Postoperative atrial fibrillation (POAF) is a common complication, arising in approximately one in three cardiac surgery patients, typically between postoperative days 2 and 4. With pharmacological treatment, it generally resolves within 24 hours of onset, with 90% of patients discharged in sinus rhythm. In non-surgical settings, atrial fibrillation (AF) is managed by either rhythm or rate control and initiation of oral anticoagulation (OAC) to prevent thromboembolic events. Consequently, anticoagulation in an immediate postoperative context may pose an increased bleeding risk. 

COMMENTARY: 

This article aims to examine recommendations from various societies regarding POAF anticoagulation management. In 2014, the American College of Cardiology, the American Heart Association, and the Heart Rhythm Society advised managing POAF similarly to non-surgical AF cases (Class IIA recommendation, level of evidence B), based on findings from the PREVENT-IV trial, which assessed saphenous vein graft patency. In this trial, 25% of patients developed POAF, and anticoagulation was prescribed at discharge. However, without a control group, the significance of this finding remains uncertain. Throughout this article, we highlight the scarcity of trials focusing on POAF, resulting in a generally low evidence level for current recommendations. This is apparent in the 2019 guideline update by the aforementioned societies, which maintained the previous recommendations, albeit emphasizing “shared decision-making and individualized treatment” and adjusting the recommendation to Class IC. 

This lack of consensus persists globally. The Society of Cardiovascular Anesthesiologists and the European Association of Cardiothoracic Anaesthetists issued a Class IIA recommendation with level of evidence B/C in 2019. In 2016, the European Society of Cardiology, the European Association for Cardio-Thoracic Surgery, and the European Heart Rhythm Association recommended OAC for POAF (Class IIA-B), later downgrading to Class IIB-B in 2020. Canadian guidelines also issued a weak recommendation, advising against OAC initiation within the first 72 hours postoperatively due to bleeding risks. 

The primary rationale for anticoagulating AF is to reduce thromboembolic events. A meta-analysis of 55 studies, including 540,209 patients, found a 28.1% incidence of POAF, translating to one stroke for every 50 POAF patients. In coronary surgery patients, POAF incidence was 25.2%, resulting in a stroke in every 69 patients, while in valvular patients, POAF incidence increased to 49%, equating to a stroke in every 36 cases. Two conclusions are evident: first, one in 50 patients who undergo surgery and experience POAF will suffer a stroke; second, stroke risk doubles for valvular patients who develop POAF. However, long-term stroke risk post-POAF remains underexplored, leaving it unclear whether AF is a consequence of underlying morbidity or potentially surgery-related. 

The benefits of OAC are also unclear, as demonstrated by three meta-analyses with conflicting findings. Two of these analyses concluded that OAC reduces thromboembolic events, with one reporting a reduction of 2 events per 1,000 patient-years and another showing a 0.6% decrease. However, both studies noted an increase in bleeding events of approximately 42 events per 1,000 patient-years. None of the meta-analyses demonstrated a significant reduction in mortality. The limitations of these analyses are rooted in the varied study quality, clinical management approaches, and inconsistencies in OAC timing and discontinuation, which hinder result interpretation. 

Registries further underscore the lack of consensus in POAF management. In the Swedish SWEDHEART registry, nearly 25,000 patients were followed longitudinally over eight years. POAF incidence reached 30%, leading to increased rates of stroke, thromboembolism, hospitalization, and heart failure (adjusted HR = 4.16), with no observed increase in all-cause mortality. OAC was associated with elevated bleeding risk (adjusted HR = 1.4). A Danish registry involving 10,500 patients observed only 8.2% of POAF cases receiving OAC, with lower thromboembolism rates than in nonvalvular AF, suggesting potential differences in pathology. 

The Society of Thoracic Surgeons (STS) registry showed a POAF incidence of 25.7% among 167,000 patients. OAC rates varied between 17% and 30%, depending on CHA₂DS₂-VASc score, with no difference in 30-day stroke readmission rates between anticoagulated and non-anticoagulated groups. However, the OAC-treated group had higher mortality at 30 days (HR=1.2) and an elevated rate of bleeding-related readmissions (HR=4.3). Notably, 74% of non-anticoagulated patients were discharged with amiodarone alone, a practice not specified in guidelines, yet this group did not demonstrate increased 30-day mortality, stroke readmission, or bleeding. 

A further STS registry analysis with propensity matching in 39,000 coronary patients found no difference in thromboembolism or stroke incidence between anticoagulated and non-anticoagulated patients. However, POAF patients discharged on OAC had higher short- and long-term mortality (HR = 1.16) and increased bleeding readmissions (HR = 1.6). These findings raise doubts regarding OAC benefits for coronary patients with POAF. 

Finally, the study highlights that patients remain within therapeutic range only 64% of the time during OAC treatment, with lower rates in the initial 3–6 months. High-bleeding-risk patients (e.g., those with renal dysfunction, heart failure, or prior stroke) exhibit the lowest time in therapeutic range. An alternative is direct oral anticoagulants (DOACs), which do not require monitoring. A meta-analysis of five randomized trials and seven observational studies showed that DOACs reduced stroke risk by 37% (NNT = 204) and bleeding by 26% (NNT = 143) compared to warfarin, with no significant difference in mortality, suggesting DOACs may be preferable to vitamin K antagonists (VKAs). 

POAF may represent a distinct entity from general AF, lacking consensus on its management due to insufficient evidence. Few studies consider POAF’s duration, frequency, and occurrence at discharge. Additionally, the CHA₂DS₂-VASc score is not tailored to the postoperative setting and does not consider reoperations, cardiopulmonary bypass duration, or concurrent antiplatelet therapy. Optimal anticoagulation duration also remains unknown. 

In conclusion, before issuing management recommendations for POAF, a comprehensive understanding of this complication is essential. Ongoing trials, such as the Anticoagulation for New-Onset Post-Operative Atrial Fibrillation After CABG (PACES), may provide guidance on optimal management strategies, potentially leading to further publications on this frequent complication. 

REFERENCE: 

Chatterjee S, Ad N, Badhwar V, Gillinov AM, Alexander JH, Moon MR. Anticoagulation for atrial fibrillation after cardiac surgery: Do guidelines reflect the evidence? J Thorac Cardiovasc Surg. 2024 Feb;167(2):694-700. doi: 10.1016/j.jtcvs.2023.03.026.