Rolling Stone Investigative Report Examines the Accuracy of AIDS Statistics in Sub-Saharan Africa

In the Nov. 22 issue of Rolling Stone, South African journalist Rian Malan investigates why different organizations have different HIV/AIDS statistics for South Africa and other sub-Saharan African countries. In July of last year -- three months after South African President Thabo Mbeki convened a panel to re-examine the relationship between HIV and AIDS -- Malan began his investigation into Africa's AIDS statistics. He, like Mbeki, began his search for information on the Internet. At the time, Malan found that according to the World Health Organization and UNAIDS, an estimated 22 million Africans were HIV-positive and 14 million additional people had died of AIDS-related complications. Through interviews with WHO and UNAIDS officials, Malan discovered that these statistics were not based on actual case reporting, as the numbers are in the West, where "[a]lmost every new AIDS case is scientifically verified and reported to government health authorities" who report their totals to WHO officials in Geneva, Switzerland. Instead, HIV/AIDS statistics for African nations are primarily derived by testing pregnant women at government-sponsored prenatal clinics. According to UNAIDS, blood is drawn for syphilis testing and then once a year, blood left over from the syphilis tests is collected and tested for HIV. The results are then "fed into a computer model that uses 'simple back-calculation procedures' and knowledge of 'the well-known natural course of HIV infection' to produce statistics for the continent."

Comparing South Africa
Public health officials can "theorize at will about the rest of Africa," but South Africa is a "semi-industrialized nation with a respectable statistical service," Malan writes. Although "coverage is far from complete," Ian Timaeus of the London School of Hygiene and Tropical Medicine and a UNAIDS consultant said, "South Africa is the only country in sub-Saharan Africa where sufficient deaths are routinely registered to attempt to produce national estimates of mortality from this source." Eighty percent of deaths are registered in South Africa compared to 1% in the rest of sub-Saharan Africa. According to the WHO model that tests pregnant women, AIDS-related deaths have tripled in South Africa over three years, from approximately 80,000 deaths in 1996 to 250,000 deaths in 1999. But when Malan compared the increase in deaths with the number of registered South African deaths for those years, he found "no such" increase, as registered deaths increased from 294,703 to 343,535 over the same time period. He contacted Statistics SA, the government's census agency, to verify the numbers and was told that they were correct.

WHO Is Correct?
Confused by the discrepancy in the numbers, Malan examined other population-based surveys in an attempt to discern sub-Saharan Africa's real HIV/AIDS figures. But the numbers gained through field research and those collected by national agencies failed to add up, so Malan began to question the validity of WHO's predictions. He notes that gross overestimates of the extent of the AIDS epidemic in the United States were made in the 1980s and early 1990s. In 1985, the CDC said that 1.5 million Americans were already infected with HIV and that two to three million would be infected within 10 years. However, a 1997 analysis found that only 450,000 people were infected in the mid-1980s. "If the numbers could be gotten so wrong in America, what are we to make of the infinitely more dire death spells cast upon the developing world?" Malan asks.

Testing Discrepancies
One reason for WHO's inflated HIV/AIDS estimates may be the type of HIV testing being done in sub-Saharan Africa, Malan theorizes. Unlike Western countries, which use two primary tests (ELISA tests) and a confirmatory test for HIV, the standard protocol in developing countries is one ELISA test, which means "almost nothing" in the United States, Malan notes. In addition, he says that false-positive test results are more common in sub-Saharan Africa due to the presence of plasmodium falciparum, a parasite that causes malaria, in many people. But Malan wonders why, if there may be so many false positives, governments do not take stronger measures to ensure accurate testing.

More Money
Higher HIV rates mean more money from foreign aid groups and other governments, which currently spend about $1 billion a year in AIDS-related funding, Malan says. The money also "translate[s] into patronage for politicians and good jobs for their struggling constituents," he adds. Still, UNAIDS' chief epidemiologist Dr. Bernhard Schwartlaender said that the agency's computer model, which is the same as the one used by WHO, is "completely dependable." He acknowledged that the model is just an estimate and over time the curves for the model and the recorded death rate will coincide. "The models may completely disagree at a particular point in time, but in the end the curves look incredibly similar. They're goddamn consistent," Schwartlaender said. However, Malan remains unconvinced and confused. "Are these numbers right?" he asks. "Who knows," he answers, adding that the "entire quest has driven me ... mad" (Malan, Rolling Stone, 11/22).