Key points are not available for this paper at this time.
BACKGROUND: Efficient urinary tract infection (UTI) diagnostics require reliable methods to rapidly exclude negative samples while accurately identifying positive cases. Automated urine flow cytometry (UFC) offers a promising approach to streamline laboratory workflows by reducing unnecessary cultures and providing early bacterial classification. This study evaluated the performance of a UFC-based screening method compared to conventional urine culture. METHODS: 4005 urine samples were analysed using UFC (Sysmex UF-5000) and compared to culture results. Receiver operating characteristic (ROC) curve analyses assessed method agreement across different patient subpopulations. Predictive values were calculated for the total population and the ten different subpopulations at different cut-offs. Carryover and cross-contamination were also examined. RESULTS: Based on our data, we propose an algorithm using bacterial count from UFC to guide urine culture decisions. A cut-off of 4000 BACT cells/µl was suggested to indicate clinically relevant bacteriuria (AUC 0.917). Pregnant women were excluded from this rule-in approach due to limited discriminatory performance in this subgroup. Additionally, the BACT-Info flag "Gram Neg?" predicted Gram-negative bacteria in 61% of culture-positive samples, with 96% concordance, offering potential early indication of Gram-negative infections. CONCLUSIONS: UFC shows potential as a useful screening tool for reducing unnecessary cultures by helping to exclude negative samples and highlighting those likely to be positive. The "Gram Neg?" flag facilitated early differentiation of Gram-negative infections, aiding timely targeted antibiotic therapy. Thus, implementing UFC-based screening in routine diagnostics could reduce laboratory workload while maintaining diagnostic accuracy and patient safety.
Sender et al. (Wed,) studied this question.