Statistical comfort?

AIDS Count: New numbers, but no consolation yet.

Last fortnight, when new data relating to the prevalence of AIDS/HIV disease in India was released, the infected population went down from earlier estimates of 5.2-5.6 million to 2-3.1 million people. Those associated with the epidemic had long suspected that the Indian numbers could be below those projected. The new estimates made official this view. But here's the thing: the decline is not in the absolute numbers of those infected. It is just that the estimates of the people who might be infected are down.

So what is it: clever semantics or more? Well, it is just that the mode of estimation changed this time around. Earlier, NACO, the nodal organisation for formulation of policy and implementation of programmes for prevention and control of HIV/AIDS in India, arrived at estimates through a random sampling of cases from around 703 sentinel surveillance sites that have been set up in areas that show high prevalence. Each site contributed 400 samples. The drug samples were collected, tested and verified. This database was then extrapolated to the denominator of a billion people to get the figure of 5.2 million people and 0.91 per cent prevalence. Prevalence, defined as the number of current cases per population at risk, has dropped to 0.36 with the new estimates. This time around, apart from the scale-up in the sentinel surveillance sites to 1,122, NACO has used additional inputs from the National Family Health Survey, which is a countrywide community-based household survey.

How reliable are these numbers? Very reliable, says K. Sujatha Rao, additional secretary and director general of NACO. Besides Indian experts, the numbers have been vetted by eight leading experts round the world, including UNAIDS, who, CDC, Imperial College of London, among others. However, as Rao points out, "Estimates are estimates and will keep changing as we get better data and insights into the epidemic." Do the new numbers change India's response to the epidemic? Rao is clear it does not. "There will be no change because the mainstay of our strategy is prevention and prevention efforts do not change with scale."

As says Ashok Alexander, who heads Avahan, the India anti-aids initiative of the Bill and Melinda Gates Foundation, the real point is not to get caught up in the single number estimation game but to look at the underlying trends in the epidemic. "About one in 6 districts in India is high prevalence. These levels are still high and not coming down fast enough yet," he adds. And that's the point.