Importance: The optimal pharmacologic thromboprophylaxis agent after total hip and total knee arthroplasty is uncertain and consensus is lacking. Quantifying the risk of postoperative venous thromboembolism (VTE) and bleeding and evaluating comparative effectiveness and safety of the thromboprophylaxis strategies can inform care.
Objective: To quantify risk factors for postoperative VTE and bleeding and compare patient outcomes among pharmacological thromboprophylaxis agents used after total hip and knee arthroplasty.
Design, setting, and participants: This retrospective cohort study used data from a large health care claims database. Participants included patients in the United States with hip or knee arthroplasty and continuous insurance enrollment 3 months prior to and following their surgical procedure. Patients were excluded if they received anticoagulation before surgery, received no postsurgical pharmacological thromboprophylaxis, or had multiple postsurgery thromboprophylactic agents. In a propensity-matched analysis, patients receiving a direct oral anticoagulant (DOAC) were matched with those receiving aspirin.
Exposures: Aspirin, apixaban, rivaroxaban, enoxaparin, or warfarin.
Main outcomes and measures: The primary outcome was 30-day cumulative incidence of postdischarge VTE. Other outcomes included postdischarge bleeding.
Results: Among 29 264 patients included in the final cohort, 17 040 (58.2%) were female, 27 897 (95.2%) had inpatient admissions with median (IQR) length of stay of 2 (1-2) days, 10 948 (37.4%) underwent total hip arthroplasty, 18 316 (62.6%) underwent total knee arthroplasty; and median (IQR) age was 59 (55-63) years. At 30 days, cumulative incidence of VTE was 1.19% (95% CI, 1.06%-1.32%) and cumulative incidence of bleeding was 3.43% (95% CI, 3.22%-3.64%). In the multivariate analysis, leading risk factors associated with increased VTE risk included prior VTE history (odds ratio [OR], 5.94 [95% CI, 4.29-8.24]), a hereditary hypercoagulable state (OR, 2.64 [95% CI, 1.32-5.28]), knee arthroplasty (OR, 1.65 [95% CI, 1.29-2.10]), and male sex (OR, 1.34 [95% CI, 1.08-1.67]). In a propensity-matched cohort of 7844 DOAC-aspirin pairs, there was no significant difference in the risk of VTE in the first 30 days after the surgical procedure (OR, 1.14 [95% CI, 0.82-1.59]), but postoperative bleeding was more frequent in patients receiving DOACs (OR, 1.36 [95% CI, 1.13-1.62]).
Conclusions and relevance: In this cohort study of patients who underwent total hip or total knee arthroplasty, underlying patient risk factors, but not choice of aspirin or DOAC, were associated with postsurgical VTE. Postoperative bleeding rates were lower in patients prescribed aspirin. These results suggest that thromboprophylaxis strategies should be patient-centric and tailored to individual risk of thrombosis and bleeding.