The incidence of septic arthritis is 4–10 per 100 000 individuals per year, a figure which increases to 30–60 per 100 000 individuals for patients having underlying joint disease or a prosthetic joint (Geirsson et al., 2008; Weston et al., 1999; Favero et al., 2008; Smith et al., 2006; He et al., 2023; Patel, 2023). Younger children and older adults are the most vulnerable groups (Roerdink et al., 2019). The prevalence of prosthetic joint infections (PJIs) ranges from 2.0 %–2.4 % after primary interventions, a figure which increases to 20 % in patients undergoing revision procedures (Signore et al., 2019). About 60 %–70 % of PJIs occur during the first 2 years post-surgery, and nearly 50 % of all revision arthroplasties are due to PJI, contributing to a 1 %–3 % annual mortality (Postler et al., 2018).

Septic arthritis (SA) on native joints is usually aetiologically monomicrobial; the pathogen profile is relatively homogenous across patient populations; and it is associated with acute clinical manifestations, with classic signs and symptoms of swelling, pain, and redness at the infected site. Conversely, PJIs may be aetiologically mono- or poly-microbial and are commonly classified according to the time of presentation from implantation of the prosthesis (Zimmerli et al., 2004; Benito et al., 2019).

A number of fastidious pathogens which are more difficult to grow in culture can also be responsible for SA and PJI, including Kingella kingae, a common cause of SA in younger children (Hunter et al., 2022; Yagupsky, 2022), and Neisseria gonorrhoeae, an important cause of septic arthritis in newborns and sexually active individuals (Kleiman and Lamb, 1973; Moussiegt et al., 2022). Moreover, in an extensive review conducted by Tande and Patel (2014) including data from over 6700 specimens, the authors demonstrated that Staphylococcus aureus was the most common pathogen in early-onset PJI. These outcomes were consistent across other prominent publications for PJIs (Benito et al., 2019; Peel et al., 2012) and SA (Horowitz et al., 2011; He et al., 2023). When acute haematogenous cases were analysed, Staphylococcus aureus alone was identified in 68 % of samples (Stefánsdóttir et al., 2009). Thereafter, the most common pathogens are Streptococcus species, Enterococcus species, and gram-negative bacilli (Tande and Patel, 2014; Stefánsdóttir et al., 2009; Wouthuyzen-Bakker et al., 2019). Most notably, polymicrobial infections contribute to more than 31 % of early infections and 15 % for all time-period infections (Tande and Patel, 2014).

Traditional culturing of pathogens for a definitive microbiological diagnosis of joint infections has several limitations. Aside from limitations related to the turnaround time, traditional cultures are associated with a substantial false negative rate (Schulz et al., 2021), most notably related to pre-analytical issues (e.g. SA patients previously treated with antibiotics have a higher risk of false negative results with culture methods. Moreover, the probability of having false-negative results of PJI cultures increases with late compared with early post-surgical period infections and ranges from 6.5 %–17 % (Stefánsdóttir et al., 2009; Aggarwal et al., 2014). In a review of 64 manuscripts, the average sensitivity and specificity were, 65.6 % and 94.4 % respectively, whereas for molecular diagnosis, sensitivity and specificity were 74.1 % and 95.2 % respectively (Cozzi Lepri et al., 2019). In another study, among 543 cases of SA, only 40 % had positive synovial fluid cultures (McBride et al., 2020).

Remarkably, molecular methods such as PCR are often unavailable in hospitals and may also take 1–2 d to obtain a diagnosis due to turnaround time. Finally, cultures are much more likely to miss polymicrobial infections (Schulz et al., 2021).

A missed or delayed diagnosis can have a tremendous impact on patients and put high financial strain on our healthcare budget. Moreover, this can result in extended rehabilitation trajectories, permanent disabilities, and psychological consequences. It is known to increase morbidity and mortality. Also, prolonged administration of inappropriate antimicrobials can result in antimicrobial resistance, addressed by the WHO as one of the main global public health threats we are facing in our society.

Molecular diagnostic methods are currently being explored in various settings, and their benefits in proving critical information about the nature of the causative organisms to optimise targeted therapy is evident (Kullar et al., 2023). The BIOFIRE^® Joint Infection (JI) Panel, a rapid syndromic multiplex-PCR assay (hereafter the JI Panel), is an in vitro diagnostic assay which detects 31 bacterial and yeast nucleic acids and selects eight antimicrobial resistance genes most involved in joint infections. The JI Panel has 91.7 % sensitivity and 99.8 % specificity. The fully integrated all-in-one system required 2 min of hands-on time and 0.2 mL of synovial fluid and provides diagnostic results within approximately 1 h (bioMérieux, 2022). In recent studies, positive percent agreement and negative percent agreement for the JI Panel and routine culture for on-panel organisms on SA and PJIs were 91.6 % and 93.0 % (Saeed et al., 2022), and the comparison of JI Panel with culture on SA and acute PJI demonstrated a sensitivity of 80.6 % and a specificity of 100.0 % for the JI Panel (Schoenmakers et al., 2023).

The purpose of this study was to evaluate the agreement between the JI Panel and the synovial fluid culture and assess the user experience with the JI Panel in a multi-national site study of real-world settings.

A total of 30 to 60 JI Panels were provided to 34 clinical sites across 19 countries in Europe and the Middle East. Instructions were provided to utilise the JI Panel per labelled indication in addition to the standard of care in patients suspected of having SA or PJI using leftover synovial fluids. The JI Panel includes 29 bacterial targets, 2 yeasts, and 8 resistance markers (details of the microorganisms are in Tables 2 and 3). The JI Panel tests were intended for investigational use only, and no data provided by the JI Panel were used for clinical decision-making. To capture the real-world data with the least amount of bias, the study allows the standards of care to vary (e.g. disease epidemiology, patient management methods, and facility capacities) across clinical sites. Data on the JI Panel results and their synovial fluid culture counterparts performed per standard of care in each centre were collected prospectively by individual investigators in each centre during approximately 16 months, from February 2021 to June 2022. Subsequently, the sites were asked to comment on whether the JI Panel results would have had an impact on patient management, in order to evaluate the user experience and possible perspectives in utilising the JI Panel in real-life use.

Only descriptive analyses were done for the data interpretation. Overall agreement was calculated at sample level, based on results obtained with synovial fluid routine culture and JI Panel for each sample; a positive result means that at least one microorganism was detected by the method. Then, the detection results by organism were assessed on if both or only one of the methods detected the microorganism. The final diagnosis of SA or PJI was not considered in this calculation. Ethical aspects, such as required approval and patient consent forms, were handled on a case-per-case basis depending on the site and country requirements. Only suspected cases of SA and PJI were included. In addition, all specimens in which cultures only grew organisms missing from the JI Panel menu were excluded for the on-panel analysis results.

There were 1527 synovial fluid analysed across the 34 participating sites, including 873 (57 %) specimens for SA and 398 (26 %) in PJI. The sites reported use of the JI Panel most commonly in patients over the age of 56 years. The distribution of joint types was the following: knee (54 %), hip (16 %), shoulder (5 %), elbow (4 %), spine (2 %), ankle (2 %), and wrist (1 %).

Table 1 presents the patient age groups, joint infection types, and involved joints.

Delay in diagnosis and inadequate antibiotic therapy in septic joints including PJI can lead to worse outcomes, and therefore enhancing microbiological diagnostic yield is an important factor in the management and outcome of septic arthritis (Costales and Butler-Wu, 2018).

We report the findings for a multi-national evaluation of the JI Panel in comparison with SFC for synovial fluid (native and PJI) samples to consider its potential impact on the management of joint infections in various laboratory settings.

At specimen level, the overall agreement between the two methods was high for both SA (88.4 %) and PJI (85.7 %), with more specimens found to be positive by the JI Panel than synovial fluid cultures. SFC is the current reference method but seems to fail to identify some microorganisms leading to a lower diagnostic yield than the JI Panel. Even if there were no discrepant analyses at the specimen level, these results are in accordance with recent publications, demonstrating that the positivity rate obtained by SFC is lower than the one obtained with the JI Panel (Saeed et al., 2022; Esteban et al., 2023).

The most frequent microorganisms identified in both types of infections were Staphylococcus aureus, Streptococcus, and Enterococcus species. The JI Panel found many additional detections of these three microorganisms, which could potentially lead to more appropriate antimicrobial treatment, a crucial element in tackling antimicrobial resistance, and better patient outcome. Additionally, the panel detected more anaerobic organisms which can be generally difficult to grow, or they can easily perish during suboptimal sample processing. More than 30 anaerobic microorganisms such as Anaerococcus prevotii/vaginalis, Finegoldia magna, Peptoniphilus, and Parvimonas micra were only detected by the JI Panel and not by SFC. While these microorganisms are less prevalent based on culture epidemiology, their detection by the JI Panel will potentially change the nature of the epidemiology of SA and PJI and possible antibiotics guidance; however we need more studies to confirm this impact. The JI Panel also identified some microorganisms that were overlooked by SFC, such as Neisseria gonorrhoeae and Streptococcus pneumoniae, which might not routinely be suspected by clinicians.

Early diagnosis of Neisseria gonorrhoeae septic arthritis is not only important for the patient but also for contact tracing as well checking other potential sexually transmitted diseases.

Regarding paediatric patients, Kingella kingae is the most prevalent pathogen for children under 5 years old. The JI Panel was able to detect three cases of Kingella kingae in paediatric SA samples that were missed by SFC as they are difficult to grow organisms.

Confirming the findings from Saeed et al. (2022), accurate identification in children with septic arthritis is critical to avoid cartilage damage and could also support an adjusted antimicrobial prescription with an earlier switch of route of administration and facilitate discharge (Alcobendas et al., 2023).

Additionally, it is important to note that the JI Panel's ability to quickly detect certain common antimicrobial resistance genes may optimise antimicrobial therapy which will impact antimicrobial stewardship, especially in regions with a high prevalence of drug resistant pathogens, as well as impact infection prevention, isolation for cases of methicillin-resistant S. aureus (MRSA), extended spectrum beta-lactamase producers (ESBL), and vancomycin-resistant enterococci (VRE). The JI Panel detected more resistance genes overall, especially for Staphylococcus aureus with methicillin resistance genes.

The study has limitations, including its retrospective nature; interlaboratory variations of SFC; and sites performing the study at different time points, with criteria for native SA or PJI that could have potentially varied from site to site based on suspicion of infection. Furthermore, as a non-interventional study, we have no prior information about antibiotic exposure and the impact of this on culture. Finally, we did not assess impact on antimicrobial management and infection prevention when a resistance gene was detected by the panel. Additionally, it is important to mention that a certain proportion of organisms identified by SFC were not detected by the JI Panel as the panel does not cover those specific organisms and not because of any limitations in the accuracy of detection, such as coagulase negative Staphylococci (apart from S. lugdunensis) and Cutibacterium acnes.

More studies such as controlled or interventional studies are needed to better evaluate the clinical impact in terms of antimicrobial resistance, surgical management, and long-term patient-reported outcomes.

Despite these limitations, the JI Panel detected more microorganisms and more resistance genes in both SA and PJI compared to synovial fluid cultures and applicable methods for resistance identification. Even if the study did not directly compare the turnaround times of SFC and the JI Panel, as per recent publications, the JI Panel, with a 1 h turnaround time, provided a faster time to results compared to a Gram stain and SFC, which can take up to 4 h and 14 d respectively (Azad et al., 2022; Berinson et al., 2023; Hoffman et al., 2023). Given the high precision of diagnosis with the JI Panel relative to SFC and the ability to quickly detect microorganisms and resistance markers, in 1 h, the JI Panel may yield a higher number of correctly identified pathogens in both septic arthritis in native joints and prosthetic joint infections, possibly leading to an optimised antimicrobial stewardship, as delayed appropriate therapy could increase the time of antibiotic therapy (Balada-Llasat et al., 2022). Hence, patients could also benefit from an earlier antibiotic de-escalation and/or switch of antibiotic (Berinson et al., 2023). This could also support the surgical strategy (Indelli et al., 2023) and appropriate patient management for better patient outcomes.

The JI Panel must always be used alongside synovial fluid culture for patients with a suspicion of SA or PJI, and the final diagnosis should follow dedicated guidelines. Preferences should not be given to a specific joint since there is no significant difference.

In this real-world evidence study, the JI Panel was provided to clinical sites across several countries to further understand its utility as assessed by the final users. The impact on patient management observed on positive samples is correlated with results obtained when culture as having a fast and accurate result could improve patient management. User experience needs to be further evaluated through surveys and testimonials to gather more informative data on its clinical utility.

In summary, the JI Panel had increased yield for on-panel organisms compared to synovial fluid cultures and with rapid turnaround times, demonstrating a potential adjuvant to standard cultures with more clinical impact on patient management.