Routine intraoperative infection testing in presumed aseptic hip and knee revision: a matched cohort retrospective...

Periprosthetic joint infections (PJIs) complicate 1-2% of primary and up to 4% of revision arthroplasties, and their diagnosis remains challenging due to the lack of a diagnostic gold standard. Differentiating PJI from aseptic failure is particularly difficult in low-grade or biofilm-related infections. Routine intraoperative microbiological screening is not recommended in presumed aseptic revisions. This study aimed to evaluate the association between routine intraoperative microbiological screening and implant survival in presumed aseptic revision arthroplasty. A secondary objective was to identify preoperative predictors of unexpected PJIs.

Retrospectively collected data of patients undergoing presumed aseptic THA and TKA revisions between 2013 and 2019 were included. Two matched cohorts were compared: a screened group (2016-2019) undergoing routine intraoperative microbiological sampling, and an unscreened one (2013-2015) without systematic screening. Kaplan-Meier survival analysis with log-rank testing assessed overall and infection-free survival. Univariate conditional logistic regression and receiver operating characteristic (ROC) analyses evaluated potential associations between preoperative serum markers and unexpected PJI.

A total of 559 patients were included, of whom 295 underwent screening and 264 did not. Unexpected infections occurred in 20.3% (n = 60) of screened patients. No significant differences in implant survival were observed between groups. Although overall and infection-free survival were numerically higher in the screened cohort, these differences were not statistically significant (p = 0.168 and p = 0.118, respectively). Stratified analyses by joint and failure mechanism showed comparable findings. C-reactive protein and erythrocyte sedimentation rate were associated with unexpected PJI (p = 0.004 and p = 0.046), but ROC analysis suggested limited discrimination of individual markers, whereas a combined serum model improved diagnostic performance (AUC = 0.833).

Routine intraoperative microbiological screening in presumed aseptic revision arthroplasty was not associated with a significant improvement in implant survival. Future studies should investigate whether selective screening improves outcomes. Level III retrospective cohort study.