2152. Detection of Uropathogens Using BD Kiestra™ Total Laboratory Automation with Urine Culture Application

Abstract Background Urine is the most frequently cultured specimen type for the majority of clinical microbiology laboratories. Typically, around 30% of cultures are positive for uropathogens with 70% yielding insignificant or mixed growth. BD is developing a software Urine Culture Application (UCA) for the BD Kiestra Total Laboratory Automation (TLA) system to screen images of urine culture plates, sort them based on growth vs. insignificant growth and also allow for presumptive pathogen identification. Methods De-identified urine specimens were inoculated onto BD BBL™ CHROMagar™ Orientation Media (CHROM; BD, Sparks, MD), CHROM/Trypticase™ Soy Agar II with 5% Sheep Blood (TSA) biplate, BD BBL MacConkey II agar, and TSA using the BD Kiestra TLA system. Plates were imaged at 24 hours using the BD Kiestra™ ReadA Compact imaging acquisition software and an algorithm was applied to the images using the UCA (Version 2.0). Semi-quantitative measurements of <100, 100–1,000, 1,000–10,000, 10,000–100,000, and >100,000 cfu/mL growth were determined by UCA for all media types and presumptive ID was determined using CHROM. Manual reading of the images by two technologists was the gold standard for comparison. For discrepant results, a third manual reader was used as an arbitrator. Results Testing between 877 and 934 urine specimens on each of five media types using UCA resulted in an exact semi-quantitative agreement with manual reading for 85.5–95.0% of specimens (Table 1). If semi-quantitative values ± one category of agreement are included, the number rises to 98.2–99.4% agreement. Using CHROM for presumptive identification of pure or predominant organisms, UCA was in agreement with manual identification in 251 of 272 cultures (92.3%). Of the 21 discrepant organisms, 19 were classified as “other” by manual reading but were identified as specific organisms by UCA. Definitive organism identification was not performed. Conclusion UCA was able to accurately categorize bacterial growth into five semi-quantitative categories using five media types. Pure and predominant uropathogens were accurately identified from CHROM using UCA. The use of UCA software application may enable laboratories to save time screening urine cultures by allowing more efficient use of technologist time. Disclosures All authors: No reported disclosures.