@article{mbs:/content/journal/jmm/10.1099/jmm.0.000714, author = "Bzdyl, Nicole M. and Urosevic, Nadezda and Payne, Ben and Brockenshire, Ray and McIntyre, Michael and Leung, Michael J. and Weaire-Buchanan, Graham and Geelhoed, Elizabeth and Inglis, Timothy J. J.", title = "Field trials of blood culture identification FilmArray in regional Australian hospitals", journal= "Journal of Medical Microbiology", year = "2018", volume = "67", number = "5", pages = "669-675", doi = "https://doi.org/10.1099/jmm.0.000714", url = "https://www.microbiologyresearch.org/content/journal/jmm/10.1099/jmm.0.000714", publisher = "Microbiology Society", issn = "1473-5644", type = "Journal Article", keywords = "sepsis", keywords = "blood culture", keywords = "polymerase chain reaction", keywords = "FilmArray", keywords = "molecular diagnostics", keywords = "bacteraemia", keywords = "bacterial identification", abstract = " Purpose. In this field trial of rapid blood culture identification (BCID), we aimed to determine whether the improved speed and accuracy of specific BCID predicted in our earlier pilot study could be obtained in regional hospitals by deploying a multiplex PCR FilmArray (Biomerieux, France) capability in their laboratories. Methods. We trained local hospital laboratory staff to operate the FilmArray equipment and act on the results. To do this, we integrated the multiplex PCR into the standard laboratory blood culture workflow and reporting procedure. Results. Of 100 positive blood culture episodes, BCID FilmArray results were correct in all 42 significant monobacterial cultures, with a fully predictive identity in 38 (90.5 %) and a partial identity in another four (9.5 %). There was one major error; a false positive Pseudomonas aeruginosa. The minor errors were the detection of one methicillin-resistant Staphylococcus aureus, which proved to be a methicillin-sensitive S. aureus mixed with a methicillin-resistant coagulase-negative staphylococcus, five false negative coagulase-negative staphylococci and one false negative streptococcus species. We found that 41/49 (84 %) clinically significant mono- and polymicrobial culture results were fully predictive of culture-based identification to bacterial species level at a mean of 1.15 days after specimen collection. Conclusions. There was a reduction of 1.21 days in the time taken to produce a definitive BCID compared to the previous year, translating into earlier communication of more specific blood culture results to the treating physician. Reduced time to definitive blood culture results has a direct benefit for isolated Australian communities at great distances from specialist hospital services.", }