In recent years, many changes have taken place regarding agents used for chemical thromboprophylaxis in elective joint replacement. Enoxaparin, Rivaroxaban, Dabigatran and Apixaban are all now recommended in NICE CG92 and their use varies nationally. Whilst data exists comparing oral anticoagulants to Enoxaparin, there is little data on the comparative efficacy of the individual oral anticoagulants.

This study analyses data from Warrington Hospital, where each of the above oral anticoagulants was used trustwide in 3 successive years following hip and knee arthroplasty. We analysed similar 4–5 month periods in 2010(Rivaroxiban), 2011(Dabigatran) and 2012(Apixaban). The study was done prospectively and data was collected contemporaneously. The total sample size was 475 patients. Data was collected through electronic hospital patient records. Patients were excluded if data was incomplete. We defined our primary outcome as any complication requiring the drug to be omitted or stopped.

We found that for Rivaroxaban, 7 of 129 patients had the drug omitted or stopped (5.4%, 95% confidence interval 1.0–9.8), for Dabigatran 19 of 150 patients, (12.7%, 95% confidence interval 6.4–19.0) and for Apixaban 10 of 196 patients (5.1%, 95% confidence interval 0.9–9.3). For Rivaroxaban and Apixaban, there were no confirmed thromboembolic events; however, for Dabigatran, there were six VTEs. All three had bleeding complications, which were well below the figures published for Enoxaparin. Apixaban registered the lowest rate in our study (5.1%). This data suggests that Apixaban is a safe oral anticoagulant in elective total knee and hip replacement.

In recent years, many changes have taken place regarding agents used for chemical thromboprophylaxis in elective joint replacement. Enoxaparin, Rivaroxaban, Dabigatran and Apixaban are all now recommended in NICE CG92 and their use varies nationally. While data exist comparing oral anticoagulants to Enoxaparin, there is little data on the comparative efficacy of the individual oral anticoagulants.

This study analyses data from Warrington Hospital, where each of the above oral anticoagulants was used trustwide in 3 successive years following hip and knee arthroplasty. We analysed similar 4–5 month periods in 2010 (Rivaroxiban), 2011 (Dabigatran) and 2012 (Apixaban). The study was done prospectively and data was collected contemporaneously. The total sample size was 475 patients. Data was collected through electronic hospital patient records. Patients were excluded if data was incomplete. We defined our primary outcome as any complication requiring the drug to be omitted or stopped.

We found that for Rivaroxaban, 7 of 129 patients had the drug omitted or stopped (5.4%, 95% confidence interval 1.0–9.8), for Dabigatran 19 of 150 patients, (12.7%, 95% confidence interval 6.4–19.0) and for Apixaban 10 of 196 patients (5.1%, 95% confidence interval 0.9–9.3). For Rivaroxaban and Apixaban, there were no confirmed thromboembolic events; however, for Dabigatran, there were six VTEs. All three had bleeding complications, which were well below the figures published for Enoxaparin. Apixaban registered the lowest rate in our study (5.1%). This data suggests that Apixaban is a safe oral anticoagulant in elective total knee and hip replacement.