Abstract
INTRODUCTION: Periprosthetic joint infection (PJI) remains a serious complication after total knee arthroplasty (TKA). While debridement, antibiotics, and implant retention (DAIR) are considered for acute PJI, success rates vary. This study aims to assess a new scoring system's accuracy in predicting DAIR success. METHODS: 119 TKA patients (2008-2019) diagnosed with PJI who underwent DAIR were included for analysis. Data were collected on demographics, laboratory values, and clinical outcomes. This was used for validation of the novel classification system consisting of PJI acuteness, microorganism classification, and host health for DAIR indication. Statistical analysis was carried out using SPSS programming. RESULTS: Mean follow-up was 2.5 years with an average age of 65.5 ± 9.1 years, BMI of 31.9 ± 6.2 kg/m(2), and CCI of 3.04 ± 1.8. Successful infection eradication occurred in 75.6% of patients. The classification system demonstrated 61.1% sensitivity, 72.4% specificity, and 87.3% positive predictive value (PPV) when the DAIR cutoff was a score less than 6. For a cutoff of less than 8, sensitivity was 100%, specificity was 37.9%, and PPV was 83.3%. CONCLUSIONS: To date, no consensus exists on a classification system predicting DAIR success. This novel scoring system, with high PPV, shows promise. Further refinement is essential for enhanced predictive accuracy.