Purpose: Urinary tract infections are the most common type of hospital-acquired infection, up to 80% of which are associated with catheterisation. The present study evaluates phenotypic and genomic characterisation of a panel of catheter associated urinary tract infection isolates from a UK hospital.  
Methods: Strains were identified and characterised utilising a range of phenotypic and genomic techniques to understand where methodologies agree. The effect of medium composition on growth and biofilm formation phenotype was also determined to evidence the importance of assay design in characterisation of bacterial isolates.
Results: No consensus was observed for any of the CAUTI isolates across five identification methods, including biochemical testing, MALDI and sequencing technologies. Comparison of EUCAST antimicrobial susceptibility testing and genotypic data for antibiotic resistance showed high concordance where strains were phenotypically resistant to multiple antibiotic classes, however discordance increased for strains that were phenotypically sensitive to range of antibiotics. Phenotypic analysis of bacterial pathogens often relies on the use of rich laboratory media; however, we observed significant differences in growth rate and biofilm formation within a range of media, with a trend towards comparatively low planktonic growth and significant biofilm biomass formation in artificial urine.
Conclusion: This study emphasises potential pitfalls of relying on a single method of species identification, with only whole genome sequencing providing accurate identification of isolates to species level. Furthermore, it highlights the continuing importance of utilising phenotypic methods to understand antibiotic resistance within clinical settings and of utilising clinically relevant conditions for pathogen characterisation.

Evaluation of phenotypic and genotypic methods for the identification and characterisation of bacterial isolates recovered from catheter-associated urinary tract infections