Vitamin K antagonists versus direct oral anticoagulants after cardiac surgery: a 31-country cohort study
Session title: Oral Anticoagulation for Stroke Prevention in Atrial Fibrillation in ‘Special’ Situations
Topic: Oral Anticoagulation
Session type: Rapid Fire Abstracts
Available from:

View abstract

R Whitlock1 , EP Belley-Cote1 , J Healey1 , PJ Devereaux1 , J Eikelboom1 , A Lamy1 , K Brady1 , B Marsella2 , GI Tagarakis3 , D Paparella4 , W Reents5 , P Punjabi6 , S Connolly1 , 1Population Health Research Institute - Hamilton - Canada , 2McMaster University - Hamilton - Canada , 3Aristotle University of Thessaloniki, Cardiothoracic Surgery - Thessaloniki - Greece , 4Santa Maria Hospital, Emergency and Organ Transplant - Bari - Italy , 5Heart Center Bad Neustadt, Cardiac Surgery - Bad Neustadt a. d. Saale - Germany , 6Imperial College London, Cardiothoracic Surgery - London - United Kingdom of Great Britain & Northern Ireland ,


On behalf: LAAOS III investigators

Citation: N/A

Background: About 10% of patients undergoing cardiac surgery have a history of atrial fibrillation (AF). Among these patients, uncertainty exists regarding the safest and most effective oral anticoagulant (OAC) during the postoperative period.

Purpose: To evaluate practice patterns regarding OAC early after cardiac surgery in patients with a preoperative history of AF and to compare the safety and effectiveness of direct oral anticoagulants (DOACs) versus vitamin K antagonists (VKAs).

Methods: We conducted a nested cohort study within the Left Atrial Appendage Occlusion Study (LAAOS) III (NCT01561651). LAAOS III included patients with AF undergoing cardiac surgery with a CHA2DS2-VASC ≥ 2. In this substudy, we examined patients without end-stage renal dysfunction (eGFR >30 mL/min/1.73m2) who were discharged on OAC. We evaluated bleeding and thromboembolism within 90 days postoperatively using logistic regression adjusting for CHA2DS2-VASC score. 

Results: Recruitment started in 2012 and completed in 2018 in 113 centres in 31 countries. Of the 4811 patients enrolled in LAAOS III, 3725 (77%) were included in this substudy. Preoperatively, 58% of patients received OAC: 56% DOACs and 44% VKAs. At hospital discharge 23% received DOACs and 77% VKAs; 55% of patients on a DOAC at baseline were switched to a VKA while 5% of patients on a VKA were switched to a DOAC. Patients discharged on a DOAC were older, had a higher CHA2DS2-VASC, and were more likely to be male. Patients having undergone an isolated coronary bypass procedure were more likely prescribed DOACs than VKAs (41% vs 23%, p<0.001) whereas isolated non-mechanical valve patients were more likely to be prescribed VKAs (43% vs 28%, p<0.001).  Switching from a DOAC to a VKA postoperatively occurred in 42% of patients In Australia/New Zealand, 49% in Europe, and 63% in North America. Major bleeding between 48 hours postoperatively and 30 days occurred in 1.5% in the DOAC group and 1.3% in the VKA group (aOR 1.14, 95%CI 0.60-2.15, p=0.69) while between 48 hours and 90 days, it occurred in 1.8% of patients in both groups (aOR 0.97, 95% CI, 0.54-1.17, p=0.91). Cardiac tamponade, the composite of stroke and systemic arterial embolism, and the composite of stroke, systemic arterial embolism and death did not differ significantly at 30 and 90 days between the DOAC and VKA groups.

Conclusions: VKAs was the dominant OAC used early after cardiac surgery, but postoperative OAC practices varied by region. After adjustment for CHA2DS2-VASC score, the early postoperative incidence of major bleeding and of the composite of stroke and systemic arterial embolism did not differ significantly when DOACs were compared with VKAs.