What medications may increase bleeding risk for individuals taking NOACs (non-Vitamin K oral anticoagulants)?  A recent article published in JAMA aimed to address this question.  Chang et al published their work on assessing the association between the use of NOACs with or without concurrent medications and risk of major bleeding.

This was a retrospective cohort study using data from the Taiwan National Health Insurance database. It included 91330 patients with nonvalvular atrial fibrillation who received at least 1 NOAC prescription of dabigatran, rivaroxaban, or apixaban from Jan 1 2012 through Dec 31 2016.

Specifically, the study evaluated the exposure of NOAC with or without the following medications which have potential drug interactions:

  • P-glycoprotein competitors (digoxin, verapamil, diltiazem, amiodarone, cyclosporine)
  • CYP3A4 inhibitors (fluconazole, ketoconazole, itraconazole, voriconazole or posaconazole), or
  • Both P-glycoprotein competitors & CYP3A4 inhibitors (atorvastatin, erthyromcyin, clarithyromycin, dronedarone, rifampin and phenytoin)

The primary outcome of measure was major bleeding which was defined as hospitalization or emergency department visit with a primary diagnosis of intrancranial hemorrhage or gastrointestinal, urogenital or other bleeding.  Note that traumatic hemorrhage was excluded in the analysis.




The authors have concluded that among patients taking NOACs for nonvalvular atrial fibrillation, the concurrent use of:

  • Amiodarone
  • Fluconazole
  • Rifampin
  • Phenytoin

was associated with increased risk of major bleeding.

Here are my thoughts:

  • It is not uncommon that we see amiodarone to be prescribed in the setting of nonvalvular atrial fibrillation along with NOACs. Should this practice be reviewed? Given this study suggests the risk of major bleeding is not significant with dronedarone, should this drug be a possible substitute? However, dronedarone is only indicated for non permanent atrial fibrillation. This is in part due to the increased risk of cardiovascular death and other complications when evaluated in the setting of permanent atrial fibrillation in PALLAS.   While the drug interactions between amiodarone and warfarin is unattractive, this combination may be a safer option to consider, in light of available evidence thus far.


  • As the authors have pointed out, there are limitations to this study including the retrospective study design, the exclusion of edoxaban as well as study population limited to Asian population, it does offer insights into potential bleeding risks when NOAC is prescribed along with these medications.


  • Finally, this study does not offer a complete picture as it only evaluates safety outcomes in terms of bleeding risk.  It is also important to consider efficacy in context of drug interactions. For example phenytoin is a CYP3A4 inducer which can reduce exposure to many NOACs. As such, the stroke or clotting risk may be increased.  When evaluating each individual patient’s risk in the context of the concurrent use medications with NOACs, both the clotting and bleeding risks must be considered.

Thank you for reading my post and hope to hear your experience with NOACs use along with these concurrent medications.