• BACKGROUND
    • The use of closed-suction drainage systems after total joint replacement is a common practice. The theoretical advantages for the use of drains is a reduction in the occurrence of wound hematomas and infection. The aim of this meta-analysis was to determine, on the basis of the evidence from randomized controlled trials, the advantages and adverse effects of surgical drains.
  • METHODS
    • All randomized trials, as far as we know, that compared patients managed with closed-suction drainage systems and those managed without a drain following elective hip and knee arthroplasty were considered. The trials were identified with use of searches of the Cochrane Collaboration with no restriction on languages or source. Two authors independently extracted the data, and the methods of all identified trials were assessed.
  • RESULTS
    • Eighteen studies involving 3495 patients with 3689 wounds were included in the analysis. The pooled results indicated that there was no significant difference between the wounds treated with a drain and those treated without a drain with respect to the occurrence of wound infection (relative risk, 0.73; 95% confidence interval, 0.47 to 1.14), wound hematoma (relative risk, 1.73; 95% confidence interval, 0.74 to 4.07), or reoperations for wound complications (relative risk, 0.52; 95% confidence interval, 0.13 to 1.99). A drained wound was associated with a significantly greater need for transfusion (relative risk, 1.43; 95% confidence interval, 1.19 to 1.72). Reinforcement of wound dressings was required more frequently in the group managed without drains. No difference between the groups was seen with respect to limb-swelling, venous thrombosis, or hospital stay.
  • CONCLUSIONS
    • Studies to date have indicated that closed suction drainage increases the transfusion requirements after elective hip and knee arthroplasty and has no major benefits. Further randomized trials with use of larger numbers of patients with full reporting of outcomes are indicated before the absence of any benefit, particularly for the outcome of wound infection, can be proved.