Utility of digital images captured after 4 h of incubation on a microbiology laboratory automation system in guiding the work-up of subcultures from positive blood cultures

Abstract

Initial workup [e.g., identification (ID) and/or antimicrobial susceptibility testing (AST)] of bacterial growth on solid media traditionally occurs 16–24 h after sub-culturing of positive blood cultures (BC). Early ID and AST can be facilitated by reviewing digital images captured using a Microbiology Laboratory Automation (MLA) system. The goal of this study was to evaluate the utility of images captured at 4 h on the WASPLab MLA system for rapid bacterial ID and AST. Retrospective review of all positive BCs results between 1 January 2021 and 31 July 2022 was performed. WASPLab App data were extracted to determine the decision (e.g., perform ID and/or AST or re-incubate plates) made from the 4 h image. Culture results were extracted from the laboratory information system (LIS). A total of 6,845 BCs flagged positive during the study period. The 4 h images for 1,476 cultures (21.6%) were reviewed: 1,200 cultures were re-incubated due to insufficient growth and 276 cultures (4.0%) were sent for ID and/or AST. ID by mass spectrometry was in 100% agreement with that of the molecular BC identification panels. Overall categorical agreement between AST results from the 4 h and overnight growth was 98%. The 4 h images for the remaining 5,369 cultures (78.4%) were not available for review during the day shift. Implementing early reading times for BCs on MLA allows for rapid and accurate ID and AST results. However, optimization of the reading schedule to align with the laboratory’s operation schedule is key to realizing the full potential of early reading times. IMPORTANCE: In recent years, an increasing number of clinical microbiology laboratories have adopted laboratory automation for processing and incubation of specimens submitted for bacterial culture. At our institution, we implemented the Copan WASPLab in 2018 for all cultures, including positive blood cultures. Given that positive blood cultures start with a higher biomass of organisms, the first image capture was set up to occur after 4 h of incubation. In this study, we investigated the utility of this early 4 h image by capturing and calculating the percentage of useful actions taken based on growth identified on the image and the yield of both new identification by MALDI-TOF MS and valid and accurate antimicrobial susceptibility testing (AST) results. We found that while the 4-hour time point provided accurate, early identification and AST results, the overall yield was minimal. From a practical standpoint, this review prompted us to discontinue capture and review of this time point. While our staffing model is likely responsible for this low yield, we hope that our experience would help other laboratories decide how to implement WASPLab workflow for positive blood cultures. Thus, we believe that this information will be of interest to the readers of JCM.