Optimization of sampling protocol, specimen types, and sampling sites in periprosthetic joint infections: a retrospective study

BACKGROUND: Intraoperative culture is critical for diagnosing periprosthetic joint infection (PJI) and antibiotic guidance, yet its value is limited by high culture-negative rates. To improve microbial yield, we developed a multi-site intraoperative sampling protocol (MSP) with following objectives: (1) to evaluate whether MSP could improve the yield of cultures; (2) to identify the optimal specimen types and anatomical sites for culture. METHODS: A retrospective study was conducted on 163 PJIs from 2015 to 2025. Among these cases, 69 underwent traditional intraoperative sampling protocol (TSP), while 94 underwent MSP. Pathogen detection rates and contamination rates were compared between the two groups. Additionally, culture positivity rates were compared across various specimen types and anatomical sites within MSP group. RESULTS: The MSP showed significantly higher pathogen detection rates than TSP (P = 0.003), consistent across knee (P = 0.038) and hip PJIs (P = 0.034). The adoption of MSP (OR = 2.430, P = 0.030) was independent predictor of pathogen detection. Regarding specimen type, synovial fluid (unadjusted P < 0.001, adjusted P = 0.001), pseudocapsule (unadjusted P < 0.001, adjusted P = 0.001), and interface membrane (unadjusted P = 0.001, adjusted P = 0.004) had higher culture positivity rates than synovium. Besides, synovial fluid had higher culture positivity rates than sinus tract (unadjusted P = 0.012, FDR-adjusted P = 0.046). Regarding sampling site, for hip, femoral neck specimens demonstrated higher culture positivity rates than medullary cavity (unadjusted P = 0.008, adjusted P = 0.040) and superficial layer specimens (unadjusted P = 0.008, adjusted P = 0.040). For knee, femoral intercondylar area specimens showed higher culture positivity rates than tibial medullary cavity (unadjusted P = 0.003, adjusted P = 0.037), infrapatellar fat pad (unadjusted P = 0.002, adjusted P = 0.037). CONCLUSIONS: The MSP enhanced the yield of cultures without increasing the contamination rate. The optimal specimen types for culture included synovial fluid, interface membrane, and pseudocapsule. For hip, the optimal anatomical site for culture was femoral neck. For knee, the optimal anatomical site for culture was intercondylar area.