Process of estimating the number of people living with HIV in India.

This article attempts to appraise the methodology of estimating the people living with AIDS in India. In this estimation, the primary demographic data were obtained from Registrar General of India and Sample Registration System. HIV prevalence data was drawn from HIV Sentinel Surveillance conducted at 704 sites in 2005. The HIV prevalence at each site was estimated using the formula p +/- 1.645O(pq/n), where p represents the proportion of HIV positives and n the number of attendees at the site. Statewise prevalence for each risk group was estimated with the help of the formula giving 90% confidence internal for median prevalence as Median of the proportion positive +/- 1.645 x 1.2533O(pq/n), where 1.2533 was the adjustment for large variations involved. If the estimated median prevalence comes out to be zero, it was replaced by the actual prevalence of the low prevalent states for each risk group. In case of general population, the urban-rural ratio came out to be 1:1 in case of high prevalence states and 2.4:1 in case of other states. Having adjusted for age, the sex differential was 1.2:1 in gereralised epidemic states, 2:1 in concentrated epidemic states and 3:1 in low epidemic states. Overall additions were made for FSW by applying to them the HIV prevalence observed in FSWsites under TI. Prevalence in child population was done by applying the GFR to the women with HIV and then multiplying it by 0.3 (percentage infected).