U.N. Revises HIV Prevalence Estimates

Overall number reduced

CORRECTION APPENDED

The Joint United Nations Program for HIV/AIDS (UNAIDS) acknowledged last week that it had overestimated the prevalence of the global HIV epidemic, confirming a Harvard School of Public Health professor’s three-year-old claim that the UNAIDS figures were too large.

UNAIDS decreased its estimate of the number of people living with HIV from 39.5 million in 2006 to 33.2 million in 2007.

In previous reports, UNAIDS had estimated national rates by extrapolating data collected at public health clinics, where young women who thought they might be pregnant or infected with a sexually transmitted disease were given anonymous blood tests. But that method gave extra statistical weight to groups with higher HIV rates, such as prostitutes and drug abusers, and virtually disregarded rural women.

This year’s report uses national household surveys, which are believed to be more representative of the prevalence of HIV. Under the new method, the number of HIV victims in India dropped by more than half, to 2.5 million.

Daniel T. Halperin, a professor at the School of Public Health, co-authored an article three years ago asserting that the old estimate was too high. [SEE CORRECTION BELOW]

“The method adopted before is the easiest way of getting statistics, but the current method is better,” Halperin said.

India dropped in a ranking of countries with the greatest number HIV victims, from first place to third. South Africa and Nigeria claim the most victims.

Unlike India, some African countries, including Uganda, Zimbabwe, and Kenya, are witnessing a drop in new cases unrelated to changes in methodology.

“The disease has been around longer in these countries,” Halperin said. “The drop is caused by change of people’s behavior, due to many factors such as improving AIDS education and increasing fear of the disease.”

Halperin said the decrease can be attributed to heightened awareness of the risk of multiple sexual partners, driven by education efforts led by churches, community groups, and governments.

“Over time, people are fixed to one partner, and this really slowed down the HIV infection growth,” Halperin said. “It doesn’t take much money to make these changes happen.”

Halperin does not think the report will hurt progress in fighting the epidemic, even if donors reduce future donations because of the lower estimate.

“In the short run, overestimation might be helpful to get more money, but it will cause confusion in the long run,” he said. “HIV infection rates don’t go down just because of the money.”

CORRECTION

The Nov. 26 article "U.N. Revises HIV Prevalence Estimates" incorrectly stated that Daniel T. Halperin is a professor at the School of Public Health. In fact, he is a senior research scientist.