Résumé
<p><b>OBJECTIVES</b>To investigate the different effects of closed suction drainage and non-drainage for total knee arthroplasty(TKA) and to provide reference information for the choice of clinical treatment.</p><p><b>METHODS</b>Randomized controlled trials (RCTs) of closed suction drainage versus non-drainage for TKA were collected from the Cochrane Library, PubMed, EMBase, Springer, CBM, CNKI, VIP and WANFANG database. Methodological quality of the RCTs was independently assessed using the Consolidated Standards of Reporting Trials (CONSORT) checklist. Data analysis was performed by RevMan Version 5.1.6 based on the methods recommended by the Cochrane Collaboration.</p><p><b>RESULTS</b>Twenty-one RCTs without bias were finally enrolled, and 1920 enrolled knees were identified into drainage group (979 knees) and non-drainage group (941 knees). A lower incidence of soft tissue ecchymosis was demonstrated in the closed suction drainage group (OR = 0.30, 95%CI: 0.24 - 0.49); however, compared with the non-drainage group, more loss of blood (MD = 320.03, 95%CI: 235.31 - 404.76) and more need of homologous blood transfusion (OR = 1.83, 95%CI: 1.26 - 3.29) were found in the closed suction drainage group. In addition, there were no significant differences of postoperative infection (OR = 0.53, 95%CI: 0.22 - 1.32), deep venous thrombosis (OR = 1.00, 95%CI: 0.46 - 2.18), and the joint range of motion (MD = -0.04, 95%CI: -1.11 - 1.02) between the two groups.</p><p><b>CONCLUSION</b>Based on the current evidence, no obvious advantage is demonstrated for closed suction drainage, in comparison with non-drainage for TKA.</p>