Background We assessed the impact of a large-scale, targeted HIV prevention program among female sex workers (FSWs) in Karnataka state, south India.

Methods In the context of an HIV prevention program involving 18 districts, funded by the Bill & Melinda Gates Foundation, we conducted initial (R1) and two follow-up (R2 and R3) survey rounds (integrated behavioural and biological assessments), involving random samples of urban FSWs in three selected districts. The total urban FSW population size of the three districts is estimated at 8600. Initial surveys were conducted at 7–19 months following program commencement in 2004–2005, and each follow-up survey at two to three year intervals thereafter. Multivariable logistic regression models were used, with weights to account for a complex sampling design. Results presented are for the three districts combined.

Results 1242 FSWs participated in R1, 1238 in R2 and 1248 in R3. In R3, 98% and 92% of FSWs reported contact by a peer educator and having visited a project STI clinic, respectively. HIV prevalence declined steadily from R1 to R3, and in R3 compared to R1, there were reductions in HIV prevalence (24.8% vs 13.3%, adjusted OR (AOR)=0.59, 95% CI 0.41% to 0.86%, p=0.005); active syphilis (12.7% vs 8.8%, AOR=0.44, 95% CI 0.31% to 0.61%, p<0.001); and high-titre syphilis (6.6% vs 3.2%, AOR=0.33, 95% CI 0.21% to 0.51%, p<0.001). Trichomoniasis (Tv), chlamydial infection (Ct) and gonorrhoea (Ng) were assessed in one district in R3. From R1 to R3, Tv declined from 32.9% to 3.5% (AOR=0.07, 95% CI 0.04% to 0.15%, p<0.001); Ng from 5.4% to 0.9% (AOR 0.22, 95% CI 0.0.07% to 0.71%, p=0.01); and Ct from 10.8% to 4.6% (AOR 0.38, 95% CI 0.19% to 0.75%, p=0.006). Comparing R3 with R1 in all three districts, reported condom use at last sex increased significantly for repeat clients (60.1% to 93.9, AOR 11.2, 95% CI 7.2% to 17.5%, p<0.001); for occasional clients (81.0% to 88.2%, AOR 2.8, 95% CI 1.9% to 4.1%, p<0.001); and for regular (non-commercial) partners (24.4% to 55.9%, AOR 5.0, 95% CI 2.4% to 10.3%, p<0.001). Condom use overall was strongly associated with level of peer educator contact, having been exposed to condom demonstrations, and attendance at a program health clinic.

Conclusions Over 6 years, this large-scale HIV prevention program for FSWs achieved high rates of program coverage, high rates of condom use with clients, and significant reductions in HIV and STI prevalence. It will be important to sustain these efforts in the long term.