Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

In Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases

OBJECTIVES : Urine sample is the most common material tested in clinical microbiology laboratories. Automated analysis is already performed allowing a decrease in time to obtain results and in laboratory technologist (LT) workload. Interestingly, these automatic systems introduced digital imaging concepts. PhenoMATRIX™ (PHM) is an artificial intelligence software that merges picture algorithms and user rules to provide presumptive results. This study aimed at designing a tailored workflow using PHM, performing its validation, and checking its performance in routine practice.

METHODS : Two data collections including 96 and 135 urines from nephrostomy/ureterostomy and artificial bladder (US), 948 and 1257 urines from catheter (UC), and 3251 and 2027 midstream urines (MSU) were used to compare LT results with those obtained using two versions of PHM. Another 19 US, 102 UC, and 508 MSU were used to monitor performance level three months after routine implementation.

RESULTS : Before and after revisions, agreement between the first version of PHM and LT results were 83% (95CI 74.3-90.2) and 83% (95CI 75.3-90.9) (US), 66.7% (95CI 63.5-69.5) and 71.7% (95CI 68.8-74.4) (UC), 65.4% (95CI 63.8-67.1) and 76% (95CI 74.1-77.1) (MSU). The second version improved results, demonstrating 96.2% (95CI 91.6-98.8) and 97% (95CI 92.6-99.2) (US), 87.5% (95CI 85.5-89.2) and 88.9% (95CI 87.0-90.5) (UC), 91% (95CI 89.7-92.1) and 92% (95CI 91.1-93.4) (MSU) of agreement with LT results, before and after revisions. The reliability of PHM results was confirmed by a routine study demonstrating 92% (95CI 90.0-94.2) of overall agreement.

CONCLUSIONS : PHM showed high performance with > 90% of results in agreement with LT. PHM could help standardise and secure results, prioritise positive plates during analytical workflow, and likely save LT time.

Dauwalder Olivier, Michel Agathe, Eymard Cécile, Santos Kevin, Chanel Laura, Luzzati Anatole, Roy-Azcora Pablo, Sauzon Jean François, Guillaumont Marc, Girardo Pascale, Fuhrmann Christine, Lina Gérard, Laurent Frédéric, Vandenesch François, Sobas Chantal


CHROMID®CPSE, PhenoMATRIX™, WASPLab®artificial intelligence, algorithms, urine