Oral Direct Factor Xa Inhibitors Versus Low-Molecular-Weight Heparin to Prevent Venous Thromboembolism in Patients Undergoing Total Hip or Knee Replacement A Systematic Review and Meta-analysis

Background: Thromboembolic disease is the most frequent medical complication of arthroplasty. Purpose: To evaluate the benefits and harms of oral direct factor Xa inhibitors versus low-molecular-weight heparin (LMWH) in patients undergoing total hip or knee replacement. Data Sources: MEDLINE (1966 to December 2011), EMBASE (1980 to December 2011), and the Cochrane Central Register of Controlled Trials (up to December 2011), without language restrictions. References of reviews and abstracts of conferences were hand-searched. Study Selection: Randomized trials in patients undergoing hip or knee replacement that evaluated factor Xa inhibitors versus LMWH. Data Extraction: Two reviewers independently evaluated eligibility, abstracted the data, and assessed risk for bias. Data Synthesis: In 22 trials, high-quality evidence indicated that the absolute effect of factor Xa inhibitors and LMWH does not differ in terms of all-cause mortality (risk difference, 0 fewer deaths per 1000 patients [95% CI, 2 fewer to 1 more death]) or nonfatal pulmonary embolism (risk difference, 0 fewer events per 1000 patients [CI, 1 fewer to 2 more events]). Factor Xa inhibitors can prevent 4 instances of symptomatic deep venous thrombosis per 1000 treated patients (CI, 3 to 6 fewer events; high-quality evidence) but may increase major bleeding by 2 more events per 1000 patients (CI, 0 to 4 more events; moderate-quality evidence). High, but not lower, doses of oral factor Xa inhibitors increased bleeding compared with LMWH. Limitations: Most trials did not report outcome data for a substantial proportion of the patients. In 9 trials, the follow-up period was 14 days or less. Conclusion: Compared with LMWH, lower doses of oral factor Xa inhibitors can achieve a small absolute risk reduction in symptomatic deep venous thrombosis without increasing bleeding.