‍Abstract: Pseudomonas spp., especially P. aeruginosa are among the most important nosocomial pathogens, mainly as causative agents of infections of wounds, respiratory tract infections, and the urinary tract infections. Rapid and reliable antimicrobial susceptibility testing (AST) is mandatory for targeted treatment of P. aeruginosa infections. Since automated AST does not always provide reliable results for some antibiotics, additional confirmatory manual testing of P. aeruginosa is often required.

In this study, 19 P. aeruginosa isolates from routine samples were tested on a panel of 12 antibiotics using standard disc diffusion testing, and zone diameter results from three interpretation methods were compared: manual plate-in-hand reads, APAS Independence AI-algorithms, and electronic reads by a microbiologist using APAS-generated images. The APAS Independence results showed high categorical agreement (c. 95%) with manual reads, with a low error rate.

This study shows that the APAS AMR module has a high categorial agreement with manual reading of disc diffusion test for P. aeruginosa and is therefore a promising tool for the automated measurement of zone diameters in the routine microbiology. APAS AMR can flag a review for some reads contributing to errors and decrease the actual error rate further.

Poster & Oral Presentation: Wisplinghoff Laboratories, Cologne, Germany

‍Conference: ECCMID 2024, Barcelona, Spain

‍Date: April 2024

‍Authors: Popsel, J., Nowag, A., Giglio, S., Hill, R., Wisplinghoff, H., and Jazmati, N.

‍Citation: Popsel, J., Nowag, A., Giglio, S., Hill, R., Wisplinghoff, H., and Jazmati, N. 2024. Comparison of AI-assisted and manual interpretation of disc diffusion test from 12 antibiotics of Pseudomonas aeruginosa. 34th ECCMID. Barcelona.