Background: The recommended doses for a NOAC to prevent stroke and systemic embolism (SE) in patients with non-valvular atrial fibrillation (AF) are specified in the prescribing information. Prospective observational studies provide an opportunity to evaluate whether the actual dosing of these drugs in real-life conforms with the approved dosing.

Methods: This analysis assesses the impact of NOAC dosing (i.e., recommended vs. non-recommended) on all-cause mortality at 1 year in patients in newly diagnosed AF. Of patients enrolled consecutively into the Global Anticoagulant Registry in the FIELD-AF (GARFIELD-AF) between Apr-2013 and Sep-2015, 10,417 from 35 countries were eligible for this analysis: rivaroxaban (n=4490, 43.1%), apixaban (n=3283, 31.5%), dabigatran (n=2359, 22.6%) and edoxaban (n=285, 2.7%). Recommended dosing was determined according to differing country criteria, including: impaired kidney function and/or low body weight (≤60 kg) or age ≥80 yrs.

Results: Compared with patients who received the recommended dose, non-recommended dosing according to country-specific guidelines was associated with a higher all-cause mortality. The hazard ratios were: 1.51, 95% confidence interval 1.16 to 1.96 (for dosing below recommendations) and 1.57, 95% CI 0.97 to 2.56 (for dosing above recommendations) after adjusting for age, sex, ethnicity, smoking, alcohol consumption, diabetes, hypertension, history of bleeding, history of stroke/TIA, heart failure, vascular disease, and AF type. Prescribing patterns according to country-specific guidelines are reported in the figure. More than 70% of patients received the correct dose of rivaroxaban, apixaban or dabigatran. Treatment below the recommended doses with rivaroxaban or apixaban was more common in patients from Japan (38.0% [rivaroxaban], 39.4% [apixaban]) than other regions (17.8% [rivaroxaban], 24.0% [apixaban]). Patients on edoxaban were more likely to receive doses below the recommendations compared with the other NOACs; approximately two-thirds of these patients were from Japan. Few patients (3.6%, overall) were treated above the recommended doses. Of those who were treated over the recommended doses, 67.7% had moderate-to severe CKD. By comparison, 8.9% of patients on recommended doses and 7.1% on non-recommended low-doses had moderate-to severe CKD.

Conclusion: Since the introduction of NOACs, most patients receive the recommended NOAC doses according to country-specific guidelines. Treatment above the recommended doses is relatively rare compared with non-recommended low dosing. Prescription of non-recommended doses is associated with an increased the risk of death compared with patients on recommended doses, even after adjusting for baseline factors.