Twelve studies (n=402,908) were included; one RCT (n=70), four cohort studies (n=400,700) and seven case-control studies (n=2,138).
The RCT provided no information about methods of randomisation, allocation concealment or blinding. The quality of the cohort and case-control studies was variable: only one study adjusted for age, socioeconomic factors and ethnicity, and four studies did not adjust for any confounding factors.
The RCT showed no statistically significant reduction in the incidence of UTI with circumcision (OR 0.13, 95% CI: 0.01, 2.63, p=0.18).
The cohort studies showed a statistically significant reduction in the incidence of UTI with circumcision (OR 0.13, 95% CI: 0.07, 0.24, p<0.00001). There was significant heterogeneity between these studies, which was eliminated on the removal of one study from the analysis.
The case-control studies showed a statistically significant reduction in the incidence of UTI with circumcision (OR 0.13, 95% CI: 0.07, 0.23, p<0.00001).
Overall, across study designs, there was a statistically significant reduction in the incidence of UTI with circumcision (OR 0.13, 95% CI: 0.08, 0.20, p<0.00001). There was significant heterogeneity between these studies, again eliminated by the removal of the cohort study.
The NNT was 111 for the general population, 11 for those with past UTI, and 4 for boys with high-grade vesicoureteric reflux.