Africa: Row over which AIDS statistics to believe

Original Post Date: 2004-02-27  Posted By: Jan

From the News Archives of: WWW.AfricanCrisis.Org
Date & Time Posted: 2/27/2004 2:26:58 PM
Africa: Row over which AIDS statistics to believe

[So what are the real AIDS statistics? Well, in much of Africa, it is difficult to get proper statistics and no doubt most studies estimate as best they can. It is my belief that African countries feel vulnerable about AIDS and that they attempt to downplay its significance.

However, I feel that studies, like those cited by people like Dr Du Plessis (see the African Crisis archives for: South Africa: Beyond Democracy) are probably the closest. Generally speaking, if one speaks to doctors, nurses or people who work in our blood bank, they will tell you that the number of people infected with HIV are huge in number and that our state hospitals are overflowing with such patients. So when I see AIDS statistics I tend to believe the higher numbers quoted.

I suspect that most African governments are trying to cover this up because they fear “recolonisation” if AIDS should prove to be devastating.
Riaan Malan the “commentator” cited below may be one of those people trying to help to blur the results of scientific studies in the government’s favour by using various bogus “logic”. Jan]

Nairobi – Aids has Africa in its grip, but how tightly?

The exact answer varies according to which survey you consult – a fact of supreme concern to experts and governments fighting the pandemic on the world’s poorest continent.

Look at one set of figures and Africa is facing a disaster that threatens the existence of entire countries. Look at another and the crisis, while horrific, does not appear so widespread.

‘There is no perfect method of measuring something as elusive and stigmatised as HIV’
A bad-tempered debate about methodology flared in January after research in Kenya appeared to show that United Nations estimates of infection in the east African country could be inflated.

For some, the controversial finding cast doubt on the world body’s widely-used estimates for sub-Saharan Africa as a whole.

Emotions run especially high in South Africa where the government has been accused of moving too slowly against the disease. Commentator Rian Malan said the epidemic was being exaggerated, especially in South Africa, to pull in aid money.

“What if we wake up five years hence to discover the problem has been blown out of all proportion by unsound estimates, causing upwards of $20-billion to be wasted?” he wrote.

But a US researcher working in Kenya says no method of measuring HIV and Aids prevalence gets it absolutely right, and both the main systems at the centre of the dispute have benefits.

‘People of high risk fear the test’
“There is no perfect method of measuring something as elusive and stigmatised as HIV infection,” said Lawrence Marum, an epidemiologist at a research station in Kenya run by the US government’s Centres for Disease Control (CDC).

The established method is to extrapolate from tests from pregnant women at medical clinics. Another method less widely used is based on tests of volunteers, both men and women.

“Neither method is flawed but both are intrinsically biased. Both are equally useful sources of information and both are conducted to a high international standard,” Marum said.

Marum, whose station is in Kenya’s worst Aids-affected region, has worked on both systems in collaboration with other research institutions in the country where 1,5 million people have died from Aids.

The latest controversy stems from a recent official survey in Kenya carried out using the volunteer model that reported a much lower incidence of HIV infection than expected.

The research by the Kenya Demographic and Health Survey (KDHS) conducted in 2003 placed prevalence at 6,7 percent, much lower than the 9,4 percent rate reported by UNAids and the World Health Organisation.

Marum said UNAids’s sentinel surveillance, which tests for HIV in mothers visiting antenatal clinics, was biased because it excluded women past child bearing age and young women who were sexually inactive.

The KDHS model surveys women between 15 and 49 and men between the ages of 15 and 54. But it, too, was also biased because 14 percent of women and 13 percent of men in the survey refused to have their blood samples taken, he said.

“People of high risk fear the test. People who have high risk behaviour are more likely to be missed in a survey,” Marum said.

He said if the HIV infection rate in those whose blood was not tested by the survey was higher than those who were, then the KDHS data could well match the more alarming results collected by UNAids.

Another difference between the two methods is that volunteers make up KDHS data whereas testing under the UN system is done without the knowledge of the pregnant women.

Researchers say that trends can easily be monitored by the antenatal surveillance because it is done annually, whereas the KDHS is carried out only once in five years.

The sentinel surveillance has been the standard method of monitoring spread of the virus in Kenya since 1990.

The KDHS found an HIV prevalence rate of 8,7 percent among women – the same as UNAids estimates – but found a rate of 4,5 percent in men, significantly lower than expected.

James Mutunga of the Kenya Medical Research Institute said only three other surveys similar to KDHS have been carried out worldwide – in the Dominican Republic, Mali and Zambia.

All three found slightly lower rates of HIV infection, compared to the figures given by the UN.

Source: IOL
URL: http://www.iol.co.za/index.php?click_id=68&ar…br>

xxxxx Sign Up