A novel approach to evaluating the benefit of post-urinary tract infection renal ultrasonography, using decision curve analysis.

ABSTRACT

BACKGROUND: The benefit of post-urinary tract infection (UTI) sonography to detect clinically significant renal abnormalities remains a subject open to debate. Decision curve analysis (DCA) is a novel method for evaluating the clinical usefulness of diagnostic tests. Our objective was to determine, using DCA, the benefit of post-UTI sonography and of post-UTI sonography with biological markers of inflammation to predict the risk of recurrence of febrile UTI in children aged 2 to 24 months without known uropathy. METHODS: We retrospectively analyzed all children aged 2 to 24 months, without known uropathy, who presented with a first episode of febrile UTI between 2009 and 2012 and followed them for 30 months. We then used DCA to estimate the benefit of post-UTI sonography or post-UTI sonography + biological markers of inflammation for detecting the risk of recurrence. RESULTS: A total of 318 children [144 boys (45.3 %) and 174 girls (54.7 %)], with a mean age of 6.9 ± 5.6 months, were identified. Of these, 210 children presented with a significant inflammation [66.2 %; 95 % confidence interval (CI) 61.0-71.4], and 30 (9.4 %; 95 % CI 6.2-12.6) presented with abnormal post-UTI sonographic findings. Eighteen (5.7 %; 95 % CI 3.1-8.2) children presented with recurrent UTI at 30 months. CONCLUSIONS: There were significantly more recurrences in those children who presented with abnormal sonographic findings than in those who did not (relative risk 7.68; 95 % CI 3.03-19.46). However, taking into account the effect of false-positives and false negatives, the DCA revealed that for threshold probabilities of >30 %, at which patients/doctors are concerned about unnecessary interventions (whether tests or treatments), neither post-UTI sonography nor post-UTI sonography + biological markers of inflammation have sufficient value to improve care.