Abstract
Urinary tract infections are among the most common reasons for antimicrobial treatment, and early diagnosis could have a significant impact by enabling rapid administration of the adapted antibiotic and preventing complications. The current delay between sample receipt and pathogen identification is about 24 to 48 h, which could be significantly shortened by use of an accurate direct method. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is already used for the identification of pathogens in clinical laboratories and constitutes a promising tool for direct diagnosis. A simple preparation protocol was established for the processing of urine samples prior to MS analysis. MALDI-TOF spectra collected directly from 1,000 infected urine samples were used to create a specific reference database (named Urinf). A prospective study was then carried out to evaluate the Urinf database and compare the results obtained with the standard database provided by Bruker on the Biotyper Real Time Classification software. Seven hundred eighty urine specimens were processed and analyzed according to our method. Among them, almost 90% of 500 infected monobacterial samples could be correctly diagnosed with the Urinf database, compared to 50% using the standard database. The identification of Enterobacteriaceae, Staphylococcus aureus, Staphylococcus saprophyticus, Pseudomonas aeruginosa, Enterococcus faecalis, and Enterococcus faecium was greatly improved but not for Staphylococcus epidermidis The creation of a database adapted to a particular type of clinical sample has great potential to increase both the rate and rapidity of pathogen identification. Sensitivity still remains to be improved for bacterial species that exhibit few specific peaks on mass spectra.