This study comes as a great contribution to the current controversy over the common clinical problem of medical management of AF patients needing antiplatelet therapy. This is the largest epidemiological study made on this issue， based on the medical registries of the entire Danish population. It is also a real-life patients study combining data from different centres and including all kind of patients regardless health insurance coverage or other confounders. What this paper tells us is that more drugs could mean less for AF patients， a statement already raised by another Scandinavian study published in 2013， the WOEST trial， an open-label， multicenter， randomised controlled study according to which adding aspirin on top of OAC plus clopidogrel post PCI in AF patients confers no thrombo-prophylaxis and carries an increased bleeding risk. The study by Lamberts et al. compared numerous antithrombotic regimens and found that the combination of OAC plus clopidogrel seems to present the best risk-benefit profile. Last but not least， the selection of antithrombotic treatment by the clinicians was not influenced by the predicted stroke risk as assessed by the CHADS2 score， a validated tool that has been in clinical practice since 2001. Such a seemingly random treatment of AF patients should alert clinicians. Adherence to the guidelines for thrombo-prophylaxis in AF patients remains an issue as demonstrated by Warlé-Van Herwaarden et al. in 2012， and is probably a cause of frequent and severe iatrogenic complications in the AF patient population.

　　Some limitations of the study should be noted. First of all， selection bias could not be excluded as triple antithrombotic therapy could be preferred for “healthier” patients with less medical co-morbidities. Secondly， no novel anticoagulants (NOACs) were studied， as observation terminated at 2009. NOACs are currently approved for non-valvular AF and their clinical use is growing but data is still scarce regarding the risk-benefit ratio when combined with antiplatelets. Thirdly， outcomes are only studied at one year after MI or PCI. It still remains unclear whether triple antithrombotic therapy should be used in the acute phase after MI or PCI (eg 2 weeks to 3 months).  It could be asserted that an initial clinical benefit could be alleviated by an increased bleeding risk afterwards. This study would be much more significant if it could compare outcomes in patients receiving these antithrombotic regimens in 3 as well as in 12 months.

　　Despite the aforementioned limitations， this is an important contribution that helps physicians select an optimal antithrombotic therapy for AF patients post MI or PCI. In the absence of prospective studies that could provide high-quality data， clinicians should seriously revise the widely established clinical practice of prolonged triple antithrombotic therapy in AF patients post MI or PCI. If there is still a role for triple therapy， it should be limited in the acute phase， according to current guidelines. OAC and clopidogrel combination seems to be equally effective to triple antithrombotic therapy. However， as we entering the era of NOACs， OAC-based regimens could soon prove obsolete and research for combined therapies including the novel agents is urgently needed.