Objectives In the ‘Comparison of an Oral Factor Xa Inhibitor With Low Molecular Weight Heparin in Patients With Cancer With Venous Thromboembolism’ (SELECT-D) trial, rivaroxaban showed relatively low venous thromboembolism (VTE) recurrence but higher bleeding compared with dalteparin in patients with cancer. We aim to calculate the cost-effectiveness and budget impact of rivaroxaban compared with dalteparin in patients with cancer at risk of recurrent VTE. Setting We built a Markov model to calculate the cost-effectiveness from a societal perspective over a 5-year time horizon for the Dutch healthcare setting. Participants A hypothetical cohort of 1000 cancer patients with VTE entered the model with baseline characteristics based on the SELECT-D trial. Intervention Six months of treatment with rivaroxaban (15 mg two times per day for first 3 weeks followed by 20 mg once daily) was compared with 6 months of treatment with dalteparin (200 IU/kg daily during month 1 followed by 150 IU/kg daily). Primary and secondary outcome measures The primary outcome of the cost-effectiveness analysis was the incremental cost-effectiveness ratio (ICER). The robustness of the model was evaluated in probabilistic and univariate sensitivity analyses. A budget impact analysis was performed to calculate the total annual financial consequences for a societal perspective in the Netherlands. Results In the base case and all scenarios, rivaroxaban were cost-saving while also slightly improving the patient’s health, resulting in economically dominant ICERs. In the probabilistic sensitivity analysis, 77.8% and 98.7% of the simulations showed rivaroxaban to be cost-saving and more effective for a 5-year and 6-month time horizon, respectively. Rivaroxaban can save up to €11 326 763 (CI €5 164 254 to €17 363 231) in approximately 8000 cancer patients with VTE per year compared with dalteparin based on a 1-year time horizon. Conclusions Treatment with rivaroxaban is economically dominant over dalteparin in patients with cancer at risk for recurrent VTE in the Netherlands. The use of rivaroxaban instead of dalteparin can save over €10 million per year, primarily driven by the difference in drug costs.

Objectives For the non-vitamin-K oral anticoagulants, data on bleeding when used for 42 days as thromboprophylaxis after total knee arthroplasty (TKA) are scarce. This pilot study assessed feasibility of a multicentre randomised clinical trial to evaluate major and clinically relevant non-major bleeding during 42-day use of dabigatran, nadroparin and rivaroxaban after TKA. Patients and methods In 70 weeks, between July 2012 and November 2013, 198 TKA patients were screened for eligibility in the Martini Hospital (Groningen, the Netherlands). Patients were randomly assigned to dabigatran (n=45), nadroparin (n=45) or rivaroxaban (n=48). The primary outcome was the combined endpoint of major bleeding and clinically relevant non-major bleeding. Secondary endpoints of this study were the occurrence of clinical venous thromboembolism (VTE) (pulmonary embolism or deep venous thrombosis), compliance, duration of hospital stay, rehospitalisation, adverse events and Knee Injury and Osteoarthritis Outcome Score (KOOS). Results The primary outcome was observed in 33.3% (95% CI 20.0% to 49.0%), 24.4% (95% CI 12.9% to 39.5%) and 27.1% (95% CI 15.3% to 41.8%) of patients who received dabigatran, nadroparin or rivaroxaban, respectively (p=0.67). Major bleeding was found in two patients who received nadroparin (p=0.21). Clinically relevant non-major bleeding was observed in 33.3% (95% CI 20.0% to 49.0%), 22.2% (95% CI 11.2% to 37.1%) and 27.1% (95% CI 15.3% to 41.8%) for dabigatran, nadroparin and rivaroxaban, respectively (p=0.51). Wound haematoma was the most observed bleeding event. VTE was found in one patient who received dabigatran (p=0.65). The presurgery and postsurgery KOOS qQuestionnaires were available for 32 (71%), 35 (77%) and 35 (73%) patients for dabigatran, nadroparin and rivaroxaban, respectively. KOOS was highly variable, and no significant difference between treatment groups in mean improvement was observed. Conclusions A multicentre clinical trial may be feasible. However, investments will be substantial. No differences in major and clinically relevant non-major bleeding events were found between dabigatran, nadroparin and rivaroxaban during 42 days after TKA. KOOS may not be suitable to detect functional loss due to bleeding. Trial registration number NCT01431456.