Performance of Direct Identification and EUCAST Rapid Antimicrobial Susceptibility Testing in Bloodstream Infections Using Sepsityper® and Disk Diffusion Methods

Background/Objectives: Rapid pathogen identification and targeted antimicrobial susceptibility testing (AST) in bloodstream infections are critical to reducing sepsis-related mortality. This study aimed to evaluate the performance of direct MALDI-TOF MS identification using the rapid Sepsityper® kit and to determine the diagnostic accuracy and reliability of the EUCAST Rapid Antimicrobial Susceptibility Testing (RAST) protocol at various incubation periods (4, 6, 8, and 16–20 h) from positive blood culture bottles. Methods: A total of 71 clinical isolates recovered from positive blood cultures in the Medical Microbiology Laboratory of Ondokuz Mayıs University Faculty of Medicine were included. Samples prepared directly using the rapid Sepsityper® kit were identified via MALDI-TOF MS, while the EUCAST RAST protocol was simultaneously performed on Mueller–Hinton agar using disk panels guided by Gram stain results. Results were analyzed for Categorical Agreement (CA), major errors, and very major errors using routine automated AST data as the reference standard. Results: The rapid Sepsityper® kit successfully identified 51 of 71 isolates (71.8%) at the species level within 30–60 min. The overall categorical agreement of EUCAST RAST with the automated system increased from 67.5% at 4 h to 76.6% at 6 h, 82.7% at 8 h, and 87.0% at 16–20 h. With prolonged incubation, the major error rate dropped significantly from 51.3% to 3.8%. However, the very major error rate remained at 8.1% in final readings, with these errors predominantly clustering in carbapenems and fluoroquinolones among Klebsiella pneumoniae strains. Conclusions: The rapid Sepsityper® kit provides species-level identification in under an hour in approximately three-quarters of cases, facilitating the early de-escalation of empirical therapy. Although EUCAST RAST categorical agreement improves progressively with time, conventional methods are still recommended for confirmation due to the risk of very major errors, particularly in detecting carbapenem resistance.