Patient Care

Should Patients Who Develop Postoperative Atrial Fibrillation Start Anticoagulation?



A 66-year-old man with diabetes mellitus type 2 and hypertension underwent left total knee replacement. Several hours after surgery, the patient developed atrial fibrillation (AF). He was asymptomatic, and reversible causes of AF were ruled out. Approximately 18 hours later, he spontaneously reverted back to sinus rhythm. Should this patient, who has no known prior history of AF and a CHA2DS2-VASc score of 3, be started on anticoagulation?


Image Credit:

Image Credit:

Hospitalists are commonly consulted for evaluation and management of postoperative atrial fibrillation (POAF). The incidence of new-onset AF associated with non-cardiac surgery is approximately 2% and may be more frequent in an elderly population.1 The increased adrenergic tone associated with surgery is thought to elicit AF in some patients. POAF has also been associated with positive fluid balance, electrolyte abnormalities, and hypoxemia.2 Some of these patients will spontaneously revert back to sinus rhythm after these issues are reversed. Others will go on to develop chronic or paroxysmal AF that persists indefinitely. It is also likely that some patients with POAF, in fact, already had asymptomatic AF that was simply undetected prior to hospitalization.

Hospitalists are faced with the difficult task of determining which patients with POAF will benefit from either short-term or long-term anticoagulation. This has not been well studied in postsurgical patients, in contrast to medical patients in whom stroke risk from AF has been very well-characterized. The decision may be further complicated by bleeding risk (associated with either some surgeries or with patient-dependent factors).3

It is worth noting that following major cardiac or thoracic surgery, POAF is common; the incidence ranges from 10% to 60%. In these cases, POAF may be triggered by transient atrial ischemia or by postoperative inflammation and may have a different natural history from POAF in non-cardiac surgery patients in terms of both reversibility and stroke risk. More retrospective data are available regarding cardiothoracic surgery patients.

Previous American Heart Association (AHA) and American College of Cardiology (ACC) guidelines stated that POAF lasting longer than 48 hours warranted anticoagulation. This recommendation was removed from the newest update. The 2014 updated AHA/ACC guidelines are less absolute and now state only that “it is reasonable to administer antithrombotic medication in patients who developed postoperative AF, as recommended for nonsurgical patients” (Level of Evidence: B) in regard to cardiothoracic surgery.4

There is no specific recommendation regarding POAF for non-cardiac surgery patients. The current guidelines are likely purposefully vague due to the lack of direct evidence. The following is a review of the existing literature and a suggested approach to anticoagulation in POAF.


How common is postoperative atrial fibrillation? New-onset AF during hospitalization is known to occur in association with many acute conditions including surgery, infection, and myocardial infarction. About half of the cases of in-hospital new-onset AF are associated with surgery. AF is more commonly seen in surgery that involves the thoracic cavity and cardiac structures. In a cross-sectional epidemiologic study of 22 million patients in California, 20.8% of patients undergoing cardiac surgery developed POAF compared with only 1.3% of patients undergoing non-cardiac surgery.5 A smaller study of non-cardiac surgery patients found a 30-day POAF incidence of 0.37%.2

Does postoperative atrial fibrillation increase the short-term risk of stroke? A major concern in AF is the risk of stroke. It is well-established that prolonged or recurrent AF increases the risk for stroke over months or years, but do short episodes of POAF increase stroke risk to a significant degree? Most of the studies in the literature focus on perioperative stroke risk specifically in cardiothoracic surgery. A prospective study of 4,000 patients undergoing cardiac surgery found that the in-hospital postoperative stroke risk was 3.3% in patients with POAF compared to 1.4% in patients without POAF (P<0.01).6 Similar outcomes were seen in a VA study looking at patients who underwent open heart surgery: Stroke risk was 5.3% at six months in POAF patients compared to 2.4% in those without POAF.7 Another study of coronary artery bypass graft (CABG) patients with a follow-up of almost six years showed a stroke risk of 12.1% in POAF patients compared to 8.4% in those without POAF.8

It is not clear that all of the increase in stroke risk is a direct effect of POAF. Indeed, in a retrospective analysis of almost 3,000 CABG patients, 1.1% suffered a stroke during their hospital stay. Fewer than half of those had a cardiac rhythm other than sinus rhythm. In the 15 stroke patients who developed POAF, nine presented with stroke symptoms prior to the first episode of AF.9 The authors suggest that aggressive anticoagulation for POAF would not have prevented most of these events.

Furthermore, the rate of in-hospital stroke after non-cardiac surgery is probably much lower, though it has not been as well studied. These data raise some questions as to the benefit of anticoagulation in the immediate postoperative period, though it is difficult to draw firm conclusions without randomized data.

What about non-cardiac surgery? There is less evidence available for patients undergoing non-cardiac surgery, but the few studies that do exist also point to higher stroke risk in patients with POAF. A large population-based study using ICD codes found that the one-year risk of stroke for patients with POAF after non-cardiac surgery was 1.47% compared to 0.36% in non-cardiac surgery patients without POAF (P<0.001). Based on these data, the long-term stroke risk after POAF in non-cardiac surgery patients is similar to that of medical AF patients with a CHA2DS2-VASc score of 2. The authors of this study suggest that transient POAF after non-cardiac surgery may carry a long-term stroke risk similar to any other AF diagnosis.10 However, this study design is subject to significant ascertainment bias (i.e., they may have unintentionally captured some patients with preexisting or prolonged AF), and further research is needed to better delineate this risk.

Does increased stroke risk translate into increased mortality? In a retrospective study of 17,000 patients, El-Chami et al found that POAF after CABG was associated with decreased survival after one year (90% versus 96%) and 10 years (55% versus 70%).11 However, those patients who develop POAF may be sicker overall.

Another study showed that death due to stroke occurred in 4.2% of POAF patients compared to 0.2% of non-POAF patients in a five-year period.12 Based on these studies, POAF is likely associated with increased mortality, but there may be other unaccounted variables. Nevertheless, the increased mortality associated with POAF in these populations is similar to that seen for non-surgical population-based studies13 and provides support that those with newly diagnosed AF in the post-surgical setting should at least be followed closely to assess for recurrence.

What is a patient’s risk of developing atrial fibrillation later in life? When we choose to anticoagulate patients with POAF, we then have to determine whether they should be committed to long-term anticoagulation. It is thought that many cases of POAF are transient; however, some patients will go on to have persistent or paroxysmal AF after discharge.

A retrospective study examined 571 patients who underwent CABG, 30% of whom had POAF during the index admission. After five years of follow-up, 25% of those with POAF were diagnosed with paroxysmal or persistent AF after discharge compared to only 3% of patients without POAF. Researchers did this by looking at the most recent ECG, if done in the last year, or by obtaining a new ECG at the five-year point.12 By this method, it is probable that some diagnoses of paroxysmal AF were missed.

In another study of about 300 CABG patients, about 20% of patients with POAF also went on to develop post-discharge AF, defined as symptomatic AF that led to medical evaluation. As in the previous study, it is likely that there were undetected episodes of AF.14 Thus, in cardiothoracic surgery patients, some but not all of whom develop POAF have recurrent or ongoing AF. For this reason, if anticoagulation is started, it may be reasonable to stop anticoagulation after weeks or months if ongoing AF is not apparent.

What is the risk of postoperative bleeding if anticoagulation is started? Any decision about the benefits of anticoagulation must be weighed against the risks, most notably the risk of serious or life-threatening bleeding. This risk may be heightened in the immediate perioperative period. Discussions should always take place with our surgical colleagues about type of surgery, intraoperative complications, and postoperative risk of bleeding.

Anticoagulation, if indicated, should not be started until postoperative bleeding risk is deemed appropriately low. That said, the 2015 BRIDGE trial (looking at the benefits and risks of “bridging” patients before surgery) provides some peripheral but meaningful information about postoperative bleeding risk. In this study, patients with preexisting AF who underwent low-bleeding-risk surgery and were bridged on day one after surgery with therapeutic doses of unfractionated or low-molecular-weight heparin had a significantly higher risk of postoperative bleeding compared to non-bridged patients, with a number needed to harm of 50.15 It may be reasonable—and likely safer—to wait a couple days to start anticoagulation for patients with POAF.

What is the expert’s opinion? We asked one of our cardiac electrophysiologists what her approach is to this situation. In general, if a patient has a low stroke risk and is in AF for fewer than 24 hours, it is reasonable to defer anticoagulation and follow as an outpatient. Regardless of risk, if AF is sustained for more than 24 hours, we recommend at least four weeks of anticoagulation and close outpatient follow-up, which should include a period of ambulatory monitoring to determine the need for continued anticoagulation. We also recommend considering what comprises the patient’s stroke risk.

For example, if the CHA2DS2-VASc score is 2 but the points come from being a female with coronary artery disease, we would consider forgoing anticoagulation but arranging for an outpatient cardiac monitor with cardiology follow-up. If the patient has a history of stroke or TIA, we recommend continuing anticoagulation indefinitely.

Back to the Case

Given our patient’s episode of POAF lasted fewer than 24 hours, it would be reasonable to hold off starting anticoagulation, but he should be followed as an outpatient with ambulatory monitoring at a minimum, monitoring for recurrence. If he were to develop recurrent AF, then he would warrant anticoagulation based on an annual stroke risk of 3.2% as determined by a CHA2DS2-VASc score of 3.

Bottom Line

Our strategy is as follows: If a patient has a low stroke risk (i.e., CHA2DS2-VASc score <2) and is in AF for fewer than 24 hours, anticoagulation is not started, but outpatient follow-up is arranged to monitor symptoms. Regardless of stroke risk, if a patient is in AF for more than 24 hours, we initiate and continue anticoagulation for a minimum of four weeks and arrange outpatient follow-up with a period of ambulatory monitoring to determine need for continued anticoagulation. If a patient has a high stroke risk (CHA2DS2-VASc >2) or if their risk factors include a history of stroke or TIA, anticoagulation is started and continued indefinitely. Risk-benefit discussion is held with the patient, especially with regard to bleeding risk, prior to anticoagulation initiation. If the individual patient’s situation presents further nuance, we ask for the assistance of our cardiology or cardiac electrophysiology colleagues.

Final Thought

None of the mentioned studies investigated or included newer oral anticoagulants. Risk-benefit ratios may change (potentially considerably) with these agents. Further study is needed. We expect, in due time, studies will look at the question of POAF in regard to newer anticoagulant agents, and perhaps then our decision making will change. TH

Dr. Evavold is a resident in the hospitalist training program, while Dr. Lessing and Dr. Merritt are hospitalists in the Department of Internal Medicine at the University of Colorado. Dr. Tzou is a cardiologist in the section of electrophysiology at the University of Colorado.


  1. POISE Study Group, Devereaux PJ, Yang H, et al. Effects of extended-release metoprolol succinate in patients undergoing non-cardiac surgery (POISE trial): a randomised controlled trial. Lancet. 2008;371(9627):1839-1847. doi:10.1016/s0140-6736(08)60601-7.
  2. Christians K, Wu B, Quebbeman E, Brasel K. Postoperative atrial fibrillation in noncardiothoracic surgical patients. Am J Surg. 2001;182(6):713-715. doi:10.1016/s0002-9610(01)00799-1.
  3. Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138(5):1093-1100. doi:10.1378/chest.10-0134.
  4. Fleisher L, Beckman J, Brown K, et al. ACC/AHA 2007 guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 2002 guidelines on perioperative cardiovascular evaluation for noncardiac surgery). Circulation. 2007;116(17):e418-e500. doi:10.1161/circulationaha.107.185699.
  5. Walkey A, Benjamin E, Lubitz S. New-onset atrial fibrillation during hospitalization. J Am Coll Cardiol. 2014;64(22):2432-2433. doi:10.1016/j.jacc.2014.09.034.
  6. Creswell L, Schuessler R, Rosenbloom M, Cox J. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg. 1993;56(3):539-549. doi:10.1016/0003-4975(93)90894-n.
  7. Almassi G, Schowalter T, Nicolosi A, et al. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg. 1997;226(4):501-513.
  8. Horwich P, Buth K, Légaré J. New onset postoperative atrial fibrillation is associated with a long-term risk for stroke and death following cardiac surgery. J Card Surg. 2013;28(1):8-13. doi:10.1111/jocs.12033.
  9. Kollar A, Lick S, Vasquez K, Conti V. Relationship of atrial fibrillation and stroke after coronary artery bypass graft surgery: when is anticoagulation indicated? Ann Thorac Surg. 2006;82(2):515-523. doi:10.1016/j.athoracsur.2006.03.037.
  10. Gialdini G, Nearing K, Bhave P, et al. Perioperative atrial fibrillation and the long-term risk of ischemic stroke. JAMA. 2014;312(6):616. doi:10.1001/jama.2014.9143.
  11. El-Chami M, Kilgo P, Thourani V, et al. New-onset atrial fibrillation predicts long-term mortality after coronary artery bypass graft. J Am Coll Cardiol. 2010;55(13):1370-1376. doi:10.1016/j.jacc.2009.10.058.
  12. Ahlsson A, Fengsrud E, Bodin L, Englund A. Postoperative atrial fibrillation in patients undergoing aortocoronary bypass surgery carries an eightfold risk of future atrial fibrillation and a doubled cardiovascular mortality. Euro J Cardiothorac Surg. 2010;37(6):1353-1359. doi:10.1016/j.ejcts.2009.12.033.
  13. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation. 1998;98(10):946-952.
  14. Antonelli D, Peres D, Freedberg N, Feldman A, Rosenfeld T. Incidence of postdischarge symptomatic paroxysmal atrial fibrillation in patients who underwent coronary artery bypass graft: long-term follow-up. Pacing Clin Electrophysiol. 2004;27(3):365-367. doi:10.1111/j.1540-8159.2004.00443.x.
  15. Douketis J, Spyropoulos A, Kaatz S, et al. Perioperative bridging anticoagulation in patients with atrial fibrillation. N Engl J Med. 2015;373(9):823-833. doi:10.1056/nejmoa1501035.

Key Points

  1. POAF is associated with an increased risk of stroke in the perioperative period.
  2. Some, but not all, patients with POAF will go on to have recurrent or chronic AF.

Additional Reading

Next Article:

   Comments ()