(P-065) Safety of Rivaroxaban in First-time Users Treated for Deep Vein Thrombosis and Pulmonary Embolism

Background: A post-authorisation study programme of rivaroxaban use was conducted between 2011–2020.

Objectives: To study major bleeding events among first-time users of rivaroxaban (RVX) or standard of care (SOC: vitamin K antagonist) for deep vein thrombosis (DVT) and pulmonary embolism (PE) treatment in routine clinical practice.

Methods: This cohort study used data from four European countries: IQVIA Medical Research Data-UK (IMRD-UK) in the UK, the German Pharmacoepidemiological Research Database (GePaRD), the Dutch PHARMO Database Network and Swedish national health registries. Patients ≥2 years (no recent history of cancer) with a first prescription of RVX or SOC from December 2011–December 2017 temporally close to diagnostic codes for DVT/PE were followed until the date of each safety outcome (intracranial [ICB], gastrointestinal [GIB], urogenital [UGB] or other [OB] bleeding), death, or study end (December 2017 or December 2018, country-dependent).
Unadjusted incidence rates (IR; [95% CI]) of bleeding events leading to hospitalization in the first treatment period were calculated for RVX and SOC users; nested case-control analyses estimating adjusted relative risk using the odds ratio (OR; [95% CI]) were performed for each safety outcome per database.

Results: Overall, 44 737 (RVX) and 45 842 (SOC) first-time users were identified. IR estimates for ICB ranged from 0.17 (0.06–0.38) to 0.41 (0.04–1.2) for RVX users and from 0.16 (0.03–0.4) to 0.50 (0.43–0.60) for SOC users. IR of GIB ranged from 0.58 (0.36–0.90) to 1.13 (1.00–1.28) for RVX users and from 0.37 (0.20–0.62) to 0.94 (0.83–1.06) for SOC users.
In the nested case-control analysis, OR for ICB associated with current use of RVX vs non-use ranged from 1.56 (0.98–2.50) to 2.54 (0.74–8.69) and was statistically significant in one database, while OR for SOC use vs non-use ranged from 1.53 (0.42–5.52) to 2.42 (1.62–3.61) and was statistically significant in two databases. For GIB, the ORs for RVX use vs non-use ranged from 2.24 (1.79–2.82) to 4.10 (1.90–8.87), statistically significant in 3 databases; OR for SOC use vs non-use ranged from 2.24 (1.76–2.84) to 6.76 (2.20–20.80) and were statistically significant in 3 databases.

UGB and OB outcomes will also be presented.

Conclusions: Observed incidence rate estimates were broadly in line with those from previous observational studies and clinical trials. Current use of SOC was more often associated with statistically significant increased relative risk of ICB than RVX use. Current use of RVX or SOC were similarly associated with a statistically significant increased relative risk of GIB. Observed variations may reflect data source and healthcare system differences.