone follow-up every 4 weeks, Schulman cardioversion (as see section and colleashowed that the 12-week-interval group had similar TTR and bleeding/embolic events. The Amer- ican College of Chest Physicians recommends a follow-up Since the analysis by the Atrial Fibrillation Investigators interval of up to 12 weeks if INR is stable.
(AFI) of the first 5 prospective, randomized, clinical trialscomparing oral anticoagulants with placebo and sometimeswith aspirin, we have learned that not all patients with atrial fibrillation have the same risk of The AFI demonstrated that stroke risk may be stratified by several 7.2. Anticoagulation within the first 48 hours of POAF factors, including a previous thromboembolic event or ) should be considered based on theCHA2DS2-VASc risk score () of the patientfor stroke weighed against the risk of postoperativebleeding (LOE C).
7.2.1. For risk assessment, the following may serve as a guide: CHA2DS2-VASc risk score forstro(LOE A): S ¼ 0: no anticoagulation recommendedS ¼ 1: anticoagulation should be considered if its benefits outweigh the risk of bleeding S ¼ 2: anticoagulation is highly recommended if FIGURE 10. Stroke risk stratification in atrial fibrillation. From: Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk stratifi- its benefits outweigh the risk of bleeding cation for predicting stroke and thromboembolism in atrial fibrillation us- 7.2.2. The presence of impaired renal function should ing a novel risk factor-based approach: the euro heart survey on atrial weigh in favor of anticoagulation. Caution should fibrillation. Chest. 2010;137:263-72. HTN, Hypertension; MI, myocardial be exercised when patients on dialysis are consid- infarction; PAD, peripheral arterial disease.
The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines not require anticoagulation. For a CHA2DS2-VASc score congestive heart failure, poor left ventricular function, of 1, oral anticoagulation could be considered, and for a and age 65 years or older. Stroke risk was further CHA2DS2-VASc of 2 or more, oral anticoagulation is stratified to mild, moderate, and severe categories. Other generally indicatAlthough the recently published risks such as coronary artery disease, peripheral vascular American College of Cardiology Foundation/American disease, gender, thyrotoxicosis, rheumatic mitral valve Heart Association/Heart Rhythm Society AF guidelines disease, and hypertrophic cardiomyopathy were also suggested that for a CHADS-VASC score of 1, it is rea- important risk markers to consider.
sonable to consider not using antithrombotic therapy or To translate these risks derived from group data to aspirin if the burden of cardiovascular disease is otherwise the individual, there are now several stroke risk stratifica- lowHowever, most thoracic surgical patients who tion schemes available. Initially, most guidelines adopted develop (or have) AF would likely have an indication the CHADS2 stroke risk stratification scheme, and it has CHA2DS2-VASc risk been widely used for many years.Recently, the scoring system. Because of its ease of use, and its wide acceptance, the use of the CHA2DS2-VASc score is () has taken prominence, having being adopted recommended for the assessment of stroke risk.
by the European Society of Cardiology, and the American Several scoring systems have been reported to identify College of Cardiology Foundation/American Heart Asso- the risk of bleeding after the initiation of anticoagula- ciation/Heart Rhythm Society guidelines.A major These scoring systems are not recommended reason for its use is that it is better at sorting out those for routine use as standard practice. However, among the patients with low stroke risks who do not really need risk factors, end-stage renal disease on hemodialysis is anticoagulation for prophylaxis and those who do. Thus, considered to pose a significant risk for bleeding when these a CHADS2 score of 0 or 1 is associated with an annual patients are anticoagulated. In a retrospective review of risk of 1.9% to 2.8%, not really small risks at all.
1626 dialyzed patients and nondialyzed patients, anticoagu- However, when applying the CHA2DS2-VASc stroke risk lation for dialyzed patients did not decrease the risk of scheme to the same patients, the CHA2DS2-VASc score stroke, but increased the incidence of bleeding episodes is anywhere from 0 to 4. A CHA2DS2-VASc score of 3 or 4 carries an indication for use of anticoagulationtherapy, whereas a CHA2DS2-VASc score of 0 does recommendation to consider the use of anticoagulationtherapy.
7.3. New oral anticoagulants (dabigatran, rivaroxiban, The recommendations using the CHA apixiban) are reasonable as an alternative to are that if the CHA warfarin (for patients who do not have a 2DS2-VASc score is 0, the patient does TABLE 8. Commonly used anticoagulants Mode of clearance Recommended doses Significant limitations Vitamin K antagonist Variable (monitor INR) Multiple food and drug interactions, need for frequent INR monitoring and dose adjustments Thrombin inhibitor 150 mg twice a day 75 mg Interaction with inhibitors of twice a day for CrCl P-gp, no established antidote, not recommended in severe renal failure Factor Xa inhibitor 20 mg daily, 15 mg daily for Interaction with inhibitors of CrCl 15-50 mL/min P-gp and CYP3A4, no established antidote, not recommended in severe Factor Xa inhibitor 5 mg twice a day 2.5 mg twice Interaction with inhibitors of a day (AF) for at least 2 of P-gp and CYP3A4, no the following: age >80 y, established antidote, not body weight<60 kg, Cr recommended in severe INR, International normalized ratio; CrCl, creatinine clearance; P-gp, P-glycoprotein; CYP3A4, cytochrome P450 3A4; AF, atrial fibrillation; Cr, serum creatinine.
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines prosthetic heart valve, hemodynamically significant investigated in the Apixaban for Prevention of Stroke in valve disease, and/or severe renal impairment or risk Subjects with Atrial Fibrillation (ARISTOTLE) noninfer- of gastrointestinal bleeding(LOE B).
iority trialPatients (n ¼ 18,201) with AF and at least 7.4. It is reasonable to continue anticoagulation therapy for 1 additional risk factor for stroke were randomized to 4 weeks after the return of sinus rhythm because of the receive apixaban 5 mg orally twice daily or warfarin possibility of slowly resolving impairment of atrial titrated to an INR of 2.0 to 3.0 for a median follow-up contraction with an associated ongoing risk for of 1.8 years. The risk of ischemic or hemorrhagic stroke or systemic embolism in the apixaban group was signifi- cantly lower than in the warfarin group, as was the inci-dence of death from any cause. The incidences of major bleeding and hemorrhagic stroke were also significantly Newer oral anticoagulant drugs have recently become available, including dabigatran, rivaroxaban, and apixa- difference between the groups in the incidence of ischemic ban. Dabigatran is an oral direct thrombin inhibitor, or uncertain type of stroke.
whereas rivaroxaban and apixaban are factor Xa inhibi- Patients who received standard anticoagulation on tors. Compared with warfarin, these agents offer the discharge from the hospital can return for cardioversion advantage of not requiring monitoring of the INR. The ef- between 3 and 12 weeks after initiation of anticoagulant ficacy of dabigatran for stroke prevention in nonsurgical, therPatients who convert to sinus rhythm but nonvalvular AF was compared with that of warfarin in are experiencing intermittent paroxysms of AF may be the RE-LY trial,which was a prospective noninferiority considered for anticoagulation for 1 month after the return study that randomized 18,113 patients into 3 groups: dabi- of sinus rhythm because it has been shown that impaired gatran 110 mg twice daily or dabigatran 150 mg twice atrial mechanical function can persist for several weeks daily, administered in blinded fashion, or warfarin titrated after the return of sinus rhyt to an INR of 2.0 to 3.0, administered in unblinded fashion for a median duration of 2 years. Dabigatran 150 mg twicedaily significantly reduced the risk of stroke or systemic embolism by 34% compared with warfarin. There was 7.5. New oral anticoagulants should be avoided for patients no significant difference in the incidence of death in either at risk for serious bleeding (including gastrointestinal dabigatran group compared with the warfarin-treated bleeding) as they cannot be readily reversed. However, group. There was no difference in the incidence of major their use may be recommended in situations where bleeding between the groups on warfarin and dabigatran achievement of a therapeutic INR with warfarin has 150 mg twice daily. However, there was a significantly proved to be difficult(LOE C).
lower incidence of hemorrhagic stroke in dabigatran 150mg twice daily group compared with the warfarin group.
The efficacy of rivaroxaban for reducing risk of stroke A large phase 2 randomized control study, the RE- ALIGN trial, studied patients who underwent implanta- compared with that of warfarin in the ROCKET AF tion of mechanical valve (aortic or mitral) or had In this noninferiority study, 14,264 patients with undergone implantation of mitral bileaflet valve less AF were randomized in double-blind fashion to receive ri- than 3 months before randomization.The trial was varoxaban 20 mg orally daily or warfarin titrated to an terminated because of an increase in strokes (5% vs INR of 2.0 to 3.0 for a median treatment period of 590 0%), myocardial infarction, and major bleeding (4% vs days (median follow-up, 707 days). Compared with 2%) in the dabigatran group. Currently, dabigatran is warfarin, rivaroxaban significantly reduced the risk of contraindicated and should not be used in patients with mechanical valves.
intention-to-treat analysis, there was no significant differ-ence between rivaroxaban and warfarin in the incidence 8. Recommendations for Long-Term Management of stroke or systemic embolism. There was no difference and Follow-up of Patients With Persistent New-Onset between the groups in the incidence of major and nonmajor clinically relevant bleeding. However, the inci- dence of intracranial hemorrhage and fatal bleeding was complications, and those requiring long-term management significantly lower in rivaroxaban-treated patients. The ef- of antiarrhythmics and anticoagulants are likely to ficacy of apixaban compared with that of warfarin for benefit from cardiology follow-up after their discharge stroke prevention in nonsurgical, nonvalvular AF was The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines FIGURE 11. Recommendation for the postdischarge follow-up for patient with new onset postoperative atrial fibrillation (POAF). Post-op, Postoperative; EF, ejection fraction; NSR, normal sinus rhythm; ECG, electrocardiography.
although this was as an older study, 80% were given digi-talis. The duration was between 1 and 12 days, with an 8.1. Postdischarge follow-up and management recom- average of 2 days. It has been estimated that approximately mendations for persistent new-onset POAF.
50% of episodes of POAF spontaneously convert to normal sinus rhythm within 12 hoursGiven the relatively short 8.1.1. For patients who have a complicated in-hospital duration of POAF in most cases, it is unclear when the first course related to their POAF, who have underlying postoperative visit to the surgeon in uncomplicated cases structural heart disease, or who experience should be. The 2010 guidelines of the Canadian Cardiovas- sequelae of AF, such as myocardial infarction or cular Society recommend that medical management of AF decreased LVEF, follow-up with cardiology should and anticoagulation should be reassessed at 6 to 12 weeks be arranged at the time of discharge (LOE C).
postoperatively, although this was intended primarily for cardiac surgical patients. This was considered a strong 8.1.2. Patients with well-controlled new-onset POAF recommendation, with moderate evidence but no reference (either converted to sinus rhythm or with good was giKowey and colleaguesreported a retro- rate control) may be seen in routine follow-up spective analysis of 116 patients with POAF after coronary by the surgical team without cardiology follow- bypass surgery. There were 36 patients treated with antiar- rhythmic and rate control drugs compared with 76 treated with rate control agents alone. Only 1 patient in each group Most cases of POAF are self-limiting and even when pre- was still in AF at 6-week follow-up. In another study, Izhar sent at discharge, will have resolved by the time of follow- and collearandomized 129 patients after coronary up. There is little literature regarding postdischarge risks for bypass surgery who had converted to sinus rhythm before general thoracic patients specifically. Recommendations discharge to 1, 3, or 6 weeks of antiarrhythmic therapy.
regarding cardiology follow-up for complicated patients There was no difference in the rate of recurrent AF with seem self-evident. The appropriate timing for cardiology 0, 2, and 0 patients in the 3 groups. Most of the patients follow-up should be individualized before discharge. For were managed with amiodarone. On this basis, it seems uncomplicated patients, there is some evidence for guid- reasonable that the timing of routine surgical follow-up ance. After lung resection, Rena and colleaguesdemon- should be dictated by surgical considerations, and the pres- strated that 98% of POAF resolved after discharge, ence or absence of AF assessed at that time.
The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines Despite the relatively self-limiting nature of most cases of POAF, the long-term significance of a single episode of 8.3. Management of anticoagulation POAF is unknoAhlsson and colleagiesstudied the late outcome of patients who developed POAF after cor-onary bypass surgery and found that the development of AF 8.3.1. For patients who are started on anticoagulants, the was a risk factor for late mortality. Whether this applies to anticoagulation should continue for a minimum patients after noncardiac thoracic surgery is unknown.
of 4 weeks after return to normal sinus rhythm However, it seems prudent to ensure communication with is documented (LOE C).
the primary care physician for vigilant follow-up of cardio- vascular and AF risk factors.
8.3.2. More prolonged anticoagulation (longer than 4 weeks after return to normal sinus rhythm) can be beneficial in the presence of stroke riskfactors (CHA 8.2. Management of antiarrhythmic medications 2DS2-VASc score) or if the patient had a previous stroke. The concomitant presence of mild or moderately impaired kidney function 8.2.1. For patients who have converted to sinus rhythm weighs in favor of a longer period of anticoagula- before hospital discharge, it is reasonable to tion (LOE B).
consider discontinuation of antiarrhythmic medi-cations 4 weeks after ECG documented return of 8.4. Recommendations for long-term management of normal sinus rhythm or at the first postoperative new-onset persistent POAF visit (usually 2-6 weeks after discharge) (LOE C).
8.4.1. Patients with new-onset POAF persisting for or 8.2.2. For patients with new-onset POAF who were dis- recurring after 4 to 6 weeks (or at the time of charged in AF but who are in normal sinus rhythm the first postoperative visit) can benefit from (ECG confirmed) at the first postoperative visit, it referral to a cardiologist for long-term manage- may be reasonable to instruct the patients to ment of stroke risk as well as antiarrhythmic or self discontinue the antiarrhythmic medications anticoagulant medications (LOE C).
4 weeks after the visit if no signs of AF recur(LOE C).
The ideal duration of anticoagulation after POAF is un- known. European guidelines have concluded that there is There is no clear evidence to guide duration of antiar- insufficient evidence to make any recommendation.
rhythmic therapy after noncardiac thoracic surgery. Landy- Others have concluded that it is reasonable to continue anti- coagulation for 4 weeks, on the basis that atrial contraction underwent coronary artery bypass grafting with ambulatory is impaired long after the AF has ceased.All the current Holter monitoring, including 3 patients who developed evidence in this area is level C.
spontaneous symptomatic AF and received digitalis forrate control. Sixteen patients (group 2) continued taking RECOMMENDATIONS FOR FUTURE AATS digoxin for 8 weeks after surgery, 13 patients (group 3) discontinued digoxin treatment 5 weeks after surgery, and The task force recommends the establishment of a high- 14 patients (group 4) discontinued digoxin treatment fidelity thoracic surgery database that uses the uniform 3 weeks after surgery. Twenty-four–hour Holter monitoring definitions and monitoring strategies recommended here, indicated that asymptomatic AF was common in the treat- stratifies by surgery type, and systematically documents ment groups after digitalization just before discharge from the occurrence, duration, and complications of POAF and hospital. Recurrence of AF was rare after discharge. Yilmaz its treatment. The aim would be to develop risk prediction and colleaperformed a similar smaller study, with models, and eventually randomized interventional trials, 120 patients who had converted to sinus rhythm (pharma- for the prevention and treatment of POAF, specific to cologically or with DC cardioversion) enrolled in a prospec- thoracic surgery. This could be most readily accomplished tive randomized trial of placebo or 1 of 3 drugs (amiodarone, by enriching the STS data collection system.
verapamil, or quinidine) postdischarge. Patients underwent24-hour Holter monitoring 6 times over 9 months postopera- RECOMMENDATIONS FOR THE USE OF THE tively. Recurrent AF usually developed within 15 days of discharge. AF occurred in only 1 patient (3.33%) in group These guidelines are best used as a guide for practice and 1, and 2 each (6.66%) in each of the drug groups.
teaching. The applicability of these recommendations to the The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines individual patient should be evaluated on a case-by-case ba- sis, and only applied when clinically appropriate. In addi- tion, these guidelines can serve as a tool for uniform practices, to guide preoperative evaluations, and form the basis of large, multicenter cohort studies for the thoracic surgical community.
The task force received no financial support. AATS provided teleconferencing and covered the cost of a 1-day face-to-face conference for the participants. The members of this task force had no conflicts of interest related to any of the 88 recommendations made here; all their other poten- tial conflicts of interest were disclosed in writing A heartfelt thank you goes to Mr James M Bell for his artistic contribution in finalizing the figures and tables and to Mr Matt Ea- ton (AATS) for his role in coordinating all taskforce meetings and The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3 Clinical Guidelines The Journal of Thoracic and Cardiovascular Surgery c September 2014 Clinical Guidelines The Journal of Thoracic and Cardiovascular Surgery c Volume 148, Number 3

Source: http://i.heartabc.com/218a204826973604386ca3d9e54c7702.pdf