Application of postoperative autotransfusion in total joint arthroplasty reduces allogeneic blood requirements: a meta-analysis of randomized controlled trials

Abstract Background Total joint arthroplasty is associated with significant blood loss and often requires blood transfusion. However, allogeneic blood transfusion (ABT) may lead to severe problems, such as immunoreaction and infection. Postoperative autotransfusion, an alternative to ABT, is controversial. We conducted a meta-analysis to evaluate the ability of postoperative autotransfusion to reduce the need for ABT following total knee arthroplasty (TKA) and total hip arthroplasty (THA). Methods Systematic literature searches for randomized controlled trials were performed using PubMed, Embase, and the Cochrane Library until February 2016. Relative risks (RRs) and weighted mean differences with 95% confidence intervals (CIs) were calculated using fixed-effect or random-effect models; we also evaluated publication bias and heterogeneity. Results Seventeen trials with a total of 2314 patients were included in the meta-analysis. The pooled RRs of ABT rate between autotransfusion and the regular drainage/no drainage groups for TKA and THA were 0.446 (95% CI = 0.287, 0.693; p < 0.001) and 0.757 (95% CI = 0.599, 0.958; p = 0.020), respectively. In the subgroup analysis performed in TKA patients according to control interventions, the pooled RRs were 0.377 (95% CI = 0.224, 0.634; p < 0.001) (compared with regular drainage) and 0.804 (95% CI = 0.453, 1.426, p = 0.456) (compared with no drainage). In the subgroup analysis performed for THA, the pooled RRs were 0.536 (95% CI = 0.379, 0.757, p < 0.001) (compared with regular drainage) and 1.020 (95% CI = 0.740, 1.405, p = 0.904) (compared with no drainage). Conclusions Compared to regular drainage, autotransfusion reduces the need for ABT following TKA and THA. This reduction is not present when comparing autotransfusion to no drainage. However, the reliability of the meta-analytic results concerning TKA was limited by significant heterogeneity in methods among the included studies