Background To evaluate the impact of Avahan, the India AIDS Initiative of the Bill & Melinda Gates Foundation, two rounds of integrated biological and behavioural assessment (IBBA) surveys were conducted in 2006–07 and 2009–10 among high-risk men who have sex with men (MSM-T) across 14 districts in south India to measure the prevalence of HIV and sexually transmitted infections (STIs), and related risk factors (n∼4,300 per round). We studied the determinants for decline in HIV prevalence among MSM-T.

Methods Random slope multilevel models were developed using logistic regression procedures to examine the effects of round, district level and individual variables on HIV prevalence among MSM. The model also controlled for baseline HIV prevalence.

Results HIV prevalence among MSM-T significantly reduced from 13.1% in round 1 to 11.4% in round 2 (p = 0.017). MSM-T who were aged 25 and above, whose self- reported receptive sexual role (Kothi), who had male sexual clients, who were illiterate, who first had sex before age 15 and who ever used condoms, were significantly more likely to have HIV. Surprisingly, MSM-T who had paid a female partner were less likely to have HIV (OR: 0.71 p = 0.008). There was a statistically significant effect modification between round and syphilis, with a statistically significant decline in HIV prevalence between survey rounds among MSM-T without syphilis (OR: 0.73, p = 0.002), whereas there was a non-significant HIV increase among MSM-T with syphilis (OR: 1.36, p = 0.086).

Conclusion Overall HIV prevalence declined significantly among MSM-T from round 1 to round 2 in Avahan intervention districts in south India. The lack of decline in presence of syphilis suggests that the latter is an important marker of risky behaviour. These results, in conjunction with those obtained in other groups targeted by the intervention; suggest that Avahan had a major impact on the HIV epidemic in south India.