Septic arthritis is a diagnostic and therapeutic emergency that threatens both life and function. The primary objective of this study was to assess the efficacy on the infectious process of arthroscopic treatment in patients with septic arthritis of native joints. The secondary objective was to identify factors predicting failure to achieve infection resolution after arthroscopic treatment. We hypothesised that arthroscopy was the appropriate treatment strategy.

Forty-six cases of septic arthritis in 46 patients with a mean age of 46 years (range, 18–72 years) were retrospectively reviewed. The cause of the septic arthritis was haematogenous dissemination in 39.1% of patients, surgery in 34.8%, a local injection in 19.6%, and trauma in 6.5%. The involved joint was the knee in 32 patients, the shoulder in 6, the hip in 3, the ankle in 3, and the elbow in 2. All patients underwent arthroscopic joint lavage, with or without synovectomy depending on the Gächter stage. Dual antibiotic therapy was given routinely after the procedure. For each patient, we assessed time to treatment, intraoperative findings according to the Gächter classification, cultures of drainage-fluids, and whether repeat arthroscopic lavage was required. Recovery of the infection was defined as absence of clinical or laboratory signs of infection at last follow-up.

Mean follow-up was 42 months (range, 1–120). Mean time from symptom onset to arthroscopic treatment was 7.5 days. Full recovery of the infection was achieved in 93% of patients, although 25% required more than one arthroscopic lavage. Factors significantly associated with arthroscopic treatment failure were Gächter stage III or IV and positive drainage-fluid cultures after 24h.

Arthroscopic treatment is indicated in all patients with septic arthritis on native joints. The procedure should be repeated if the initial course is unfavourable.

IV. Retrospective study.