To recommend the proper anticoagulant drug and its dose for patients with atrial fibrillation (AF) and heart failure (HF), we conducted a network meta-analysis (NMA) to make the comparisons among non-vitamin K antagonist oral anticoagulants (NOACs) and warfarin with regard to efficacy (stroke or systemic embolism) and safety (major bleeding).
We searched PubMed, EMBASE, Web of Science and Cochrane Library with the items: “dabigatran, edoxaban, apixaban, rivaroxaban, warfarin, atrial fibrillation and heart failure” through April 14, 2020, focusing on the RCTs comparing the effect of NOACs to warfarin in patients with AF and HF. The NMA was performed based on R (version3.5.1) recalling JAGS (version4.3.0) with gemtc package. Moreover, NetMetaXL (version1.6.1) and winBUGS (version1.4.3) were employed to obtain the cumulative ranking curve area (SUCRA) of the anticoagulants.
There was a high probability that dabigatran150 (SUCRA 0.82) ranked the first for the most effective drug, followed by apixaban (SUCRA 0.81), edoxaban60 (SUCRA 0.57) and rivaroxaban (SUCRA 0.52). However, with respect to safety for preventing major bleeding, edoxaban30 (SUCRA 0.99) ranked as the safest drug, followed by apixaban (SUCRA 0.71), edoxaban 60 (SUCRA 0.59) and dabigatran150 (SUCRA 0.55).
Apixaban, edoxaban60 and dabigatran150 were more likely to become the choice for preventing stroke or systemic embolism and major bleeding in patients with AF and HF. Nevertheless, more trials need to be performed to focus on the effect of NOACs on the efficacy outcome due to the sparse data. In addition, caution should be excised over selecting the NOAC and its dose on account of the lacking head-to-head comparisons.

Copyright © 2020 Elsevier Ltd. All rights reserved.

Author