Abstract
BACKGROUND: The optimal initial management of adult native knee septic arthritis remains debated. We compared outcomes across arthroscopic irrigation and debridement, open arthrotomy, and aspiration-led medical therapy, and developed a locally calibrated Weighted-Incidence Syndromic Combination Antibiogram (WISCA) to guide empiric antibiotics. METHODS: Single-center retrospective cohort of 136 adults (2018-2025). Index modalities were open arthrotomy (n = 88), arthroscopy (n = 32), and medical therapy (n = 16). Primary outcomes were unplanned reintervention, number of procedures, complications, and hospital length of stay (LOS). Secondary outcomes included conversion to open surgery, knee range of motion, and in-hospital mortality. Multivariable models (bias-reduced logistic for binary endpoints, Poisson/negative-binomial for counts, and Gamma log-link for LOS) adjusted for demographics, comorbidities, and symptom-to-treatment time. WISCA estimated regimen-level empiric coverage using a hierarchical Bayesian model weighted by local pathogen incidence and susceptibility. RESULTS: Arthroscopy required fewer procedures (median 2.0 vs 2.5) and had shorter LOS (median 5.5 vs 11.5 days) than arthrotomy (both p ≤ 0.002). Adjusted analyses confirmed fewer procedures (incidence rate ratio 0.67, 95 % confidence interval 0.51-0.88) and shorter LOS (ratio of means 0.37, 95 % confidence interval 0.26-0.53) with arthroscopy, without a clear increase in unplanned reintervention. Medical treatment and aspiration had the highest conversion to surgery (37.5 %) and more downstream complications. Longer symptom duration, diabetes, and cirrhosis independently predicted greater procedure burden. The WISCA favored β-lactam-anchored regimens with anti-staphylococcal coverage for most patients and supported initial narrow-spectrum β-lactam choices. CONCLUSION: In routine care, arthroscopic irrigation and debridement achieves infection control with lower resource use than open arthrotomy and should be the preferred first-line surgical strategy for most presentations. Open arthrotomy remains important for anatomically complex or advanced disease, and aspiration-only strategies should be limited to carefully selected, closely monitored patients. A locally calibrated WISCA can support rational empiric choices and timely de-escalation alongside early source control.