Zhipeng Wu, Yi Zhang, Zhaowei Jiang, Dong Lv, Zan Shen, Bangjian He
Although robotic-assisted total knee arthroplasty (R-TKA) enhances the accuracy of component alignment, its relationship with postoperative infection risk remains uncertain. This study systematically compares infection rates following R-TKA versus manual TKA (M-TKA). A systematic search of four bibliographic databases was undertaken to retrieve comparative studies that evaluated postoperative infection rates in patients undergoing R‑TKA versus M‑TKA. Eligible studies were required to report at least one infection‑related endpoint, including those explicitly reporting zero infection events. In total, 87 studies satisfied the selection criteria, and data were combined using a random‑effects meta‑analytic framework to account for between‑study heterogeneity. R-TKA was associated with a reduced overall infection rate compared with M-TKA (P = 0.009) and a reduced risk of PJI (P = 0.007). These associations remained consistent across exploratory subgroup analyses based on clinically relevant factors (P < 0.05). Follow-up stratified analyses showed comparable periprosthetic joint infection (PJI) risk between R-TKA and M-TKA within 90 days and at 1- and 2-year follow-up. In contrast, R-TKA was linked to a reduced risk of PJI after more than 2 years of follow-up (P = 0.02). Conversely, there was no statistically significant difference in surgical site infection between the groups (P = 0.12). This study indicated that R-TKA was not linked to an increased 2-year infection risk and was associated with a reduced incidence of PJI in longer-term follow-up, suggesting comparable short-term infection outcomes and a potential association with lower long-term PJI risk compared with conventional M-TKA. The study protocol was prospectively registered in the PROSPERO database (Registration No. CRD420251114140).