Laboratory automation reduces time to report of positive blood cultures and improves management of patients with bloodstream infection

• DOI: 10.1007/s10096-018-3377-5
• Publication Date: 2018-09-14T12:39:08Z
• Authors: Giuseppe Vittorio...; Francesco Di Donato; Riccardo Paggi; Chiara Gabrielli; Alessandra Belati; Giuseppe Rizza; Martina Savoia; Antonella Repetto; Elio Cenci; Antonella Mencacci
• Keywords: Automation, Laboratory; 0303 health sciences; Time Factors; Bacteria; Bacteremia; Microbial Sensitivity Tests; Bacterial Physiological Phenomena; Anti-Bacterial Agents; Bacterial Typing Techniques; 3. Good health; 03 medical and health sciences; BD Kiestra; Blood culture; Bloodstream infections; Laboratory automation; Time to report; Early Diagnosis; Blood Culture; Humans

The impact on time to results (TTR) and clinical decisions was evaluated for mono-microbial positive blood cultures (BC) processed using the BD Kiestra Work Cell Automation (WCA) system. Positive BC were processed by the WCA system by full-automatic subculture on solid media and digital imaging after 8 h of incubation (8-h method) followed by identification (ID) and antimicrobial susceptibility testing (AST). To evaluate the accuracy of the 8-h method, ID and AST from 8-h and overnight incubated colonies were compared for the same organisms. To evaluate its clinical impact, results from 102 BC processed by the 8-h method (cases) were compared with those from 100 BC processed by overnight incubation method (controls) in a comparable period. Identification after 8-h and overnight incubation gave concordant results in 101/102 (99.0%) isolates. Among a total of 1379 microorganism-antimicrobial combinations, categorical agreement was 99.4% (1371/1379); no very major error, 7 major errors, and one minor error were observed. TTR in cases (32.8 h ± 8.3 h) was significantly (p < 0.001) shorter than in controls (55.4 h ± 13.3 h). A significant reduction was observed for duration of empirical therapy (cases 54.8 h ± 23.3 h vs controls 86.9 h ± 34.1 h, p < 0.001) and 30-day crude mortality rate (cases 16.7% vs controls 29.0%, p < 0.037). Automation and 8-h digital reading of plates from positive BC, followed by ID and AST, greatly reduce TTR and shorten the duration of antimicrobial empiric therapy, possibly improving outcome in patients with mono-microbial bloodstream infections.