Timing of follow-up voiding cystourethrogram in children with primary vesicoureteral reflux: development and application of a clinical algorithm.

BACKGROUND AND OBJECTIVES: Of children diagnosed with urinary tract infection, 30% to 40% have primary vesicoureteral reflux (VUR). For the majority of these children, treatment involves long-term prophylactic antibiotics (ABX) and a periodic voiding cystourethrogram (VCUG) until resolution of VUR as detected by VCUG. Radiation exposure and considerable discomfort have been associated with VCUG. To date, no clear guidelines exist regarding the timing of follow-up VCUGs. The objective of this study was to develop a clinically applicable algorithm for the optimal timing of repeat VCUGs and validate this algorithm in a retrospective cohort of children with VUR.

METHODS: Based on previously published data regarding the probability of resolution of VUR over time, a decision-tree model (DTM) was developed. The DTM compared the differential impact of 3 timing schedules of VCUGs (yearly, every 2 years, and every 3 years) on the average numbers of VCUGs performed, years of ABX exposure, and overall costs. Based on the DTM, an algorithm optimizing the timing of VCUG was developed. The algorithm then was validated in a retrospective cohort of patients at an urban pediatric referral center. Data were extracted from the medical records regarding number of VCUGs, time of ABX prophylaxis, and complications associated with either. VUR in patients in the cohort was grouped into mild VUR (grades I and II and unilateral grade III for those < or =2 years old), and moderate/severe VUR (other grade III and grade IV). Kaplan-Meier survival curves were created from the cohort data. From the survival curves, the median times to resolution of VUR were determined for the cohort, and these times were compared with the median times to VUR resolution of the data used for the DTM. The numbers of VCUGs performed, time of ABX exposure, and costs in the cohort were compared with those that would have occurred if the algorithm had been applied to both mild and moderate/severe VUR groups.

RESULTS: Using an algorithm that results in a recommendation of VCUGs every 2 years in mild VUR would reduce the average number of VCUGs by 42% and costs by 33%, with an increase in ABX exposure of 16%, compared with a schedule of yearly VCUGs. For moderate/severe VUR, a VCUG performed every 3 years would reduce the average number of VCUGs by 63% and costs by 51%, with an increase in ABX exposure of 10%. Applying this algorithm to the retrospective cohort consisting of 76 patients (between 1 month and 10 years old) with primary VUR would have reduced overall VCUGs by 19% and costs by 6%, with an increase in ABX exposure of 26%. The patterns of VUR resolution, age distribution, and prevalence of severity of VUR were comparable between previously published results and the retrospective cohort.

CONCLUSIONS: Delaying the schedule of VCUG from yearly to every 2 years in children with mild VUR and every 3 years in children with moderate/severe VUR yields substantial reductions in the average numbers of VCUGs and costs, with a modest subsequent increase in ABX exposure.