J Orthop Traumatol. 2026 Jun 5. doi: 10.1186/s10195-026-00937-3. Online ahead of print.
ABSTRACT
BACKGROUND: Robotic-assisted total knee arthroplasty (RA-TKA) is increasingly being adopted for its ability to enhance bone-resection accuracy and component alignment. However, whether these technical gains influence the risk of periprosthetic joint infection (PJI) remains uncertain, especially in the context of prolonged operative duration. This study aimed to compare the 1-year rate of PJI following conventional total knee arthroplasty (cTKA) and RA-TKA in a propensity-score-matched cohort.
METHODS: We retrospectively reviewed 1284 consecutive patients who underwent primary TKAs at a single centre between 2021 and 2023. The patients were stratified according to surgical technique (cTKA versus RA-TKA) and subsequently matched 1:1 using propensity score analysis (age, sex, body mass index [BMI], American Society of Anesthesiologists [ASA] score, Charlson Comorbidity Index [CCI] score, CCI components and smoking), resulting in 522 pairs (1044 patients) for the final comparative analysis. Operative time and 1-year PJI were assessed using multivariable logistic regression. Infections were stratified according to timing: ≤ 90 days and from 90 days to 1 year after surgery.
RESULTS: The 1-year rate of PJI was 0.77% (4/522) after RA-TKA and 0.96% (5/522) after cTKA (P = 1.000). All PJIs in patients who underwent RA-TKA occurred within 90 days, whereas PJIs in patients who underwent cTKA occurred in both time windows. Multivariable logistic regression analysis did not identify surgical modality as an independent predictor of PJI (adjusted odds ratio [OR] 0.75, 95% confidence interval [CI] 0.22-2.90; P = 0.57). The median operative time was longer in the RA-TKA group than in the cTKA group (115 (range, 90-145) versus 85 (range, 60-105) min; P < 0.001).
CONCLUSIONS: RA-TKA was associated with a longer operative time, while no statistically significant difference in 1-year PJI rates was detected compared with cTKA. Nevertheless, these findings should be interpreted cautiously given the limited number of infection events.
LEVEL OF EVIDENCE: Level 3, non-randomised observational study.
PMID:42247101 | DOI:10.1186/s10195-026-00937-3
