ObjectiveTo compare the effect on blood loss after total knee arthroplasty (TKA) between rivaroxaban and enoxaparin. MethodsA retrospective analysis was made on the clinical data of 107 patients (121 knees) with osteoarthritis undergoing primary TKA between January 2010 and October 2012. According to different perioperative anticoagulants, the patients were divided into the rivaroxaban group (51 cases, 57 knees) and the enoxaparin group (56 cases, 64 knees). There was no significant difference in gender, age, height, weight, body mass index, osteoarthritis classification, and disease duration between 2 groups (P>0.05). The total blood loss, hidden blood loss, dominant blood loss, and the percentage of hidden blood loss were compared between 2 groups. The bleeding events were recorded within 35 days after operation. ResultsThe dominant blood loss of enoxaparin group was significantly higher than that of rivaroxaban group (t=3.025, P=0.003), but the percentage of hidden blood loss of enoxaparin group was significantly lower than that of rivaroxaban group (t=4.361, P=0.000); no significant difference was found in the total blood loss and hidden blood loss between 2 groups (P>0.05). The incidence of bleeding event in rivaroxaban group (15.69%; including 1 case of incision bleeding, 4 cases of melena, and 3 cases of haematuria) was significantly higher than that in enoxaparin group (3.57%; including 1 case of haematuria and 1 case of melena) (χ2=4.624, P=0.032). ConclusionRivaroxaban does not increase the risk of hidden blood loss for TKA when compared with enoxaparin, but enoxaparin can increase the risk of dominant blood.