Objectives: (1) To determine the degree of agreement in the classification of susceptibility, intermediate, and resistance between the direct antimicrobial susceptibility testing method from positive blood samples and the routine antibiotic susceptibility testing method from colonies using the BD Phoenix M50 system; (2) To determine the rate of very major errors and major errors between two methods.
Subjects and methods: 195 blood samples from January 2023 to December 2025 at Vinmec Times City International General Hospital were automatically incubated using the BacT/Alert 3D 120 system, showing positive results for the first isolated, preliminary staining for Gram-negative bacilli, and bacterial identification after subculture belonging to Enterobacterales and P.aeruginosa. The samples were subjected to direct antimicrobial susceptibility testing based on CLSI M100, and blood agar was subcultured for the identification and routine antibiotic susceptibility testing steps.
Results: The agreement on classification between the direct antibiotic susceptibility testing method from blood bottles and the routine antibiotic susceptibility testing method for Enterobacterales and P.aeruginosa strains of 95%, and 100%, respectively. Most antibiotic-bacteria combinations showed very major error rates and major error rates below 3%.