Factors Predicting VCUG Results in Children with Studies Ordered for Urinary Tract Infection: An Institutional Experience

Back to 77th Annual Meeting
Back to Program Outline

Background: Urinary tract infection (UTI) is a frequent infectious disease and source of febrile illness in children. The optimum evaluation strategy of children with UTI is undecided and varies depending on location and provider preference. In the United States voiding cystourethrogram (VCUG) is frequently employed to try and identify those with UTI at risk for subsequent renal damage. This test, while well established, is invasive and is associated with a significant radiation exposure. We hypothesized that there would be factors that could be identified that would be predictive of a negative study, allowing for more judicious use of this imaging modality.
Methods: Charts were reviewed for 1073 patients who underwent voiding cystourethrogram (VCUG) at our institution from 2004 to 2007. Age, sex, indication for study, outcome of study (positive for any degree of reflux), whether patient was sedated, whether bladder emptied, whether patient had voiding dysfunction, and infectious organism were recorded based on final radiology reports and clinic notes. Five-hundred-eighty-six patients were identified who were under 18 years of age at time of their study and had an indication of UTI or febrile UTI. VCUG outcomes were compared to age, sex, sedation, emptying, and indication (insufficient data for voiding dysfunction and organism were obtained) by univariate (t-test for age, chi-2 otherwise) then multivariate logistic regression analysis.
Results: Of the 586 studies analyzed 173 (30%) were positive. Univariate analysis demonstrated that VCUG was significantly more likely to be positive with female gender 31% vs 19% for males (p=0.031) and indication of febrile UTI 43% vs. 23% for UTI alone (p<0.005). Age, use of sedation during the study and whether the bladder emptied had no impact on study outcomes. (Table 1) Multivariate analysis demonstrated that males were 48% less likely to have a positive study (OR= 0.52, CI 0.29-0.91). When adjusted for other factors studies were 2.5 times more liked to be positive with febrile UTI (OR= 2.5, CI 1.74-3.70).
Conclusions: VCUG is less likely to be positive in male patients and those with indication of UTI alone. In such patients other imaging modalities may be warranted. Sedation does not appear to impact VCUG results and seems a logical adjunct in the use of this invasive test. Further study of VCUG compared to less invasive tests to predict renal damage potential seems warranted given the overall low positivity rates.
Table 1

	Positive	Negative	OR (CI)
Sex			0.52 (0.29-0.91)
Male n=94	19 (20%)	75
Female n=492	154 (31%)	338
Age			0.70 (.0.39-1.29)
<=1 year old	42	86
> 1 year old	131	327
Sedated			0.88 (0.51-1.51)
Yes	118	287
No	55	126
Bladder Empty			1.37 (0.94-2.00)
Yes	99	264
No	74	149
Indication for Study			2.5 (1.74-3.70)
UTI alone n=387	88 (23%)	299
UTI and fever n=199	85 (43%)	114

Back to 77th Annual Meeting
Back to Program Outline