Abstract. Introduction: Clinical criteria are included in all current periprosthetic joint infection (PJI) definitions, with sinus tract and purulence among the most unequivocal indicators of infection. However, evidence regarding their diagnostic accuracy remains limited. This study evaluated their diagnostic performance, associated microbial profile, and clinical outcomes. Methods: This retrospective study included 463 revision hip and knee arthroplasties between 2015 and 2023. A total of 245 (52.9 %) were classified as infected according to the 2021 European Bone and Joint Infection Society (EBJIS) definition. Receiver-operating-characteristic (ROC) curves and their areas under the curve (AUCs) were used for diagnostic accuracy calculations. Follow-up analysis was conducted using the Kaplan–Meier survival estimator. Results: Sensitivities for sinus tract and purulence were 12.7 % (95 % CI: 8.8–17.5) and 35.1 % (29.1–41.4), with specificities of 100 % (98.3–100) for both. AUCs were 0.546 (0.521–0.571) and 0.695 (0.653–0.737). The highest prevalence of sinus tract was observed in Candida albicans (50 %), polymicrobial infection (33 %), and Cutibacterium acnes (30 %), whereas purulence was most prevalent in Streptococcus spp. (73 %), Staphylococcus aureus (72 %), Enterobacteriaceae (50 %), and polymicrobial infections (43 %). Both were associated with lower 3-year infection-free survival; this was significant for sinus tract (p=0.011) and showed a trend for purulence (p=0.068). Conclusion: A sinus tract can be considered a confirmatory criterion for PJI due to its excellent specificity. Although highly specific, purulence remains subjective, with limited sensitivity, and should, if considered, be regarded only as a suggestive criterion. Both parameters were associated with lower survival rates and a consistent trend towards treatment failure.
Luger et al. (2026) studied this question.