Warfarin is associated with higher rates of epistaxis compared to direct oral anticoagulants: A nationwide propensity score-weighted study

Background Although epistaxis is one of the most common side effects of oral anticoagulation, it is unclear whether epistaxis rates vary between different oral anticoagulants (OAC). Objective To compare rates of clinically relevant epistaxis between OAC. Methods Epistaxis event rates were compared between new users of apixaban, dabigatran, rivaroxaban, and warfarin in a nationwide population-based cohort study over a 5-year study period, 2014-2019. Data was collected from the Icelandic Medicine Registry and the five major hospitals in Iceland. Inverse probability weighting (IPW) was used to yield balanced baseline characteristics, and epistaxis rates were compared using Kaplan-Meier survival estimates and Cox regression. Results During the study period, 2098 patients received apixaban, 474 dabigatran, 3106 rivaroxaban, and 1403 warfarin. In total, 93 patients presented with clinically relevant epistaxis, including 11 (12%) major epistaxis events and one fatal epistaxis episode. Furthermore, seven patients (9%) with non-major epistaxis later presented with major bleeding during the follow-up period. Warfarin use was associated with higher rates of epistaxis compared to apixaban (2.2 events per 100-person years (events/100-py) vs. 0.6 events/100-py, hazard ratio [HR] 4.22, 95% confidence interval [CI] 2.08-8.59, p < 0.001), rivaroxaban (2.2 events/100-py vs. 1.0 events/100-py, HR 2.26, 95% CI 1.28-4.01, p = 0.005), and dabigatran (2.2 events/100-py vs. no events, HR n/a, p < 0.001). Conclusion Warfarin treatment was associated with higher rates of clinically relevant epistaxis compared to direct oral anticoagulants.

Automated summary

This study aims to compare the rates of clinically relevant epistaxis between different oral anticoagulants. The results show that warfarin treatment is associated with higher rates of clinically relevant epistaxis compared to direct oral anticoagulants.