Targeted Indication of Imaging for Detection of Vesicoureteric Reflux after Pediatric Febrile Urinary Tract Infections Based on a Multiparametric Computational Tool.

2020 
OBJECTIVES Diagnostic workup after febrile urinary tract infections (fUTIs) in children remains a matter of debate. The authors aimed to evaluate multiple parameters in order to design a predictive tool enabling a targeted indication of voiding cystourethrography (VCUG). METHODS Records of 383 consecutive children who underwent a VCUG as well as a dimercaptosuccinic-acid (DMSA) scan after febrile urinary tract infections (fUTIs) at a single institution between 04/2009 and 06/2014 were reviewed. Twenty parameters were recorded. After regression analysis, 6 parameters were incorporated into a computational tool aiming at a targeted indication of an eventual VCUG. The performance of the tool was prospectively tested on 100 patients. RESULTS Postpyelonephritic alterations on DMSA, duplex systems, age 1-3 y, duration of fever >3 d, >2 fUTIs before VCUG and abnormal sonography findings were identified as significant predictors (p < 0.05 each); the presence of bladder and bowel dysfunction (BBD) was negatively associated with vesicoureteric reflux (VUR). The resulting computational tool achieved an Area under the curve (AUC) of 0.686 (CI 0.633-0.740). Prospective evaluation (100 new patients) revealed a sensitivity of 85.1%, a specificity of 49.1%, a positive predictive value of 59.7% and a negative predictive value of 78.7%. CONCLUSIONS The differentiated indication of a VCUG based on the use of a tool was efficient in optimizing the specificity of the diagnostic algorithm after fUTIs. The tool outperformed other common clinical approaches in terms of VUR detection and VCUG frequency. After validation and further refinement in a multicentric approach, this strategy could significantly enhance VUR detection whilst reducing the number of VCUGs.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    28
    References
    2
    Citations
    NaN
    KQI
    []