Back

Accurate, cost-effective, and a rapid turnaround of four hours for complicated UTI diagnosis using metagenomics

Bellankimath, A. B.; Branders, S.; Kegel, I.; Ali, J.; Asadi, F.; Johansen, T. E. B.; Imirzalioglu, C.; Hain, T.; Wagenlehner, F.; Ahmad, R.

2025-04-25 urology
10.1101/2025.04.24.25326310 medRxiv
Show abstract

Urinary tract infections (UTIs) affect 405 million people worldwide. Current diagnostics for UTIs rely on cultures, which can take 2 to 4 days. This study evaluated eleven culture independent methods for sample preparation from 77 complicated UTI patient samples, followed by real-time nanopore sequencing and data analysis. The metagenomic results were highly consistent with culture-based clinical routines. The optimized in-house method demonstrated a mean accuracy score of 99% for pathogen identification and 90% for antimicrobial susceptibility profiling. The methods robustness is highlighted by accurately identifying the pathogen with as few as 33 bacterial cells/{micro}L and a low bacterial-to-leukocyte ratio limit of 0.04. Additionally, mNGS identified 13 pathogens that routine diagnostics missed, which were subsequently confirmed by Vivalytic or PCR. This method is up to 30% cheaper than published studies and commercial kits. AUROC analysis indicated that DNA yield and flow cytometry can be used for pre-screening to reduce costs, which is crucial for clinical adoption. This research highlights the rapid diagnosis of UTIs in clinical settings using a cost-effective and scalable method that requires four hours from sample collection to informed decision making. Furthermore, it aims to improve antimicrobial and diagnostic stewardship by reducing empirical treatment and ensuring more judicious antibiotic use.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.