Abstract
AIMS: Chronic periprosthetic joint infections (PJIs) caused by low-virulence pathogens can be challenging to differentiate from aseptic failure. The aim of this study was to assess the diagnostic accuracy of the most commonly used tests for diagnosing these PJIs. METHODS: A consecutive series of 420 patients undergoing revision total hip or knee arthroplasty were studied. Coagulase-negative staphylococci, Cutibacterium spp., enterococci, Actinomyces spp., and fungi were classified as low-virulence pathogens. Of the 207 PJIs defined by the European Bone and Joint Infection Society (EBJIS) criteria, 60 were chronic infections caused by low-virulence pathogens. A total of 213 cases were classified as aseptic, resulting in a total of 273 cases included in the analysis. The performance of established test methods was assessed using receiver operating characteristic (ROC) curves. RESULTS: The calculated synovial fluid percentage of polymorphonuclear neutrophils (SF-%PMN) cut-off of > 67% demonstrated the best preoperative performance with a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) of 97.1% (95% CI 83.9 to 100), 97.2% (95% CI 91.9 to 100), 97.1% (95% CI 91.6 to 100), 97.2% (95% CI 91.9 to 100), and 0.972 (95% CI 0.933 to 1.000), followed by SF white blood cell count (SF-WBC) > 2,403 cells/µl with estimates of 85.4% (95% CI 71.1 to 93.4), 100% (95% CI 94.6 to 100), 100% (100), 93.3% (95% CI 88.0 to 98.5), and 0.927 (95% CI 0.872 to 0.982). While a %PMN cut-off of > 67% outperformed the established threshold of > 80% (p = 0.010), the calculated SF-WBC cut-off of > 2,403 G/l did not perform significantly better than the established threshold of > 3,000 G/l (p = 0.266). Intraoperatively, permanent paraffin-fixed sections showed the best accuracy with performance values of 89.3% (95% CI 78.1 to 95.3), 100% (95% CI 97.7 to 100), 100% (100), 97.1% (95% CI 94.8 to 99.4), and 0.946 (95% CI 0.906 to 0.987), followed by frozen sections (85.7% (95% CI 67.7 to 94.6), 99.2% (95% CI 95.1 to 100), 95.8% (95% CI 87.8 to 100), 96.9% (95% CI 94.0 to 99.9), and 0.922 (95% CI 0.853 to 0.991)). CONCLUSION: Synovial fluid %PMN and WBC, and permanent histological sections demonstrated the highest diagnostic performance for diagnosing PJI caused by low-virulence pathogens. Intraoperative frozen sections proved to be a valuable intraoperative tool, particularly in cases with unclear preoperative diagnosis. A lower cut-off for SF-%PMN may be considered to more accurately diagnose these infections.