Seven studies (1,520 participants) were included in the review: five published RCTs, one unpublished RCT and one non-RCT. The quality of the included studies varied. Most studies had low risk of detection and performance biases and high risk of selection and attrition biases. Randomisation and adequate allocation concealment were inadequate in four studies.
The meta-analysis showed an overall reduction of symptomatic urinary infection with prophylactic antibiotics compared to controls (RR 0.45, 95% CI 0.28 to 0.72; Ι²=16%; seven studies). The absolute reduction of symptomatic urinary tract infection was 5.8% and the number needed to treat to prevent one symptomatic urinary tract infection was 17 (95% CI 12 to 30).
Similar results were observed when removing the non-RCT or the unpublished study. The result favoured the subgroup of surgical patients; the subgroup of mixed populations showed no significant effect. The funnel plot suggested some evidence of publication bias.