Discriminating Osteoarticular Infection Bacterial Aetiology Using Immune-Inflammatory Biomarkers

Identifying the causative pathogen of paediatric osteoarticular infections (OAIs) is essential for guiding management. Kingella kingae infections are generally milder than pyogenic infections, which may have more severe and lasting consequences. Because standard inflammatory markers have limited discriminatory value, we assessed whether complete blood count-derived immune-inflammatory biomarkers (IIBs) could provide a simple, low-cost diagnostic aid. We retrospectively reviewed children younger than 5 years admitted to a tertiary hospital with confirmed OAIs. K. kingae cases were included from 2007 to 2025 and pyogenic cases from 1997 to 2025. Admission clinical data, conventional inflammatory markers, and six IIBs-NLR, MLR, PLR, SII, SIRI, and PIV-were analysed. The cohort included 118 K. kingae and 39 pyogenic OAIs. Pyogenic infections were associated with higher admission temperature and CRP levels and lower lymphocyte counts. All six IIBs were significantly increased in pyogenic infections and showed excellent univariable discrimination, with AUCs of 0.92-0.94. In a common complete-case cohort, the clinical model incorporating age, admission temperature, and CRP achieved an optimism-corrected AUC of 0.848. Adding lymphocyte count or an individual IIB increased corrected AUCs to 0.951-0.962; apparent AUCs were significantly higher than that of the clinical model after Holm adjustment. The abnormal-IIB-count model achieved the highest corrected AUC of 0.970; apparent sensitivity and specificity at the Youden-optimal threshold were 97.3% and 90.3%, respectively, but it did not significantly outperform the model with lymphocyte count alone. Routine CBC-derived IIBs may improve early differentiation between K. kingae and pyogenic OAIs, although external validation is required before clinical implementation.