The increasing prevalence of multi-drug-resistant bacteria responsible for bloodstream infections (BSIs) makes therapeutic choices progressively more complex. Fast microbiology quickly detects the presence of pathogens and clinically relevant determinants of antibiotic resistance, offering the potential for early administration of antibiotics. In this retrospective observational study, we comparatively evaluated the performances of FilmArray and the current standard method using blood samples collected from intensive care unit (ICU) patients with suspected BSI. A full agreement with the standard was observed in 97/102 samples (95.1 ± 4.2%), a mismatch in 3/102 samples (2.9 ± 3.2%) and detection failure in 2/102 cases (1.96 ± 2.7%). Statistical analysis demonstrated a near-perfect/perfect level of agreement between the two methods, with an overall degree of agreement of 95%. The high performance demonstrated by the FilmArray could allow a "watch and wait" approach helping clinicians in decision-making processes related to choice and initiation of the antimicrobial therapy, thus avoiding ineffective and excessive use of drugs.
© 2022. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.