Abstract
The diagnosis of periprosthetic joint infection (PJI) is challenging, often requiring multiple clinical specimens and diagnostic techniques, some with prolonged result turnaround times. Here, the diagnostic performance of the Investigational Use Only (IUO) BioFire Joint Infection (JI) Panel was compared to 16S rRNA gene-based targeted metagenomic sequencing (tMGS) applied to synovial fluid for PJI diagnosis. Sixty synovial fluid samples from knee arthroplasty failure archived at -80°C were tested. Infectious Diseases Society of America (IDSA) diagnostic criteria were used to classify PJI. For culture-positive PJI with pathogens targeted by the JI panel, JI panel sensitivity was 91% (21/23; 95% confidence interval [CI], 73 to 98%), and tMGS sensitivity was 96% (23/24; 95% CI, 80 to 99%) (P = 0.56). Overall sensitivities of the JI panel and tMGS for PJI diagnosis were 56% (24/43; 95% CI, 41 to 70%) and 93% (41/44; 95% CI, 82 to 98%), respectively (P < 0.001). JI panel and tMGS overall specificities were 100% (16/16; 95% CI, 81 to 100%) and 94% (15/16; 95% CI, 72 to 99%), respectively. While the clinical sensitivity of the JI panel was excellent for on-panel microorganisms, overall sensitivity for PJI diagnosis was low due to the absence of Staphylococcus epidermidis, a common causative pathogen of PJI, on the panel. A PJI diagnostic algorithm for the use of both molecular tests is proposed.