Background: Patients scheduled for primary hip or knee replacement are frequently treated with long-term curative oral anticoagulants (LT-COA), vitamin K antagonists (VKA) or direct oral anticoagulants (DOA). We conducted a prospective, controlled study comparing patients treated with LT-COA undergoing total hip replacement (THR) or total knee replacement (TKR) with patients not treated with LT-COA undergoing the same procedure, in order to estimate the adjusted and unadjusted risks of postoperative haemorrhagic or thrombotic complications associated with LT-COA.
Hypothesis: The rate of post-operative complications was higher in the group treated with LT-COA.
Patients and methods: This was a prospective, comparative, non-interventional study comparing patients undergoing primary THR or TKR arthroplasty who were treated with LT-COA with those who were not treated with curative-dose oral anticoagulants (control group). A total of 325 patients were assessed at 3 months, 215 (66%) in the control group, 123 with THR and 92 with TKR, and 110 (34%) in the LT-COA group, 68 with THR and 42 with TKR. Patients in the control group were more likely to be female, and were also slightly younger and in better pre-operative health. In the LT-COA group, 55 (50%) patients were treated with VKA and 55 (50%) with DOA. The surgical technique, implants and management of perioperative anticoagulants were left to the discretion of the practitioners. The primary endpoint was the occurrence of a HTR (haemorrhagic or thrombotic complications, or revision surgery) complication within 3 months of surgery.
Results: Of the 325 patients analysed, 77 (24%) had a HTR complication, 50 (23%) in the control group and 27 (25%) in the LT-COA group (OR = 0.93; 95% CI: 0.54-1.59 (p = 0.80)). In multivariable analysis, the risk of a HTR complication was not significantly different between the groups (adjusted OR = 0.85; 95% CI: 0.45-1.61 (p = 0.62)). Age (p = 0.99) and sex (p = 0.97) were not associated with the occurrence of a HTR complication. In contrast, ASA score (p = 0.017), and type of surgery (THR or TKR, p = 0.019) were significantly associated with the occurrence of a HTR complication in the multivariable analyses. In the LT-COA group, the rate of HTR complications during hospitalisation was significantly higher (p = 0.0076) for patients who had undergone preoperative relay (discontinuation of anticoagulants and initiation of LMWH (Low Molecular Weight Heparin)) (n = 15 out of 41 patients, 37%) compared with those who had not (n = 9 out of 69 patients, 13%). Similarly, HTR complications during hospitalisation were significantly more frequent (p = 0.0363, Fisher's exact test) for patients treated with VKA (n = 19 out of 55 patients, 35%) compared with those treated with DOA (n = 8 out of 55 patients, 15%).
Discussion: The rate of HTR complications was not significantly different between the control and LT-COA groups. The use of LMWH as a relay was associated with a significant increase in the HTR complication rate. The increased risk of haemorrhage associated with the use of LMWH may be due to residual anticoagulation at the time of surgery, earlier postoperative anticoagulation, or dual transitory anticoagulation with VKA and LMWH.
Level of evidence: III; comparative prospective study.
