Do spermicides containing nonoxynol-9 prevent sexually transmitted infections? A meta-analysis.

BACKGROUND AND OBJECTIVES There are conflicting opinions on whether to recommend spermicides containing nonoxynol-9 for prevention of sexually transmitted diseases, including human immunodeficiency virus (HIV). GOAL To systematically review and summarize the medical literature on the effect of spermicides containing nonoxynol-9 on prevention of gonorrhea, chlamydial infection, and HIV. STUDY DESIGN Meta-analysis. Potential articles were identified through computerized literature searches. Articles were included if the design was clinical trial, cohort, case-control, or cross-sectional; original outcome data were presented for gonorrhea, chlamydial infection, or HIV; and spermicides containing nonoxynol-9 were used separately from other barrier methods. Study characteristics including design, population, spermicide dose, and delivery method were abstracted. Relative risks (RR) and 95% confidence intervals (CI) were determined from information published in the study or from study authors. Summary risk estimates were computed for clinical trials. RESULTS Twelve eligible articles were identified, including six clinical trials and six observational studies. Eleven articles evaluated gonorrhea; each found a reduced risk of infection with spermicide use. Among six clinical trials, the summary RR was 0.62 (95% CI, 0.49-0.78). The five articles evaluating chlamydial infection also found significant reduction among spermicide users, with the four clinical trials having a summary RR of 0.75 (95% CI, 0.62-0.91). The degree of spermicidal protection against gonorrhea and chlamydial infection did not differ materially among studies with different study designs or spermicidal preparations (i.e., sponge, gel, suppository, or film). Three studies evaluated degree of protection according to consistency of use and found stronger protection with more consistent use. The two studies on spermicidal prevention of HIV had contrasting results: An observational study found a significant protective effect (RR = 0.1), whereas a clinical trial found a nonsignificantly increased risk (RR = 1.7). CONCLUSIONS Nonoxynol-9-containing spermicides have an appreciable protective effect against both gonorrhea and chlamydial infection, and wider use of spermicides might substantially reduce the incidence of these diseases. However, insufficient data exist to judge their effect on HIV transmission, and more research on the effect of nonoxynol-9 on HIV transmission is urgently needed to make evidence-based clinical decisions and public health recommendations in the future.