POAF after CABG: early risk, landmark bias, and the case for selective anticoagulation

This is correspondence linked to ‘Real-world anticoagulant use and clinical outcomes in postoperative atrial fibrillation after coronary artery bypass grafting: a nationwide practice-based cohort from the Netherlands Heart Registration’, by M.R.D van de Kar et al. https://doi.org/10.1093/ehjqcco/qcaf124. We have read with great interest the national study by van de Kar and colleagues on the real-world use of oral anticoagulants (OAC) and clinical outcomes in patients with post-operative atrial fibrillation (POAF) after CABG in the Netherlands. Among 44 601 CABG patients, 24.8% developed POAF and 43.6% received OAC at discharge; despite raw lower rates of thromboembolic mortality among OAC users, the 30-day landmark analysis did not show differences in survival at 1 year (peak OAC prescribing at 53.7% in 2022).1 We believe the work offers an important contribution by documenting practice at a national level, but some methodological choices warrant discussion. First, the 30-day landmark, while useful to mitigate mortality bias and dispensing data limitations, might attenuate a potential early benefit of OAC during the period of highest thromboembolic risk after surgery; a time-dependent exposure analysis and competing risks models could reduce misclassification and clarify whether the absence of differences between 30 days and 1 year reflects therapeutic neutrality or measurement limits.