COMPARISON OF INFECTION DEFINITIONS IN DIAGNOSING LOW-GRADE PERIPROSTHETIC JOINT INFECTIONS

Aim The publication of infection definitions by expert societies has standardized the diagnosis of periprosthetic joint infections (PJI) in the last two decades. While some studies have been undertaken to confirm and compare their clinical usefulness, detailed analyses of “low-grade” PJI are still lacking. The aim of this study was to assess and compare the results of the various infection definitions in this challenging subgroup of patients. Methods From 2015–2024, 463 patients with revision surgery after total hip and knee arthroplasty were included. A standardized workup was performed, and the PJI-definitions of the European Bone and Joint Infection Society (EBJIS) 2021, International Consensus Meeting (ICM) 2018, Musculoskeletal Infection Society (MSIS) 2011 and 2013, Infectious Diseases Society of America (IDSA) 2013, as well as the newly proposed Unified Definition were used for the diagnosis of PJI. Low-grade infections were defined as chronic infections (symptom duration >3 weeks) with low-virulence microorganisms (coagulase-negative staphylococci, Cutibacterium spp., Enterococcus spp., Actinomyces spp ., and fungi). Results After excluding acute infections, 278 cases were included in the final analysis; 214 were not classified as “confirmed infected” by any of the infection definitions. The Unified, EBJIS, ICM 2018, MSIS 2011 and 2013, and IDSA definitions classified 62 (22.3%), 63 (22.7%), 58 (20.9%), 61 (21.9%), 54 (19.4%), and 62 patients (22.3%) as septic, respectively ( p >0.05). Concordance between all definitions (infected/not infected) was 96.4% (268/278). The EBJIS definition identified all infections diagnosed by the other definitions. One case diagnosed as infected by the MSIS 2013, ICM 2018, and EBJIS Definitions was missed by the Unified Definition. No infection was diagnosed exclusively by a single definition. 3 (1.1%), 25 (9.0%), and 30 cases (10.8%) had an uncertain diagnosis according to the Unified, EBJIS, and ICM 2018 definitions ( p =0.001). The 3 “probable” infections according to the Unified Definition were classified as “infection likely” by EBJIS. Conclusions In chronic PJI caused by low-virulence microorganisms, the diagnosis remains challenging and depends on the chosen definition. Considerable heterogeneity was observed in uncertain diagnoses across definitions, with significantly fewer uncertain diagnoses when using the Unified Definition. Differences in preoperative sensitivity between definitions may partly be explained by how individual diagnostic criteria are weighted, such as the classification of synovial fluid leukocyte count as a suggestive rather than a standalone criterion. However, as there is no universally accepted gold standard for diagnosing true PJI, these findings should be interpreted with caution. In clinical practice, it is important to understand the differences between the available criteria to accurately interpret the diagnostic findings and their impact on clinical outcomes.