Role of urinary tract infection in bladder cancer: a systematic review and meta-analysis.

PURPOSE

We sought to examine the literature reporting the effect of urinary tract infection (UTI) on non-schistosomiasis-related UBC (UBC) through a systematic review and meta-analysis.

METHODS

A predefined study protocol was developed according to PRISMA. Medline and Scopus were searched for all studies investigating exposure to UTI with UBCNS as the primary outcome. Potential studies were screened against eligibility criteria. Clinical heterogeneity was assessed and groups with more than two studies were evaluated by random effect meta-analysis. Study-level bias was assessed with the Newcastle-Ottawa Scale (NOS). In cases of substantial between study heterogeneity (I > 50%), predefined sensitivity and subgroup analyses were performed.

RESULTS

Of 16 eligible studies, eight case-control studies spanning four decades and five countries were suitable for quantitative analysis. Main analysis favored exposure to UTI increasing risk of subsequent UBC (RR 1.33 [95% CI 1.14-1.55]). This effect was no longer statistically significant after excluding studies published prior to year 2000 and at high risk of bias. Between study heterogeneity was considerable for nearly all analyses and not reduced by predefined sensitivity or subgroup analyses.

CONCLUSION

Exposure to UTI favors increased risk for UBC, particularly in men, but these effects were statistically insignificant when pooling data from the most recent and highest quality studies. These data do not support findings of previously published studies, that report on heterogenous populations with poor definitions of UTI and minimal control for important confounders. Results from previous studies should be viewed as hypothesis generating. This review highlights the need for higher quality investigation.