PB0580 - CEDAR STUDY: Multi-centred Retrospective Cohort Study on Appropriate Dosing in Apixaban and Rivaroxaban

Background: Apixaban and rivaroxaban are two commonly used factor Xa inhibitors and dose deviation from labeled recommendations have been reported in real-world studies. However, clinical implications of dose deviation remain ambiguous.

Aims: To compare efficacy and safety outcomes between labeled and nonstandard doses of apixaban and rivaroxaban in patients diagnosed with venous thromboembolism and/or atrial fibrillation in Singapore.

Methods: This is a retrospective, multi-centered cohort study conducted in four public hospitals between January 2015 to December 2020. Primary efficacy outcome includes ischemic stroke and systemic embolism [SSE]; while safety outcome is defined as major bleeding. Follow-up is a year from index diagnosis or until the occurrence of any primary outcome. To adjust for confounders, inverse propensity treatment weighting (IPTW) adjusted logistic regression was carried out.

Results: There were 3812 patients prescribed with apixaban (n = 1839) and rivaroxaban (n = 1973). Median age was 75 years (interquartile range (IQR) 67 – 82) for apixaban and 71 years (IQR 61 – 79) for rivaroxaban. There were higher rates of nonstandard underdosing for both apixaban (underdosed: 14.2%; overdosed: 1.3%) and rivaroxaban (underdosed: 13.1%; overdosed: 6.5%). Patients with nonstandard dosing were more likely to have a history of congestive heart failure and on concurrent antiplatelets. There were no differences in IPTW-adjusted odds ratio (OR) for primary outcomes between labeled vs. nonstandard dosing of rivaroxaban (figure 1). Occurrence of SSE was similar between the two groups for apixaban but there was a significantly higher risk of major bleeding in nonstandard compared to labeled dosing (adj-OR 2.19 95% CI 1.24 to 3.13) (figure 2).

Conclusion(s): Underdosing occurred more commonly among nonlabeled dosing. Patients prescribed with nonstandard dosing of apixaban and rivaroxaban yielded similar clinical outcomes as labeled doses, except for a paradoxical increase in major bleeding for those on nonstandard apixaban dosing; thus posed no benefit in bleeding risk reduction.