OBJECTIVES: Apixaban has demonstrated non-inferiority to warfarin in preventing stroke in non-valvular atrial fibrillation (NVAF). Unlike warfarin, apixaban does not require regular monitoring of coagulation levels, potentially improving treatment persistence. This retrospective study evaluated the risk of discontinuation and health care resource utilisation (HCRU) in patients with NVAF treated with apixaban or warfarin.

METHODS: NVAF patients initiating warfarin or apixaban between December 2012 and March 2016 were identified from linked primary (Clinical Practice Research Datalink) and secondary care (Inpatient Hospital Episode Statistics) data in England. Apixaban patients were propensity-score matched 1:1 to warfarin patients to control for baseline differences. Outcomes of interest were risk of discontinuation (gap in treatment of >60 days) and rate per person-year (ppy) of all-cause hospitalisations, outpatient attendances and primary care visits (GP or practice nurse) during the first continuous treatment episode. Cox regression models compared risk of discontinuation between treatments.

RESULTS: All 1333 apixaban patients identified were matched to warfarin users: mean age (apixaban vs. warfarin: 76.4 vs. 76.5, p=0.867); percentage of females (44.7 vs. 44.6, p>0.999). Kaplan-Meier estimates suggested that at 1 year, 85.2% of apixaban users and 73.3% of warfarin users were persistent on treatment. Apixaban showed lower risk of treatment discontinuation than warfarin (hazard ratio: 0.63, 95% CI 0.53-0.74, p<0.001). The rate of primary care visits (apixaban vs. warfarin: 13.4 vs. 19.0 ppy, p<0.001) and outpatient attendance (5.84 vs. 8.69 ppy, p<0.001) was lower in apixaban compared to warfarin. Hospitalisation rates were similar in the two groups.

CONCLUSIONS: In patients with NVAF, persistence with anticoagulants is important to reduce the risk of stroke. In this study, apixaban was associated with lower risk of discontinuation than warfarin and some reductions in HCRU. These results provide initial evidence of increased persistence in NVAF patients receiving apixaban compared to warfarin in the real-world setting.