They say that different HIV test procedures in Africa produce such widely differing results that their use should be re-examined. Yet at present some people are being "frightened to death" by a positive diagnosis. By the end of last year, the National AIDS Prevention and Control Programme had received a cumulative total of 7,124 reports of full-blown AIDS since the first cases were recorded in 1985. That represents fewer than a thousand a year, relatively few in a nation of 8m people.

But, according to screening surveys conducted late last year, as many as four out of 10 sexually active people are now testing HIV positive, and a million Zambians could be infected with the virus. Those findings have horrified most politicians and AIDS workers, and spurred the government into launching a new anti-AIDS campaign.

Guy Scott, an MP and former cabinet minister, says the disease threatens to orphan 2m children, and to take the lives of large numbers of staff in companies, public utilities and government. "It is ripping through the system. It is an absolute disaster," he said.

But Dr Francis Kasolo, head of virology at the University Teaching Hospital in Lusaka, said work in his department suggests the HIV figures cannot be taken at face value.

"We have found a big problem with false positives. When we repeat the tests, there are a lot of disparities in the results. A test kit from one manufacturer behaves differently from another's." The conclusion, he said, was that "most of our results are more or less compromised".

Most of the country's 80 testing centres were unable to afford a more expensive, confirmatory procedure after an initial positive test. Even that second test, known as Western Blot, produced widely differing results.

A third, rapid test, still in use at some clinics, had been shown to produce up to 40% false positive results in patients infected with malaria. Blood "stickyness" of patients, unrelated to HIV, also produced false positives.

Dr Wilfred Boayue, the World Health Organisation's representative in Zambia, says the recent surveys show such a big increase in positive results compared with six to seven years ago, when the proportion was only about 5 to 8%, that he shares concern that the country is in the grip of an HIV epidemic.

Kasolo, however, believes changes in the type of test kit used may contribute to the changing picture. He says international aid for the developing countries is often tied to use of materials provided by the donor nation.

"Most of the kits are supplied by the donors. If one decides not to provide funds any more, we move to another who will, and the kits come from that country instead. So the kits vary a lot: reporting can be high or low, depending on the kit.

"We have had individuals tested in one laboratory, and told they are positive, who move onto another, where they are negative."

Kasolo said the picture had been further confused by a phenomenon called "transient antibody to HIV" reported at a recent international meeting. A Uganda-based professor of virology had seen that some HIV positive patients subsequently tested negative.

Kasolo agreed with a recent call by scientists in Australia for use of HIV test kits to be reappraised, in the light of evidence that many conditions apart from HIV infection such as TB, malaria, malnutrition and multiple infections can cause a person to test positive.

"It is important that we address the whole issue of HIV in Africa scientifically," Kasolo said. "There is something going on that we do not understand."

Dr Sitali Maswenyeho, a paediatrician at the University Teaching Hospital and former fellow in AIDS research at the University of Miami, said he had long argued against the HIV test. "It's non specific," he said. "The test itself is killing a lot of people here. The stigma is doing the damage.

"We have malnutrition, bad water, poor sanitation; and when on top of that you are told you have an incurable disease, that really cuts off people's lives."

Despite concerns on the tests' validity, the presence of a severe form of immune system failure, affecting mainly sexually active people, is generally acknowledged.

There is argument, however, over its causes. Kasolo questions the "new virus" theory maintaining that a variety of sexually transmitted infections might be responsible. This view is shared by many older Zambians.

David Chipanta, 22, an HIV positive man helping with the work of an AIDS education and counselling organisation, says: "People in the villages tell us it is not new, but that it has become worse because of promiscuity."

Chipanta disagrees, arguing that even in the past, people were promiscuous. But he supports the challenge to HIV testing. *