Top-Down versus Bottom-Up Approach in Children Presenting with Urinary Tract Infection: Comparative Effectiveness Analysis Using RIVUR and CUTIE Data.

Abstract

The initial imaging approach to children with urinary tract infection (UTI) is controversial. Along with renal/bladder ultrasound, some advocate voiding cystourethrogram (VCUG), ie a bottom-up approach, while others advocate dimercaptosuccinic acid (DMSA) scan, ie a top-down approach. Comparison of these approaches is challenging. In the RIVUR/CUTIE trials, however, all subjects underwent both VCUG and DMSA scan. Our objective was to perform a comparative effectiveness analysis of the bottom-up vs top-down approach. We simulated 1,000 hypothetical sets of 500 children using RIVUR/CUTIE data. In the top-down approach, patients underwent initial DMSA scan, and only those with renal scarring underwent VCUG. In the bottom-up approach, the initial study was VCUG. We assumed all children with vesicoureteral reflux (VUR) received continuous antibiotic prophylaxis (CAP). Outcomes included recurrent UTI, number of VCUGs and CAP exposure. We assumed a 25% VUR prevalence in children with initial UTI with sensitivity analysis using 40% VUR prevalence. Median age of the original RIVUR/CUTIE cohort was 12 months. First DMSA scan was performed at a median of 8.2 weeks (IQR 5-11.8) after the index UTI. In the simulated cohort, slightly higher yet statistically significantly recurrent UTI was associated with the top-down compared with the bottom-up approach (24.4% vs 18.0%, p=0.045). On the other hand, the bottom-up approach resulted in more VCUG (100% vs 2.4%, p <0.001). Top-down resulted in fewer CAP-exposed patients (25% vs 0.4%, p <0.001) and lower overall CAP exposure (5 vs 162 days/person, p <0.001). Sensitivity analysis was performed with 40% VUR prevalence with similar results. The top-down approach was associated with slightly higher recurrent UTI. Compared to the bottom-up approach, it significantly reduced the need for VCUG and CAP.