Many guidelines published on the management of UTIs in children have narrowed the indications for VCUG, emphasizing that it should be performed only in selected patients [7–10]. Numerous studies have shown that a history of UTI in children is associated with high-grade VUR [11–14]. Contrary to the literature, our study did not find an association between a history of UTIs or the number of UTIs and the frequency of severe VUR. However, the frequency of severe VUR was higher in patients with non-E. coli uropathogens (Tables 1 and 2). Similarly, a study examining 738 patients aged 0–24 months with a history of UTIs found a higher rate of VUR in those with non-E. coli uropathogens compared to those with E. coli [15]. Another retrospective study reported that the presence of non-E. coli uropathogens in cultures was the only significant predictor of severe VUR [13]. Therefore, it is more important to perform VCUG in patients with non-E. coli UTIs rather than in all patients with a history of UTIs. This approach is consistent with many current guidelines [9, 16, 17].
The selection of imaging tests following UTIs in children remains a contentious issue. Many clinical guidelines recommend performing USG for all patients after a UTI [9, 18, 19]. However, USG results are normal in over 80% of cases, leading to significant time and cost inefficiencies [20–22]. Despite the limited sensitivity of USG in detecting VUR, most clinical guidelines recommend VCUG for abnormalities detected on USG [21, 23]. Nonetheless, numerous studies have shown that USG alone is insufficient for diagnosing VUR [11, 24–29]. In our study, regression analysis revealed that the presence of any abnormality on USG was not associated with the frequency of severe VUR. This indicates that USG has a low predictive ability for severe reflux. However, the presence of HN, particularly UTD-P3 HN, on USG was associated with an increased frequency of severe VUR (Tables 1 and 2).
The approach of performing DMSA imaging first and then VCUG after a febrile UTI is known as the "top-down approach." The top-down approach assumes that clinically significant VUR will have associated changes on DMSA scan and that VUR is clinically insignificant in patients with normal DMSA scans [30]. In a prospective study evaluating the top-down approach, 85% of patients with abnormal DMSA had high-grade VUR [31]. A Cochrane review found that children with negative DMSA scans had less than 1% chance of having high-grade VUR [32]. Similarly, various studies have shown that children with severe VUR are more likely to develop renal scarring [25, 33]. In our study, the presence of multiple scars and decreased function on DMSA was associated with an increased frequency of severe reflux (Tables 1 and 2).
In a study using data from 324 children aged 2–60 months with a history of febrile UTI, a scoring system was developed, and independent risk factors for high-grade VUR were identified as recurrent UTIs, non-E. coli pathogens, and abnormal USG findings [34]. Similarly, in our model, the presence of non-E. coli uropathogens was a risk factor for severe VUR. However, instead of any abnormal USG findings, the presence of HN and specifically UTD-P3 HN were found to be significant. Additionally, multiple scars and decreased function on DMSA were also independent risk factors for severe VUR (Table 2).
In our model, the total score ranges from 0 to 6. The cut-off scores were determined as follows: low risk 0–1, moderate risk 2–3, high risk 4–6. Given that a score ≥ 4 has very high specificity for severe VUR, it may be considered to avoid performing VCUG in patients with lower scores.
Guidelines established by major institutions such as the American Academy of Pediatrics and the National Institute for Healthcare and Excellence have significantly reduced the number of unnecessary VCUGs [7, 35]. However, these guidelines have also led to delays in the diagnosis and treatment of some children with VUR. Various studies reviewing these guidelines have shown that most of the missed cases involve children with low grade VUR. Even if children with low grade VUR are not diagnosed after the first UTI, the impact on patients is minimal because antimicrobial prophylaxis is not routinely recommended for these children [16, 17, 36, 37]. The key is not to diagnose every case of VUR but to identify and treat other risk factors for UTIs, such as bladder and bowel dysfunction [38, 39].
According to our prediction model for severe VUR, patients in the low risk group (score 0–1) constituted 61% of all patients, and the rate of severe VUR in this group was 0.3%. Therefore, VCUG should not be performed in patients within the low risk group. In the moderate-risk group (score 2–3), although the frequency of severe VUR was only 14.9%, half of all severe VUR cases (42/86) were in this group. Thus, the decision to perform VCUG in this group should be carefully evaluated by clinicians. In the high risk group (score 4–6), the rate of severe VUR was 51.2%. Clinicians should not hesitate to request VCUG for patients in this high risk group. Although some cases of VUR might be missed with our model, high grade VUR cases will be detected, and unnecessary VCUGs for patients without VUR or with clinically insignificant VUR will be avoided.
Limitations of our study include retrospective design which might have precluded the fully differentiation between cystitis and pyelonephritis due to incomplete clinical and laboratory data on the patient files. In addition, information on the method of urine culture collection was not available in most patient records, so instances of growth below 100.000 colonies in samples obtained by urinary catheterization may have been missed. Furthermore, approximately half of the patients included in the study underwent DMSA imaging. Patients without a DMSA scan could achieve a maximum score 4. However, evaluation of children without DMSA scintigraphy also showed that probability of severe VUR is 20 times higher in those with a score 4 than the children with a score 0 to 3.