Rapid identification of organisms in positive blood cultures allows the delivery of appropriate antibiotic treatment earlier, resulting in better patient outcomes and reduced length of stay in hospital. In this study we assessed the reliability of the Bruker MALDI Sepsityper™ kit to give an accurate identification. We also established an algorithm for accepting the identification based on the log score and gram stain.
326 positive blood cultures were tested (BD Bactec™). No duplicates or mixed cultures were included. All blood cultures were sub cultured onto solid media and gram stains performed. All procedures and identifications using the Sepsityper kit™ and Bruker MALDI Biotyper™ were performed to the manufacturer’s specifications.
Of the 326 bottles tested, 265 gave a reliable identification to the species with a log score greater than 1.7 (81%). 263 of those were proven to be the correct identification by culture (99%).
61 bottles gave an unreliable identification due to a log score less than 1.7. Of those 61, 27 matched the gram stain and were confirmed as correct by culture. Of the 34 incorrect identifications, 30 didn’t match the preliminary gram stain and were therefore recognised as being misidentified.
By using an algorithm where the identification obtained by the Sepsityper kit™ must match the gram stain to be considered correct, regardless of the log score, we increased the number of identifications obtained from 265 (81%) to 296 (91%). 290 of those were proven to be correct by culture (98%).
In conclusion, using a log score value of 1.7 is a reliable way to ensure the identifications obtained by the Sepsityper kit™ are correct. Using the gram stain as a filter for the acceptance of identifications with a score less than 1.7 increases the number of reliable identifications obtained.