Atrial Fibrillation: Current Evidence and Management Strategies During the Perioperative Period

Kunal Karamchandani, MD, FCCP; Ashish K. Khanna, MD, FCCP, FCCM; Somnath Bose, MD; Rohesh J. Fernando, MD, FASE; Allan J. Walkey, MD, MSc


Anesth Analg. 2019;130(1):2-13. 

In This Article

Abstract and Introduction


Atrial fibrillation (AF) is the most common arrhythmia in the perioperative period. Previously considered a benign and self-limited entity, recent data suggest that perioperative AF is associated with considerable morbidity and mortality and may predict long-term AF and stroke risk in some patients. Despite known risk factors, AF remains largely unpredictable, especially after noncardiac surgery. As a consequence, strategies to minimize perioperative risk are mostly supportive and include avoiding potential arrhythmogenic triggers and proactively treating patient- and surgery-related factors that might precipitate AF. In addition to managing AF itself, clinicians must also address the hemodynamic perturbations that result from AF to prevent end-organ dysfunction. This review will discuss current evidence with respect to causes, risk factors, and outcomes of patients with AF, and address current controversies in the perioperative setting.


Perioperative atrial fibrillation (POAF) is common, with an estimated incidence of 2%–60% depending on the type of surgery.[1–4] The reported incidence is lower in noncardiac surgery and ranges from 4.8% following total joint replacements to 12%–19% for esophageal, thoracic, or abdominal surgery.[5–13] The true incidence of POAF in patients undergoing noncardiac surgery is likely underestimated because not all patients are monitored continuously in the postoperative period.[14] While the risk factors for POAF after cardiac surgery are attributed primarily to underlying cardiac disease and direct manipulation of the heart and pericardium, precipitating factors and mechanisms for POAF after noncardiac surgery are less well defined. Furthermore, no established risk scores predict POAF in patients after noncardiac surgery, and few, evidence-based strategies are available for POAF prevention.[15]

Although historically considered a self-limiting entity,[16] recent evidence suggests that POAF is associated with an increased overall risk of in-hospital morbidity and mortality.[2,17] In addition, new-onset POAF is an independent predictor of stroke, which is the primary driver of POAF mortality.[18,19] Currently, however, data for outcomes such as acute myocardial infarction (AMI), congestive heart failure, and acute kidney injury are sparse. Although POAF is more common after cardiac than after noncardiac surgery,[20] the burden of POAF after noncardiac surgery remains considerable. A retrospective 2014 review found that the most common acute condition predisposing to AF in hospitalized patients was noncardiac surgery.[21] AF risk should thus be assessed during perioperative evaluation, with attention to preventable AF triggers and insults during the perioperative period. Preventing AF is ideal because management of perioperative AF can be challenging. A recently conducted international survey, administered to members of the Society of Cardiovascular Anesthesiologists (SCA) and the European Association of Cardiothoracic Anesthesiologists (EACTA), found that, despite existing guidelines from the European Society of Cardiology and American College of Cardiology,[22,23] considerable practice variability exists with respect to POAF management in patients after cardiac surgery.[24] Because even fewer guidelines exist for patients undergoing noncardiac surgery, it is likely that similar or possibly greater practice variability also extends to this patient population.

With the increase in elderly patients undergoing surgery and the strong correlation between increasing age and likelihood of POAF,[25] the incidence of new-onset AF among surgical patients is expected to increase with time. Several clinical challenges thus arise: (1) what is the efficacy of current therapies targeting AF prophylaxis and to what degree does avoiding AF triggers reduce AF incidence; (2) what are the effects of POAF on short- and long-term morbidity and mortality and how aggressively should rapid ventricular rate (RVR), defined as ventricular rate ≥110/min,[26] be treated; (3) when and in whom stroke prophylaxis should be initiated; and (4) the optimal duration of rate-control therapy among patients who convert to normal sinus rhythm (NSR) before discharge. This review will briefly describe the pathophysiology of AF, discuss the challenges perioperative physicians face regarding prevention and management of POAF, highlight the short- and long-term outcomes in patients who develop POAF, suggest clinical strategies where appropriate, and suggest future directions for more appropriate perioperative management of POAF.

Pathophysiology of AF

Electrical impulses of the heart normally originate at the sinoatrial (SA) node and are conducted through the atrioventricular (AV) node and down to the bundle of His and Purkinje fibers to depolarize both ventricles. In AF, signals are not initiated at the SA node but, instead, are generated from all over the atria, resulting in uncoordinated atrial activity that is intermittently and irregularly conducted through the AV node. At the atrial level, AF is perpetuated by reentry and/or rapid focal ectopic firing.[27] Uncoordinated atrial discharge may result from an irregular atrial response to a rapidly discharging, regularly firing driver or a single localized reentry circuit.

Alternatively, fibrillatory activity may be caused by multiple functional reentry circuits varying in time and space. The Heart Rhythm society defines 3 types of AF: paroxysmal (may occur and terminate spontaneously), persistent (will not terminate without treatment), and permanent (will not terminate even with drug or electrical therapy).[28] Paroxysmal AF, the subtype most commonly seen in the perioperative setting, usually results from focal ectopic firing.[29] One current hypothesis is that the natural history of AF involves an evolution from paroxysmal to persistent to permanent forms via atrial remodeling caused by the arrhythmia itself and/or progression of underlying heart disease.[30–32] Thus, preventing perioperative AF may avoid long-term development of persistent and permanent AF.

AF may cause left ventricular dysfunction as a result of inappropriately rapid and/or irregular ventricular rhythms and reduced coronary blood flow due to decreased diastolic filling time.[33–35] In addition, coordinated atrial contraction, which contributes about 20% of left ventricular stroke volume at rest, is lost in AF and contributes to impaired cardiac output.[36] These effects trigger a spiral of adverse cardiac consequences, where AF-induced atrial hypocontractility leads to ventricular dysfunction, which leads to further atrial dilation, stretch, and remodeling that makes AF resistant to therapy (Figure 1).

Figure 1.

The interaction between atrial fibrillation and heart failure. AF indicates atrial fibrillation; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle. Reproduced with permission from Iwasaki YK, Nishida K, Kato T, Nattel S, "Atrial Fibrillation Pathophysiology: Implications for Management," Circulation, 2011;124:2264–2274.29

Causes and Risk Factors

The perioperative period is associated with many factors that predispose to the development of de novo AF, and precipitate RVR in patients with paroxysmal and chronic AF (Table 1). In most cases, many potential mechanisms and factors are involved. One 2012 review examining 139 patients with new-onset POAF found that 73% patients had at least 1 modifiable risk factor and that 45% had 2 or more risk factors.[37] Risk factors for POAF can be broadly categorized into patient and surgery related and are summarized in Table 1.

Surgery-related Risk Factors for POAF

Hypovolemia, intraoperative hypotension, anemia, trauma, and pain increase sympathetic activity, catecholamine release, heart rate, and arrhythmogenicity.[14] Operative stress by itself can nonuniformly shorten atrial refractoriness, perpetuating atrial arrhythmias.[38] Metabolic disturbances during surgery such as hypoglycemia, hypokalemia, and hypomagnesemia can also lead to development of AF.[39–41] In addition, by causing pulmonary arterial vasoconstriction hypoxemia may increase right ventricular pressure and right atrial stretch. Myocardial ischemia itself may alter atrial conduction. Excessive fluid shifts during surgery or afterward may increase intravascular volume and stretch the right atrium, also predisposing to the development of AF.[42] The type of surgery may also play a role, with thoracic surgery, abdominal surgery, and major vascular surgery being associated with the highest risk of developing POAF.[2,10,43]

Patient-related Risk Factors for POAF

Although multiple patient-related factors predisposing to POAF exist, increased age is the most important[25] and the incidence of AF increases with age.[44] Atrial fibrosis is more common in the aging heart and forms a substrate for the development of AF.[45] Race impacts the development of new-onset POAF, with African Americans having a lower risk.[2] As expected, a history of paroxysmal AF increases a patient's risk for development of POAF and conservative estimate suggests that 27%–67% of patients who develop POAF have a preoperative history of paroxysmal AF.[2,7]

Other patient comorbidities, including preexisting congestive heart failure,[2,10,46] ischemic heart disease,[2,47] hypertension,[2] chronic renal failure,[47] sepsis,[5,10,43,48–50] shock,[49] asthma,[10] thyroid disorders,[51,52] and valvular disease,[10] are all associated with POAF. Obstructive sleep apnea (OSA) is an independent risk factor for AF and the associated nocturnal hypoxemia can precipitate new-onset AF in hospitalized patients.[53,54]

Impact of POAF

Patients who develop POAF have higher in-hospital mortality, longer hospital lengths of stay, and increased hospitalization costs.[2,10] Those with preexisting AF who develop POAF have similar outcomes compared with patients who develop POAF de novo.[2] Short-term postoperative morbidity and mortality are similarly worse in patients who develop POAF. Analysis of the 2008 PeriOperative ISchemic Evaluation (POISE) trial suggested that patients who developed new clinically important POAF had a higher risk of stroke within 30 days after surgery (odds ratio [OR], 4.35 [1.87–10.12]).[55] A 2012 review of administrative data from nearly 370,000 noncardiac surgical patients across 375 US hospitals determined that those with POAF had higher mortality (OR, 1.72; P < .001), increased length of stay (adjusted relative difference of +24%, P < .001), and a higher cost of hospitalization (adjusted difference of +$4177, P < .001).[2] Similarly, in patients undergoing vascular surgery, myocardial ischemia in the first 30 days was more common in the POAF group (53% vs 21%; P = .01) and new-onset POAF was associated with 30-day perioperative cardiovascular events (HR = 6.0; P < .001) and a higher risk of cardiovascular events up to 1 year after surgery (HR = 4.2; P < .001).[56] In a 2019 study of patients undergoing noncardiac surgery for malignancy, 30-day complications were higher in the POAF group (OR = 2.84; P < .001).[57] Although the most common complications were related to infection and miscellaneous events, outcomes such as AMI, congestive heart failure, and acute kidney injury were also more common in patients with POAF.

The long-term risk of complications is also high in patients who develop POAF. A 2014 study assessing the long-term risk of stroke in patients after cardiac and noncardiac surgery who developed de novo POAF found that the cumulative risk of stroke at 1 year after discharge was 1.47% compared to 0.36% in those with no AF (HR = 2.0; 95% confidence interval [CI], 1.7–2.3).[17] Despite a higher incidence of POAF in the cardiac surgery group, the rate of postdischarge encounters for AF and stroke within 1 year were higher in patients who had noncardiac surgery, underscoring the relevance of POAF to long-term outcomes. In addition to stroke, older data suggest that patients with POAF have a higher risk of developing other complications such as congestive heart failure, myocardial infarction, cardiac arrest, bacterial pneumonia, and increased hospital.[10] However, these data are almost 20 years old and are unlikely to be accurate in current clinical practice. More research is needed to better characterize the long-term risk of POAF on postoperative complications and resource utilization in today's environment.

Overall, current literature suggests that POAF in noncardiac surgery is associated with increased mortality, increased hospital length of stay, and increased cost of hospitalization. Even though data on specific outcomes are sparse, it is clear that POAF impacts surgical outcomes and that attention to POAF prevention is an important element of perioperative management of high-risk patients.


Because POAF is difficult to predict, any of the clinical risk factors for POAF described in Table 1 is a potential target for intervention. Where possible, addressing patient-related factors and averting perioperative triggers of sympathetic stimulation may not only reduce the likelihood of developing de novo AF but also avoid precipitation of RVR in patients with preexisting paroxysmal and chronic AF. Perioperative considerations are summarized in Figure 2.

Figure 2.

Perioperative considerations for patients with atrial fibrillation. *β-blockers, Ca2+ channel blockers, amiodarone, digoxin. $Ketamine, adrenergic vasopressors, desflurane, glycopyrrolate, atropine. AF RVR indicates atrial fibrillation with rapid ventricular rate; AV, atrioventricular; CPAP, continuous positive airway pressure; NOAF, new-onset atrial fibrillation; OSA, obstructive sleep apnea; POCUS, point-of-care ultrasound; TEE, transesophageal echocardiography.

Overall, limited data support the routine use of pharmacologic agents for POAF prophylaxis in noncardiac surgery. A 2018 meta-analysis reviewed 21 randomized controlled trials involving medications used for preventing POAF after noncardiac surgery, and concluded that use of amiodarone, β-blockers, and statins was associated with a decrease in incidence of POAF when compared to controls.[58] Notably, 19 of the 21 trials included in this analysis involved thoracic surgical patients and thus may not be applicable to patients undergoing noncardiothoracic surgery. In light of potentially undesirable side effects of agents such as amiodarone and β-blockers, at this time, use of pharmacologic prophylaxis should remain case dependent and based on the individual risk to benefit ratio.

Perioperative Interventions

Although little literature exists for specific evidence-based perioperative interventions, general considerations include avoiding sympathetic stimulation, avoiding hyper- or hypovolemia, prompt electrolyte replacement, and avoidance of hypoxia. Intraoperative hypotension may predispose to AF. In a 1998 review of 4181 patients, intraoperative hypotension (>30% decrease in systolic blood pressure or systolic <90 mm Hg) was independently associated with persistent postoperative supraventricular arrhythmias requiring treatment.[10] More recently, a 2013 review of 33,000 patients found that duration of mean arterial blood pressure <55 mm Hg correlated with increased myocardial injury.[59]

Treating hypotension promptly may thus reduce the likelihood of POAF. Although no intraoperative literature informs the choice of vasoactive agent, vasopressin is associated with less AF in the septic surgical and nonsurgical patient when compared with catecholamine-based vasopressors such as norepinephrine.[60,61] Earlier use of vasopressin to treat hypotension may thus be preferred in patients with strong risk factors for developing POAF. Phenylephrine causes reflex bradycardia and has been effective in suppressing focal AF.[62] However, perioperative studies linking phenylephrine and POAF are lacking.

Although ketamine has not been linked to POAF, it increases sympathetic activity even at subanesthetic doses,[63] and its intraoperative use as part of multimodal pain management could potentially predispose patients at risk of developing POAF. It thus may be reasonable to avoid ketamine in elderly patients or those with a history of chronic or paroxysmal AF or AF risk factors. Glycopyrrolate, used in combination with neostigmine for reversal of neuromuscular blockade, is an anticholinergic agent that can cause tachycardia and dysrhythmias.[64] The use of sugammadex to reverse neuromuscular blockade may avoid the anticholinergic effects of glycopyrrolate, and its role as an alternative in patients at risk of developing POAF awaits further investigation.