• Use of protection (male and female condoms)
  
• STIs
  

This three-phased study examined ways to improve services for the prevention of STIs among commercial sex workers. This study was conducted in identical fashion in two sites (Tamatave and Antananarivo) in Madagascar. It tested whether condom promotion through community-based counseling or clinic based counseling leads to an increased level of protection among high-risk sex acts and a subsequent decrease in STI infection rates.

In Phase I, women at each site were randomly assigned to a study arm receiving a community-based male condom promotion intervention, or a study arm receiving community-based male condom promotion supplemented with clinic-based counseling. Randomization was accomplished by the use of sequentially-numbered sealed, opaque envelopes containing the group assignment. The randomization list was stratified by study site. Phase I began with baseline measurement of male condom use and STI prevalence. Participants returned three times, at two month intervals, for follow-up. At each visit, face-to-face interviews were conducted to estimate the proportion of protected sex acts. At the third visit only, participants were tested for gonococcal, chlamydial, and trichomonal infections. This reliance on a combination of biologic and behavioral outcomes helped to reduce any bias brought on by the inherent limitations of either of these measures on its own.

Half the women completing Phase I at each site were then randomized to a Phase II study arm receiving a community-based male and female condom promotion intervention, while the other half entered Phase II receiving the same community-based male and female condom promotion supplemented with clinic-based counseling. This sequential randomization scheme allowed measurement of the short-term (within 6 months) effect of clinic-based counseling on male condom use and STI prevalence, and the short-term effect of clinic-based counseling on male condom + female condom use and STI prevalence. A similar data collection process was used, with measurement of male and female condom use in all three visits, and STI testing in the third visit only.

In Phase III, participants were followed for an additional 6 months, in the same intervention group to which they were assigned at the beginning of Phase II, to assess the medium-term impacts of the female condom, after participants had a chance to become accustomed to the female condom.