Predictive factors of resolution of primary vesico-ureteric reflux: a multivariate analysis

To identify independent factors predicting the resolution of primary vesico-ureteric reflux (VUR) in a cohort of medically managed children.

Between 1977 and 2003, 506 children were diagnosed with VUR and were conservatively managed and prospectively followed. All of the children were maintained on antibiotic prophylaxis. Follow-up imaging consisted of voiding cysto-urethrography (VCUG) or a direct isotope cystogram at intervals of 2-3 years. The predictive factors used are based on the patient data at the time of entry in the protocol. The dependent variable was VUR resolution. The criterion for resolution was based on a single negative VCUG or direct isotope cystogram. A survival analysis identified variables significantly associated with VUR resolution. Cox's regression model was applied to identify variables independently associated with the dependent variable.

After adjustment, four variables remained as independent predictors of VUR resolution: nonwhite race, relative risk (95% confidence interval) of 1.5 (1.1-1.9; P = 0.009); mild grade of VUR, 3.3 (2.1-5.3; P < 0.001); absence of renal damage, 3.3 (2.4-4.5; P < 0.001); and absence of dysfunctional voiding, 2.0 (1.4-3.1; P < 0.001). For mild VUR, three variables were significantly associated: male gender, 1.7 (1.1-2.6; P = 0.012); absence of renal damage, 3.4 (1.8-6.4; P < 0.001); and unilateral VUR, 1.6 (1.1-2.3; P = 0.004). For moderate/severe VUR, three variables were significantly associated: nonwhite race, 1.7 (1.1-2.6; P = 0.01); absence of renal damage, 3.0 (2.0-4.4; P < 0.001); and absence of dysfunctional voiding, 2.8 (1.4-5.5; P = 0.004).

Few factors are amenable to intervention to modify the natural history of VUR. According to our findings, there are only two possible interventions: avoiding renal scars and managing voiding dysfunction.