C-reactive protein may misdiagnose prosthetic joint infections, particularly chronic and low-grade infections

• Published in: International orthopaedics 2017-07, Vol.41 (7), p.1315-1319
• Main Authors: Pérez-Prieto, Daniel, Portillo, María E., Puig-Verdié, Lluís, Alier, Albert, Martínez, Santos, Sorlí, Lluisa, Horcajada, Juan P., Monllau, Joan C.
• Format: Article
• Language: English
• Subjects: Aged; Arthroplasty, Replacement - adverse effects; Bacteria - pathogenicity; Biomarkers - blood; Blood Sedimentation; C-Reactive Protein - analysis; Candida - pathogenicity; Chronic Disease; Diagnostic Errors - prevention & control; Female; Humans; Male; Medicine; Medicine & Public Health; Microbiological Techniques; Middle Aged; Original Paper; Orthopedics; Prosthesis-Related Infections - diagnosis; Prosthesis-Related Infections - immunology; Prosthesis-Related Infections - microbiology; Retrospective Studies; Synovial Fluid - immunology; Synovial Fluid - microbiology
• ISSN: 0341-2695; 1432-5195
• DOI: 10.1007/s00264-017-3430-5

Background Periprosthetic tissue cultures, sonication and synovial fluid cultures remain the gold standard for prosthetic joint infection (PJI) diagnosis. However, some 15–20% culture-negative PJI are still reported. Therefore, there is the need for other diagnostic criteria. One point of concern relative to the different definitions of PJI is as to the inclusion of the c-reactive protein (CRP) and the erythrocyte sedimentation rate (ESR) as diagnostic criteria for PJI despite them being non-specific inflammatory blood tests. Purpose The purpose of the present study was to determine the relevance of CRP and the ESR in the diagnosis of PJI. Methods All PJI with positive cultures over a two-year period in two hospitals were reviewed. The main variables of the present study were the type of prosthesis and the CRP level. More information was recorded in those patients with normal CRP: radiographs, physical examination records and the ESR. Results Seventy-three patients were included in study. Pre-operative CRP levels were normal (lower than 0.8 mg/dl) in 23 patients, representing 32% of all PJI with positive cultures. Low virulence micro-organisms, 12 coagulase-negative staphylococci and four P. acnes , grew in most of them. They represented 70% of all PJI with normal CRP levels. In addition, 17 patients (23% of all PJI with positive cultures) had a normal ESR, a normal physical examination (they only presented with pain) and no clear loosening was observed in the radiographs. Conclusions Per the American Association of Orthopaedic Surgeons (AAOS) guidelines or the Musculoskeletal Infection Society (MSIS), 23% of the patients in the present study with PJI would never have been identified. Blood inflammatory markers such as the CRP level and ESR may not be accurate as diagnostic tools in PJI, particularly to identify low-grade and chronic PJI.