Background. The incidence of bleeding following primary TKR has increased with the use of chemical thromboprophylaxis. Our aim was to compare Clexane, Apixaban and Rivaroxaban in terms of frequency and volume of bleeding episodes, need for blood transfusion, return to theatre and incidence of VTE events. Methods. Between February and May 2014, a consecutive series of 132 primary TKRs were studied prospectively. The wound dressings of these patients were assessed daily to look for signs of bleeding and classified into: Mild (< 50p size coin), moderate (> 50p size coin) or Severe (blood seeping through the dressing). Follow up was up to minimum of 30 days post discharge. Results. Apixaban, Rivaroxaban & Clexane were used in 64, 23 and 45 patients respectively. Eleven patients had at least 1 day of mild bleeding, 8 had at least 1 day of moderate bleeding and 11 had at least 1 day of severe bleeding. Ten patients had 1 or more doses omitted because of bleeding. However, there was no statistical significance in distribution of bleeding episodes or doses omitted due to bleeding amongst the three drugs (chi squared test). There was also no correlation between number of severe bleeding episodes and the need for blood transfusion. There were two VTE events recorded; 1 PE each in the Apixaban and Rivaroxaban groups. Two cases in the Apixaban group and 1 case in the Clexane group returned to theatre for washout of haematoma. Conclusion. There was an 8% incidence of severe bleeding in our study group. The incidence of bleeding problems following TKR was similar in the Apixaban, Rivaroxaban & Clexane groups. Level of evidence. III - Evidence from case, correlation, and comparative studies

Since the establishment of our department a multi-modal approach to thromboprophylaxis that uses aspirin for chemical prophylaxis was adopted. In accordance with the latest national recommendations, our routine chemical prophylaxis following arthroplasty was changed to rivaroxaban in 2012 and then dalteparin in 2013. This study aimed to compare venous thromboembolism (VTE) rates during the use of the aspirin-based protocol used from 2004 to 2011 with recent, rivaroxaban and dalteparin-based guidelines. Outcome data from ISD Scotland was retrieved and radiology reports performed for CT pulmonary angiograms and lower limb doppler ultrasound scans in our institution were assessed to identify cases of VTE following primary hip or knee arthroplasty. The incidence of pulmonary embolism (PE) and proximal deep venous thrombosis (DVT) was calculated for each year and compared using a Chi-squared test. Additionally, the change in extended thromboprophylaxis regimen was surveyed by recording the discharge prescriptions for consecutive arthroplasty patients for March every year. There were 90 radiologically confirmed cases of DVT or PE between 2004 and 2011 (incidence of 0.71%). The DVT/PE rate was subsequently 0.67% in 2012 and 0.69% in 2013, with a further 29 cases identified. This does not represent a significant change in the venous thromboembolism rates and remains below the national incidence of VTE (1.06%). Aspirin alone was used as chemical thromboprophylaxis in 80.8% of patients from 2004 to 2011, 50.9% in 2012, and 12.1% in 2013. The incidence of VTE at our centre remains favourable to national figures, but the modification of thromboprophylaxis guidelines will incur additional financial costs and has not had a significant reduction on the rate of VTE