Nonoxynol-9 for preventing vaginal acquisition of STIs by women from men

The review found good evidence that the use of nonoxynol-9, a vaginal microbicide, does not protect women against sexually transmitted infections (STIs). The ten randomized controlled trials included in the review reported outcomes for gonorrhoea, cervical infection, trichomoniasis, bacterial vaginosis, chlamydia and candidiasis, although not all the trials assessed each of these outcomes. Use of nonoxynol-9 did not significantly reduce the risk of any of these infections. The review reports that genital lesions such as ulcers, are significantly more likely to occur in women using nonoxynol-9, but this result (relative risk [RR]: 1.17; 95% confidence interval [CI]: 1.02–1.35) is only marginally significant.

The authors conducted an extensive search of a wide variety of appropriate databases. They attempted to include both published and unpublished studies by searching conference databases and contacting researchers and funding agencies working in the field. Reporting of the results of the search could be improved by stating the number of abstracts retrieved and the degree of overlap between the databases.

One reviewer extracted the data and another checked the data extraction. No method of arbitration is reported, so it is assumed that all differences were resolved by discussion.

Meta-analysis was conducted when the trials reported the same outcomes. The review states that the fixed effects model was used because the test for heterogeneity did not reach statistical significance. However, the tests for heterogeneity for the meta-analysis of gonorrhoea, cervical infection and candidiasis are significant at the p < 0.05 level. Only the relative risk for gonorrhoea is reported using the random effects model, although this is not made explicit in the review.

The data are comprehensively presented in the text and in the Table of Included Studies and Tables of Comparisons.

The full RHL commentary also includes sections on: