Background As a part of the evaluation of Avahan, the India AIDS initiative of the Bill & Melinda Gates Foundation, a cross-sectional survey was undertaken in the general population of Bagalkot district, Karnataka state, South India, in 2009. This replicated an earlier survey in 2003 that had examined HIV prevalence and risk behaviours.

Methods The repeat survey in 2009 was conducted in the same rural and urban areas as the 2003 study. The study covered 10 rural villages and 20 urban blocks of three of the six talukas (sub-district units) in the district. In both surveys, a target sample of 6600 adult males and females was selected. Urine and blood samples were collected from all consenting participants for HIV and STI testing. An individual was deemed HIV positive, if positive on two different tests. We compared HIV and STI prevalence in 2003 and 2009. We also examined the age-specific distribution of HIV prevalence among rural and urban males and females at both time points. These analyses used logistic regression that considers survey design to adjust for characteristics of the population that may have changed between the two rounds.

Results Overall, HIV prevalence in the district declined from 3.16% in 2003 to 2.58% in 2009, although this decline was not statistically significant (OR=0.82, p=0.278). The prevalence of active syphilis was <1% and was similar in both periods. A slight decline in the prevalence of HSV-2 was observed over the period, but the difference was not statistically significant. There was a significant decline observed in HIV prevalence among persons aged 15–24, among persons below 30 years in urban areas, and among all women younger than 20 years of age (Abstract P1-S1.14 table 1). Among rural males aged 40 and above, we observed a significant increase in HIV prevalence from 0.71 to 5.33 (Abstract P1-S1.14 table 1), a level similar to that found in the younger age group in 2003.

Conclusion We observed a downward overall trend in HIV prevalence in Bagalkot district over the period examined, and a significant decline in HIV prevalence among younger age groups, where HIV prevalence more closely reflects incidence. The increase in HIV prevalence among older rural males may have been due to a cohort effect, as they aged over time. This may also have in part reflected increased survival in those age groups, as a result of the scaling-up of antiretroviral treatment programs throughout the state.