But the latest method involved both surveillance data and a population-based survey. This tested blood samples and covered about 200,000 people (ages 15-54), including face-to-face interviews, across the country between December 2005 and August 2006.

“In the earlier method, HIV surveillance among pregnant women attending antenatal clinics was used as proxy for prevalence in the general adult population. This pushed estimates higher than the ground reality since that method is suitable for countries with a generalised epidemic, not those were epidemic is low or concentrated, like in India,” said Arvind Pandey, director National Institute of Medical Statistics, the ICMR-headed body which conducted the survey.

“The latest approach depends on knowing the population size as well as behaviour of both the high-risk and general population. This time we used data from surveillance sites and also from the community based study of National Family Health Survey,” added Pandey.

“This is a much better way of estimation and we completely stand by the figures as the study this time has community-based figures,” said Denis Broun, UNAIDS country representative in India speaking to The Indian Express from Paris.

This change of approach hasn’t come easy. In fact, UNAIDS differed with official figures on the number of AIDS deaths each year. While official figures were 10,000, UNAIDS estimated that in 2004-2005, over 1 lakh people died due to the disease.

When a research article published in a British Journal claimed that the number of people living with HIV in India could be lower than government estimates, the UN warned against drawing “hasty conclusions”.

... contd.