Full bibliography

Background This study aimed to identify the success rate of debridement, antibiotics, and implant retention (DAIR) for prosthetic joint infection (PJI) in a large prospective cohort of patients undergoing total knee arthroplasty (TKA). The ability for different PJI classification systems to predict success was assessed. Methods Prospective data recorded in the Prosthetic Joint Infection in Australia and New Zealand Observational study were analyzed. One hundred eighty-nine newly diagnosed knee PJIs were managed with DAIR between July 2014 and December 2017. Patients were prospectively followed up for 2 years. A strict definition of success was used, requiring the patient being alive with documented absence of infection, no ongoing antibiotics and the index prosthesis in place. Success was compared against the Coventry (early PJI ≤1 month), International Consensus Meeting (early ≤90 days), Auckland (early <1 year), and Tsukayama (early ≤1 month, hematogenous >1 month with <7 days symptoms, chronic >1 month with >7 days symptoms) classifications. Results DAIR success was 45% (85/189) and was highest in early PJIs defined according to the Coventry (adjusted odds ratio [aOR] = 3.9, P = .01), the International Consensus Meeting (aOR = 3.1, P = .01), and the Auckland classifications (aOR = 2.6, P = .01). Success was lower in both hematogenous (aOR = 0.4, P = .03) and chronic infections (aOR = 0.1, P = .003). Conclusion Time since primary TKA is an important predictor of DAIR success. Success was highest in infections occurring <1 month of the primary TKA and progressively decreased as time since the primary TKA increased.