Value of the synovial C-reactive protein test in the diagnosis of total hip and knee periprosthetic joint infections: A case-control study.

INTRODUCTION: The diagnosis of periprosthetic joint infection (PJI) can be challenging and rests on several principles. The use of diagnostic biomarkers, such as the synovial C-Reactive Protein (CRP), seems promising. The purpose of this study was to determine whether synovial CRP was a more discriminating test than serum CRP for the diagnosis of hip and knee PJI. MATERIALS AND METHODS: In total, 194 patients were included in this single center prospective study: 42 primary arthroplasties (control group [CG]), 111 revisions for aseptic prosthesis (aseptic revision group [ARG]), and 41 revisions for septic prosthesis (septic revision group [SRG]) based on the Musculoskeletal Infection Society (MSIS) criteria. RESULTS: The serum and synovial CRP levels were significantly higher in the SRG than the other two groups (SRG serum CRP=75.6mg/L vs. ARG serum CRP=6mg/L and CG serum CRP=2.7mg/L, p<0.001; SRG synovial CRP=31.5mg/L vs. CG synovial CRP=2.6mg/L and ARG synovial CRP=1.7mg/L, p<0.001). The positive likelihood ratios (LR+) were very similar for both the synovial CRP cut-off value of 4.4mg/L (LR+=7.04; sensitivity [Se] 82.5%, specificity [Sp] 88.3%) and the serum CRP cut-off value of 9mg/L (LR+=6.3; Se 87.5%, Sp 86.1%). CONCLUSION: This study showed that synovial CRP testing was not more discriminating than serum CRP in the diagnosis of hip and knee PJI. A serum CRP level greater than 9mg/L was a sign of PJI. LEVEL OF EVIDENCE: III; case-control study.