Prevention of Venous Thromboembolism in Major Orthopedic Surgery: Bayesian Network Meta-Analysis of 21 Randomized Trials Evaluating Unfractionated Heparins, LowMolecular Weight Heparins, and New Oral Anticoagulants

Background: In major orthopedic surgery, prevention of venous thromboembolism has been based on Unfractionated Heparins (UFHs) over the past decades, then on LowMolecular Weight Heparins (LMWHs), and on New Oral Anticoagulants (NOACs) more recently. To summarize the comparative effectiveness of UFHs, LMWHs, and NOACs in this clinical indication, we applied Bayesian network meta-analysis to the clinical material (randomized studies) published in two previous reviews focused on this issue. Objectives: Our end-point was a composite of venous thromboembolism and pulmonary embolism. Materials and Methods: Our analysis was based on standard Bayesian network metaanalysis (random-effect model). Results: The analysis included 21 randomized trials for a total of 21,805 patients. Our results showed that the degree of effectiveness did not differ among UFHs, LMWHs, and NOACs. Although some trends emerged from an in-depth analysis of these data (e.g. according to the histogram of rankings), no significant differences were found (P > 0.05). Moreover, two agents among LMWHs proved to be adequately supported by randomized trials (enoxaparin and dalteparin), while limited evidence was available for other agents of this class. Conclusions: Our synthesis of the effectiveness data can be useful as an overall reference in this area and can also contribute to defining the place of further innovative treatments for this clinical indication.