More specifically, they will have to acknowledge that children who develop a UTI can be divided into 2 groups. The first, which comprises most children and the subjects of recent clinical trials, are those with low-grade VUR (grades I–III) and no or minimal kidney scarring at the time of UTI diagnosis. The evidence suggests that only 5% to 30% of these children go on to develop a second UTI, a vanishingly small percentage of them develop renal scarring of any clinical significance, and CAP has little, if any, effect on their clinical outcomes. For this reason, the recently revised AAP guideline recommended that (1) the initial work-up of children with first febrile UTI should consist of a renal ultrasound only, which should detect most high-grade VUR and significant renal scarring or anatomic abnormalities of the genitourinary tract; and (2) CAP can be deferred in children with normal renal ultrasound. With this change in recommendations around imaging of young children with first febrile UTI, the large proportion of children who are likely to have a benign clinical course will be spared an invasive voiding cystourethrogram (VCUG), years of CAP, and possibly surgery to correct VUR.

The second group of children, which represents the minority with first UTI, consists of those with high-grade VUR (grades IV–V), more extensive kidney scarring at baseline, and a predisposition to multiple breakthrough UTIs. These children are at the highest risk of suffering clinically significant renal injury, but, because there are so few of them, they are not well represented in recent clinical trials and thus there is considerable uncertainty about what, if any, therapy can effectively protect them from long-term renal insufficiency. They are also the ones most likely to find their way to the offices of pediatric urologists, and so they feature prominently in the pediatric urologic literature, including the review article by Fonseca and colleagues elsewhere in this issue. Given the problem of small numbers, and the reluctance of most physicians and parents to agree to randomization of these high-risk children, it will be difficult to conduct clinical trials to define best practices for them. Some of these children have genetically determined developmental abnormalities of the kidneys and urinary tract, which manifest after birth as VUR and renal dysplasia/hypoplasia. In these children, the UTIs are an epiphenomenon rather than the cause of the renal abnormalities, and there may be little that can be done to prevent the development of renal insufficiency. For others, surgical correction of VUR may remove an important risk factor for recurrent UTI and prevent subsequent kidney injury.

The greatest contribution of the review by Fonseca and colleagues is that it articulates the idea that the risk of recurrent UTI and subsequent renal insufficiency varies considerably among children with first UTI, and that the intensity (and potential harms) of any intervention must be matched to that risk. However, their review omits 2 important considerations: first, that most children with first UTI (and nearly all the children that a primary care pediatrician encounters) are in the low-risk group; and, second, that there is little to no high-quality evidence to support their recommendations for the small number of children in the moderate-risk and high-risk groups. Nonetheless, decisions must be made about the clinical management of these children and a standardized risk-specific treatment approach such as the one the authors outline, can provide clinicians with an opportunity to standardize care and measure and continuously improve outcomes for these children. Given the lack of feasibility to conduct randomized controlled trials in this patient population, the best hope for progress in understanding how to manage them is for pediatric urologists to establish a national registry of high-risk children, documenting their clinical characteristics, treatments, and outcomes in a thorough and consistent manner. Such registries have been successful in facilitating discussion, evaluating practices, and improving outcomes for other uncommon conditions, such as cystic fibrosis, inflammatory bowel disease, and congenital heart disease.

Although comparisons of treatments for high-risk children are difficult outside of national registries, there is still much to be learned about the factors that contribute to renal injury in these children. Clinicians and researchers must begin to think outside of the proverbial VUR box. It is beginning to be understood that there are a variety of abnormalities in host defenses that might predispose some children to recurrent UTI. Knowledge of these deficiencies in specific host defenses may lead to therapies designed to compensate for them. There is also much left to be learned about host inflammatory response to kidney infection to explain why some children suffer extensive kidney injury with pyelonephritis, whereas others with the same amount of acute inflammation avoid scarring altogether. The National Institute of Diabetes and Digestive and Kidney Diseases recently funded a trial to determine whether steroids can modify the inflammatory and scarring response to acute pyelonephritis ( clintrials.gov identifier NCT01391793). If steroids are found to be effective, future research should explore the ability of more targeted and powerful biologic agents to control the inflammation and scarring associated with kidney infection. In addition, there is still hope that advances in genetics may help to accurately stratify children according to their risk of recurrent UTI and kidney injury, even before they have their first UTI, thus allowing earlier tailoring of their therapy to their risk status.