THE MANAGEMENT OF PERIPROSTHETIC JOINT INFECTION IN NON-SPECIALIST HOSPITALS WITHIN THE UK: A MULTICENTRE STUDY

Aim Management of periprosthetic joint infection (PJI) is infrequently reported from non-specialist hospitals. We identified factors associated with treatment failure for hip or knee PJI in non-specialist hospitals in order to identify cases that should be referred to a specialist infection centre with a bone infection multi-disciplinary team (MDT). Methods Patients having surgery for EBJIS confirmed hip or knee PJI were eligible for inclusion, at 5 participating non-specialist hospitals within the UK between 2014 and 2024. Primary outcome was treatment failure (life-long antibiotic suppression, amputation, mortality within 90 days of index procedure, failure of reimplantation or surgery for confirmed recurrent/persistent infection). Secondary outcomes included duration of antibiotics post-index procedure and referral to specialist centres. A multi-variate regression assessed factors hypothesised to predict outcome, which included index procedure, JS-BACH classification, duration of antibiotic therapy, and operating surgeon experience. Results 223 patients were eligible for inclusion (108 hip; 115 knee). There were 115 (51.6%) treatment failures, and 88 (39.5%) episodes of recurrent/persistent infection. Variables predictive of treatment failure were index procedure being an open washout (OR 4.8 95% CI 1.3–17.5), having complex joint-specifics (JS2 or JS3, OR 5.7 95%CI 2.5–13.3), and compromised host health (H2; OR 3.7 95%CI 1.8–7.7, H3; OR 8.4 95%CI 1.4–50.2). Surgeon experience, and duration of post-operative antibiotics were not associated with treatment failure. The median antibiotic duration post-index procedure was 84 days (12 weeks) (IQR 42.0-90 days) and 49 patients (21.2%) ended up on life-long antibiotics. Further surgery for infection occurred in 31 patients whilst on antibiotics (35.2%) and in 39 patients (44.4%) within 6 months of stopping antibiotics. Despite 148 (66.4%) patients having a local MDT input, only 27 (12.1%) patients were referred to specialist centres for further management. The majority of these patients (23/27, 85.2%) were referred to a specialist centre after the index procedure had failed. Conclusions More than half of all patients with hip and knee PJI had treatment failure in non-specialist hospitals, which is partly attributable to case complexity. Variables of the JS-BACH classification system identifiable before surgery were associated with treatment failure. Despite involvement of a local MDT, few patients were referred for specialist treatment. Awareness of the predictive factors for failure may aid better selection for early referral from non-specialist hospitals.