Diagnosis of Late Periprosthetic Joint Infection. Which Diagnostic Algorithm to Choose

Background . Nowadays, according to the foreign and native registers data, the number of patients with periprosthetic infection (PJI) tends to increase. The early PJI diagnostics allows to provide timely effective treatment. Several widely used PJI diagnostic algorithms PJI exist. The objective of the study is comparative analysis of diagnostic value, accuracy and specificity of contemporary diagnostic algorithms. Materials and methods . A post-hoc analysis of 242 patients undergoing revision arthroplasty in 2018, held at FSFI FCTOE, was carried out. According to the study design, 127 patients were included in this study. PJI was diagnosed according to three known algorithms: ICM (International Consensus Meeting 2018), WAIOT (The World Association against Infection in Orthopedics and Trauma), EBJIS (The European Bone and Joint Infection Society 2018). Diagnostic sensitivity, specificity and total accuracy of each algorithm was carried out. The evaluation of ICM diagnostic algorithm was made with 2 variants: “not convincing = no infection”, “not convincing = infection”. The presence of infection was confirmed by bacteriological examination of synovial fluid aspirate, intraoperative biopsy of materials and sonification of explanted components. Results . The highest value of common accuracy was achieved in ICM 2018 algorithm — “not convincing = infection” was 91.3%, with sensibility and specificity — 89.3% and 93.0% respectively. The best specificity was shown by the algorithms WAIOT and ICM (“not convincing = no infection) – 95.8%, with sensibility and common accuracy — 80.4% and 89.0% respectively. The sensibility and specificity of EBJIS algorithm was 87.5% and 84.5%, respectively, the common accuracy — 85.8%. Conclusion . All included in investigation diagnostic algorithms showed high specificity values in diagnostics of hip and knee PJI without significant differences. Patients with subclinical PJI and low virulent pathogens have the biggest difficulties in PJI diagnostics. It seems that the selection of analyzed algorithms doesn’t play an important role, however PJI diagnostics requires complex approach with the use of different clinical and laboratory values.