July 11, 2005
S. Africa health dept sharply hikes AIDS estimate
By Andrew Quinn
JOHANNESBURG (Reuters) - New figures from South Africa
suggest that more than 6.5 million of the country's 47 million
people may now be HIV-positive. The figure is a sharp jump on
previous estimates and is likely to fuel debate on the extent
of the country's HIV/AIDS pandemic.
at antenatal clinics, said results indicated that between 6.29
and 6.57 million South Africans now carry the HIV virus against
5.6 million at the end of 2003.
The figures contradict a May study by Statistics South
Africa, the state statistical agency, which estimated that
about 4.5 million South Africans were infected with HIV - a
toll which would drop South Africa behind India as the country
worst hit by the global pandemic.
Extrapolations from data at antenatal clinics, where
pregnant women have their blood tested, form the basis for most
estimates of HIV prevalence in Africa, home to more than 25
million of the estimated 39 million people infected with HIV
But the method has been criticized in some African
countries as exaggerating the spread of HIV/AIDS on the
Debate flared last year with a study purporting to show
that U.N. estimates of AIDS prevalence in Kenya were inflated
and cutting the projected number of Kenyan adult HIV infections
to 1 million from as many as 3 million.
UNAIDS, the United Nations AIDS agency, dismissed the
study's conclusions as unfounded, saying differences in
methodology could account for different numbers and standing by
its original forecasts.
South African officials including President Thabo Mbeki
have also questioned the severity of the HIV/AIDS crisis,
infuriating activists who blame the government's slow response
for increasing numbers of AIDS deaths in the country.
The Department of Health study said that 29.5 percent of
the pregnant women surveyed nationwide were HIV positive, up
from 27.9 percent in 2003. In KwaZulu-Natal, the worst affected
province in the country, the HIV prevalence rate among pregnant
women was more than 40 percent, it said.
The department acknowledged that its latest estimates for
national prevalence assumed that HIV prevalence among all
pregnant women was the same as that for those visiting
antenatal clinics, and that HIV prevalence for all women
between the ages of 15-49 was the same as that in pregnant
Some statisticians questioned the results.
Professor Rob Dorrington, head of the Center for Actuarial
Science at the University of Cape Town, said the department's
new estimates were "undoubtedly too high," although he told
Business Day newspaper the mixed messages from government were
complicating South Africa's response to HIV/AIDS.
"Is it any wonder the public is confused when the same
government offers estimates that differ by between 2 million
and 2.5 million?" he said.