Abstract
OBJECTIVE: To estimate whether a gel containing the spermicide C31G was noninferior to a commercially available product containing nonoxynol-9. METHODS: Participants were healthy, sexually active women aged 18-40 years. Measured outcomes included pregnancy rates, continuation rates, adverse events, and acceptability. The primary study outcome was contraceptive efficacy. Sample size was calculated at a 2.5% significance level using a one-sided test based on assumed 6-month pregnancy probability of 15% in the nonoxynol-9 group. Sample size was sufficient to reject, with 80% power, the null hypothesis that pregnancy probability in the C31G arm would be more than 5% higher. RESULTS: Nine hundred thirty-two women were randomized in the C31G group and 633 in the nonoxynol-9 group. For randomized patients with at least one episode of coitus (modified intent-to-treat group), 6-month pregnancy probabilities were 12.0% (95% confidence interval [CI] 9.3-14.7%) and 12.0% (95% CI 8.7-15.3%) for C31G and nonoxynol-9,respectively. Twelve-month pregnancy probabilities were 13.8% (95% CI 7.6-20%) for C31G and 19.8% (95% CI 10.9-28.7%) for nonoxynol-9. Two serious adverse events were deemed possibly related to study product and neither occurred in the C31G group. Three fourths of users in either group reported that they liked their assigned study product. Approximately 40% of patients discontinued prematurely for reasons other than pregnancy with 11% lost to follow-up. CONCLUSION: C31G demonstrated noninferior contraceptive efficacy compared with nonoxynol-9. C31G may provide another marketable option for women seeking spermicidal contraception. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov, www.ClinicalTrials.gov, NCT00274261.