Synovial-Based Tests Outperform Serum Markers to Rule Out Infection in Total Knee Arthroplasty and Total Hip Arthroplasty: A Systematic Review and Meta-Analysis

Background Total knee arthroplasty (TKA) provides successful results in most patients. Periprosthetic joint infection (PJI) accounts for up to 25% of failed TKAs needing revision. In clinical practice, consensus in diagnostic strategy for excluding or diagnosing PJI is still lacking. In this systematic review and meta-analysis, we aim to provide a simplified data-driven diagnostic strategy for aseptic knee and hip revision surgeons to rule out PJI in the outpatient clinic phase. Methods A literature search in EMBASE, MEDLINE, PubMed, and Cochrane was conducted. Studies involving the diagnosis of PJI in patients with failed TKAs and total hip arthroplasties needing revision were identified. Only studies using the Musculoskeletal Infection Society criteria were included. Quality was assessed using MINORS criteria. Meta-analysis was performed for each diagnostic test identified in the included studies. Pooled estimates of diagnostic accuracy measures were calculated using a bivariate model and plotted in summary receiver–operator characteristic curves. Positive and negative predictive values were calculated in a hypothetical sample of patients with a given disease prevalence. Results Twenty-four studies met the inclusion criteria, describing a total of 2974 patients. Quality scores ranged from 13 to 19. Meta-analysis could be performed on 7 unique diagnostic tests. Highest pooled sensitivity and specificity were demonstrated for α-defensin with values of 86% and 96.6%, respectively. α-defensin and white blood cell count in synovial fluid demonstrate highest negative predictive value values. Conclusions We recommend, in a clinical setting with low–intermediate prevalence of PJI, performing arthrocentesis and joint fluid analysis using α-defensin and/or white blood cell count before revision TKA and revision total hip arthroplasty surgery to rule out PJI.