Abstract
AIMS: Three-phase bone scintigraphy (TPBS) is a simple, reliable, and straightforward imaging modality for diagnosing periprosthetic joint infection (PJI). This study aimed to evaluate the diagnostic efficacy of the combined diagnostic method of TPBS and CRP in the detection of PJI. METHODS: A retrospective analysis was performed on 317 patients undergoing revision surgery (211 hips and 106 knees) between January 2017 and January 2024, with TPBS obtained prior to surgical intervention. PJI was diagnosed according to the 2013 Musculoskeletal Infection Society (MSIS) criteria. The diagnostic performance of TPBS, CRP, and TPBS-CRP combined method was evaluated. RESULTS: In the total knee arthroplasty (TKA) cohort, TPBS-CRP achieved maximal sensitivity (100%) and negative predictive value (100%), enabling definitive exclusion of PJI, albeit with reduced specificity (50.0%). Antibiotic therapy significantly reduced CRP accuracy (79.1% vs 97.1%, p = 0.020) but minimally impacted TPBS (100% vs 88.6%, p = 0.037). For total hip arthroplasty (THA), the TPBS-CRP demonstrated a superior diagnostic efficacy (AUC = 0.846, 0.834, and 0.795) and demonstrated a more balanced sensitivity (81.6%, 67.3%, and 64.3%) and specificity (87.6%, 91.2%, and 94.7%) compared to CRP and TPBS. Antibiotic-treated patients exhibited significantly higher diagnostic accuracy for TPBS (81.5% vs 57.7%, p = 0.028), CRP (88.9% vs 59.2%, p = 0.005), and TPBS-CRP (96.3% vs 76.1%, p = 0.020) compared to untreated patients. False-positive TPBS cases in both cohorts exhibited severe bone loss and prolonged symptoms compared to true negatives (p < 0.05). CONCLUSION: TPBS demonstrates limited specificity in differentiating PJI from aseptic loosening. The TPBS-CRP synergizes localized bone metabolic profiles with systemic inflammatory biomarkers, achieving superior diagnostic accuracy. Caution is warranted in patients with severe bone loss and prolonged symptoms.