Over the past decade, anticoagulant-associated bleeding has emerged as a significant, independent risk factor for adverse cardiovascular events, and standardized protocols for specifying their adverse effects have begun to be developed, with bleeding and ischemic events as composite endpoints in clinical trials, such as “net adverse clinical events”. In addition, risk scores to assess bleeding complications have been constructed and incorporated into treatment guidelines to develop optimal drug selection and intensity of antithrombotic therapy. As bleeding risk assessment and clinical trial design evolve, new oral anticoagulants have emerged that may replace or supplement existing antithrombotic therapy. Clinical trials Atrial fibrillation thromboprophylaxis Atrial fibrillation is the most common sustained arrhythmia and an independent risk factor for stroke and death. To date, there are four large clinical trials comparing new oral anticoagulants with warfarin in patients with atrial fibrillation. A pooled analysis of these four studies showed that the new oral anticoagulant was superior to warfarin in the primary effective endpoint, with no heterogeneity between studies. For the bleeding endpoint, intracranial bleeding was significantly reduced with the newer oral anticoagulants compared with warfarin, while GI bleeding was more common and the pooled risk ratio for bleeding was lower with the newer oral anticoagulants, but there was heterogeneity between studies. In terms of effectiveness, each drug was administered at different doses and had different effects in preventing stroke or systemic embolism. For example, dabigatran 150 mg 2/day significantly reduced the risk of stroke or systemic embolism compared with warfarin, whereas dabigatran 110 mg 2/day was not significantly different from warfarin. As for safety, all new oral anticoagulants significantly reduced fatal or intracranial hemorrhage, but not compound hemorrhage or hemorrhage of lesser severity. Pooled analyses have provided clinically relevant information on the combined effects of the new oral anticoagulants compared with warfarin, but studies directly comparing the advantages and disadvantages of the new oral anticoagulants are lacking. Given that large randomized comparative trials are unlikely to be possible given the expense and time involved, indirect comparisons or the use of new analytical techniques may be useful for comparative inferences between different novel oral anticoagulants. One of the indirect comparison studies showed that each new oral anticoagulant was similarly effective in preventing stroke or systemic embolism, but apixaban had the lowest rate of severe bleeding, whereas dabigatran or rivaroxaban had a higher and similar rate of bleeding.