Abstract
Aim
Microbiological culture of intraoperative periprosthetic tissue samples (IPTS) is one of the main criteria in diagnosing prosthetic joint infections (PJI) as stated by different guidelines. The current techniques are labor-intensive, prone for contamination and show low sensitivity. The aim of this study was to evaluate the added value of beadmill processing of IPTS and culturing in blood culture bottles (BCBs) over the conventional method of standard agar and broth alone.
Method
We conducted a single-center prospective study from May 2017 to January 2018 at the GZA Hospitals, a secondary care hospital (1012 beds) in Antwerp, Belgium. IPTS from patients undergoing revision arthroplasty were consecutively processed. Each IPTS was aseptically divided in two equal parts: one was processed by direct inoculation on agar and in broths (non-homogenized method); the other was transferred in a sterile vial with saline solution and glass beads (EOLabs), homogenized using a mechanic cell disruptor for 30s (Disruptor genie, Scientific Industries), 2mL of the suspension was inoculated in (an)aerobic BCBs, agar plates and broths (homogenized method). Agar plates were incubated for 4d; broths and BCBs in BacT/Alert (bioMerieux) for 14d. Micro-organisms were identified using MALDI-TOF MS (Bruker). Sensitivity (Se) and specificity (Sp) were calculated against the IDSA definition of PJI for different culture sets: non-homogenized and agar/broth; homogenized processing and agar/broth, agar/broth/BCB, agar/BCB. Ethics committee approved the study.
Results
Overall, 122 IPTS from 32 episodes from 29 patients were included; 14 subjects met the IDSA PJI criteria. No difference (Se71.4%, Sp88.8%) was observed between the conventional method and the homogenized method for the agar and broth set. An increased sensitivity was observed for the homogenized method with addition of BCBs (Se85.7%) in contrast to agars and broths alone (Se71.4%). The homogenized method with BCBs and agar plates showed an excellent specificity and positive-predictive value, indicating the lower contamination risk and facilitating the microbiological diagnosis. Adding broths to this combination increases the false positivity rate (Sp94.4%). A false positivity rate of 15/122 IPTS was observed for broths alone, in contrast to 2/122 for BCBs alone. Also, one case (1/14) would have been missed when using the homogenized method with agars and broths alone.
Conclusions
Beadmill processing of IPTS and culturing in BCBs is a sensitive and highly specific culture method for diagnosis of PJI. The superior specificity versus conventional methods minimizes false positive results, which frequently lead to erroneous clinical decisions. Furthermore, this makes semi-automated laboratory processing possible.
