Suction drainage is associated with early hemoglobin decline without detectable clinical benefit after total joint arthroplasty: a propensity score–matched study

Abstract Purpose To evaluate the impact of routine postoperative suction drainage on perioperative outcomes following primary total hip arthroplasty (THA) and total knee arthroplasty (TKA) using a propensity score–matched cohort. Methods A retrospective cohort study included 2,339 consecutive patients who underwent primary THA or TKA between January 2020 and May 2026. Patients were categorized according to postoperative suction drain use. Propensity score matching using age, sex, and body mass index, with exact matching for procedure type, yielded 630 matched pairs (1,260 patients). Primary outcomes were postoperative hemoglobin (Hb) change and estimated blood loss (EBL). Secondary outcomes included transfusion, length of stay, and postoperative infectious, wound, and thromboembolic complications. Subgroup analyses were performed according to procedure type, antiplatelet use, and anticoagulant use. Results Following propensity score matching, the suction drain group demonstrated a greater reduction in Hb on POD1 (2.0 ± 1.0 vs. 1.8 ± 0.9 g/dL, p = 0.005), whereas Hb change on POD2 was comparable between groups. Estimated blood loss, transfusion rate, length of stay, postoperative wound complications, thromboembolic events, and acute periprosthetic joint infection were similar between groups. Procedure-specific subgroup analyses demonstrated generally consistent findings in both THA and TKA. Among patients receiving antiplatelet or anticoagulant medications, postoperative hemoglobin decline, estimated blood loss, transfusion rate, and length of stay were comparable between the suction drain and no-suction drain groups. Conclusions Routine suction drain use following primary total joint arthroplasty was primarily associated with early postoperative laboratory differences, without meaningful differences in clinically relevant perioperative outcomes. Level of Evidence: Retrospective cohort study, Level Ⅲ