Background: Urinary tract infections (UTI) are among the most common infections in children. The primary tool to detect UTI is dipstick urinalysis; however, this has limited sensitivity and specificity. Therefore, urine culture has to be performed to confirm a UTI. Urinary volatile organic compounds (VOC) may serve as potential biomarker for diagnosing UTI. Previous studies on urinary VOCs focused on detection of UTI in a general population; therefore, this proof-of-principle study was set up in a clinical high-risk pediatric population.
Methods: This study was performed at a tertiary nephro-urological clinic. Patients included were 0-18 years, clinically suspected of a UTI, and had abnormal urinalysis. Urine samples were divided into four groups, i.e., urine without bacterial growth, contamination, colonization, and UTI. VOC analysis was performed using an electronic nose (eNose) (Cyranose 320®) and VOC profiles of subgroups were compared.
Results: Urinary VOC analysis discriminated between UTI and non-UTI samples (AUC 0.70; p = 0.048; sensitivity 0.67, specificity 0.70). The diagnostic accuracy of VOCs improved when comparing urine without bacterial growth versus with UTI (AUC 0.80; p = 0.009, sensitivity 0.79, specificity 0.75).
Conclusions: In an intention-to-diagnose high-risk pediatric population, UTI could be discriminated from non-UTI by VOC profiling, using an eNose. Since eNose can be used as bed-side test, these results suggest that urinary VOC analysis may serve as an adjuvant in the diagnostic work-up of UTI in children.
Keywords: bacterial growth culture; electronic nose; urinary tract infection; volatile organic compounds.