DEFINING SUCCESS AND FAILURE IN PROSTHETIC JOINT INFECTIONS: A META-EPIDEMIOLOGIC STUDY TOWARDS A CORE OUTCOME SET

Aim A consensus‑driven definition of prosthetic joint infection (PJI) outcomes is essential for unified data collection, cross‑study comparisons, and evidence‑based guidelines. Such consistency refines treatment strategies, informs decisions, and enhances patient care. We systematically summarized the evidence — using a rigorous meta‑epidemiological approach — on how PJI outcomes are defined, building on analyses of native vertebral osteomyelitis and postoperative spinal infection definitions. Methods We searched Cochrane, Embase, Medline, and Scopus without date or language limits for publications that provided a specific PJI outcome definition. Eligible randomized controlled trials and cohort studies reported at least ten PJI cases. Following meta‑epidemiological guidelines adapted from PRISMA, we extracted verbatim definitions of PJI outcomes and categorized them into thematic criteria using a saturation approach. A network plot displayed the frequency and co‑occurrence of criteria across studies (Figure 1). Results Of 7,584 screened studies, 552 (7.3 %) met inclusion criteria. This pilot analysis covers 49 randomly selected studies (8,162 patients), mostly retrospective (48/49, 98 %), concerning mixed joint replacements (25/49, 51 %), two‑stage exchanges (15/49, 31 %), or mixed surgeries (24/49, 49 %). 25/49 studies (51 %) exceeded two-years follow-up. Only 6/49 (12.2 %) used a society‑endorsed definition, and 13/49 (26.5 %) cited a reference. Overall, 32/49 (65.3 %) defined “success”, 35/49 (71.4 %) “failure”, and 16/49 (32.7 %) both. Patient‑reported outcomes appeared in 14/49 (28.6 %), employing 14 PROMs, most frequently UCLA, HHS, and KSS; 13/49 (26.5 %) assessed functional outcomes. Sixteen outcome criteria clustered into seven domains: clinical signs, microbiology, imaging, histopathology, antibiotics‑related, surgical‑related, and mortality. Network analysis revealed strong links among clinical, microbiology, surgical, imaging, and mortality domains, whereas histopathology and antibiotics‑related criteria were infrequent. Conclusion Substantial heterogeneity in PJI outcome definitions and limited uptake of standardized criteria remain. Reliance on clinical and microbiological measures impedes comparability, particularly given underused histopathology, antibiotic‑related endpoints, and patient‑reported measures, whose perspectives may diverge from clinical parameters. A unified, consensus‑driven approach would enhance evidence synthesis and ultimately improve patient care in prosthetic joint infections. For any figures or tables, please contact the authors directly.