2154. Rapid Microorganism Identification From Blood and Enrichment Fluid Cultures Using MALDI-TOF Mass Spectrometry Following Abbreviated Incubation on Chocolate Agar Plates

Abstract Background The more rapidly a pathogen can be identified in cases of sepsis, the more swiftly targeted therapy can be commenced. We aimed to evaluate the accuracy and clinical utility of a method for rapid microorganism identification from blood and sterile site fluid enrichment cultures using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) in a high-throughput regional microbiology laboratory. Methods A drop of blood or fluid from cultures flagging positive was inoculated onto chocolate agar and incubated until MALDI-TOF MS analysis 5–6 hours later. Results obtained with conventional overnight incubation were used as controls. The number of isolates reliably and correctly identified, and recommendations for changes to clinical management were recorded. A pilot study was undertaken to demonstrate feasibility using a cut-off score of ≥2.0, following which the method was integrated into the laboratory workflow using a cut-off score of ≥1.7 (identification to genus level) because patient management can often be altered with this degree of accuracy. Results In the pilot study, 154/210 blood cultures (73.3%) were reliably identified (score ≥2.0) and, of these, 151 (98.1%) were correctly identified. The operational study included 144 blood cultures and 14 sterile site fluid enrichment cultures. 121 (76.6%) were reliably identified (score ≥1.7) and, of these, 117 (96.7%) were correctly identified. Four isolates (2.5%) were yeasts, 104 (65.8%) were Gram-positive bacteria and 50 (31.6%) were Gram-negative bacteria. 78 of the 104 bacteria (75.0%) were correctly identified. The median score for Gram-positives was 1.89 (IQR 0.36). For Gram-negatives, 47 of the 50 isolates (94.0%) were correctly identified. The median score for Gram-negatives was 2.29 (IQR 0.14). No yeasts were reliably identified. Recommendations for changes to clinical management were made for seven (3.8%) and nine patients (6.0%) in the pilot study and operational study, respectively. The method was easily integrated into the laboratory workflow and did not require additional staff. Conclusion We build on previous work by providing a simple method for rapid micro-organism identification which allows easy integration into a busy laboratory, and which improves clinical management. Disclosures All authors: No reported disclosures.