Rapid and accurate identification of bacterial pathogens in bloodstream infections is a critical step in management [1]. Indeed, timely administration of appropriate antibiotics has been well documented as a strong predictor of sepsis survival [2] and increased mortality of bloodstream infections (BSIs) is often related to delayed, insufficient, or inappropriate anti-infective treatment [3], [4], [5].

To decrease the time necessary for initiating appropriate antimicrobial therapy, some laboratories have implemented a continuous workflow for bacterial identifications and antimicrobial susceptibility testing. This continuous workflow allows to obtain earlier appropriate treatment and an increase rate of narrow-spectrum antibiotic prescription [6,7]. However, such systems are expensive and demand a complete and sometimes complicated reorganization of the laboratory at the time of implementation of changing.

Recently, new rapid diagnosis tests and particularly molecular approaches have emerged to provide pathogen identification and to identify the presence of some resistance genes or resistance profiles [8,9]. Multiplex PCR are increasingly used in clinical practice for diagnosing central nervous system infections, particularly meningitis [10] and pneumonia whether they are community- or hospital-acquired [11]. These multiplex PCR have good analytical performance. Concerning BSIs, the new version of the BIOFIRE Blood Culture Ientification (BCID2) Panel was recently evaluated. The panel identifies 33 pathogens and 10 antibiotic resistance genes with a few minutes of hands-on time [12]. The sensitivity and the specificity for detecting in-panel targets were 98.8% and 99.6% for pathogens and 100% for resistance genes [12]. In another study including 180 positive blood cultures, BCID2 Panel results were concordant with the results of identification of bacterial colonies after subcultures on agar media in 88.3% of positive samples [13]. A retrospective simulation study of time to effective, optimal and de-escalation antimicrobial therapy with BCID2 Panel demonstrated its usefulness for shortening the time of optimal and de-escalation therapy [14]. However, few data concerning the evaluation of the impact of BCID2 Panel on the decrease of time of accurate antimicrobial treatment in the real life are available.

Our objective was to assess the effectiveness of BCID2 panel for the early implementation of appropriate antimicrobial therapy.