PRETORIA (AlertNet) - I lost my virginity about statistics during the 1992 census in Zimbabwe.

The census worker is a bright, cheerful university student. She wants to tick off all the members of my household as Europeans. I object. I am from Uruguay and my family has lived there for 150 years. My baby daughter is half-Uruguayan and half-Palestinian. Her nanny is a Guarani Indian from Paraguay. Baffled, the census worker calls her supervisor.

A bright, assertive, rotund civil servant strides in. She listens. She nods. "I see your point. But in this questionnaire, you can either be African, Asian or European." End of story. End of my faith on censuses. And I had earned an A+ with honours on my stats class at journalism school.

Two years later, Mozambican refugees in Zimbabwe go back home. They register for repatriation. It turns out there were about 30,000 ghost refugees. No wonder shops in Manica province stocked up so much USAID cooking oil and tinned fish. The camps -- and their managers -- had been getting extra food for a decade.

To make it short, I take statistics with a grain of salt. So I was not much perturbed when last week in its annual report UNAIDS revised downwards the HIV prevalence rates in Burundi, Kenya, Mali, Niger, South Africa, Zambia and Zimbabwe.

Does it matter if 200,000 or 300,000 people perished in Hiroshima? Does it matter if Rwanda's "genocidaires" slaughtered 800,000 or one million? Not really. The tragedy, the horror and the indignation do not decrease with the number of victims. Same with AIDS.

No, what bothers me most about AIDS data is how it is used and misused.

Today, among its AIDS Facts, The Pretoria News reported: "One out of two South African boys aged 15 will die of AIDS." No source, no explanation. Just factoids masquerading as facts. Just scaring young people. "Well, if I have to go, I will go with a blast," they must think. Throw condoms and caution away.

When fundraising, NGOs and foundations paint the most wretched picture ever of Africa. They make it look like only orphans and AIDS-stricken people remain in the villages. So does the Western media. The New York-based Village Voice has called it "a continent of orphans", the Economist "the hopeless continent", and so forth.

Take, for example, the well-intentioned UN special envoy for HIV/AIDS, Stephen Lewis. Every now and then he jets in for a Magical Misery Tour. Afterwards, a desolate Lewis describes to the press corps in Johannesburg the horrors he has witnessed. As I listen, I wonder: do I live on the same continent he describes? Are African women threatened with disappearance or is this hype? I understand his job is to see the world through the eyes of the virus, but he is Mr. Doom-and-Gloom himself.

Only researcher Alex de Waal surpasses Lewis in apocalyptic predictions for HIV/AIDS. He thrives on the worst case scenario for Africa. The food shortages of 2002 would become "the famine of mass starvation, the new variant famine." AIDS will turn women into virtual slaves, men with AIDS will go on the rampage, authority will collapse and civil wars explode. Well, not quite yet.

Aid agency press releases should be used in journalism courses as a study on statistics out of context, unexplained, distorted and manipulated. About a year ago, a WFP press release spoke of "the vanishing families of Zambia." Yes, Zambian families struggle, but vanish? Makes them sound like the black rhino.

Let us remember: AIDS statistics estimates and projections are based on computer-assisted modelling. Journalists should take them as approximations, only.

UNAIDS and the WHO revised their figures for the seven African countries because population-based household surveys have supplied more varied and precise data than antenatal clinics. In Zambia, additionally, several urban areas had been misclassified as rural, which tend to have higher prevalence rates.

One country's revision, though, I find worrisome. In Zimbabwe, HIV prevalence among people aged 15-49 dropped from 32% to 24%.

The reason for the new figure is that "samples were estimated using a more specific test kit," according to a study published in the Lancet, the British-based medical journal. It goes on to cite "poor handling and storage of samples or the use of out-of-dare tests or poor quality of reagents" as factors that distort results.

In plain speak, re-testing found a lot of false positives in Zimbabwe in 2000. If surveillance testing quality is so poor, how is it in the voluntary testing and counselling centres? How many false positives -- or negatives -- are circulating? I tested for HIV in Harare the same year. Can I trust the result?

This is weird. Zimbabwe' s rate should be going up, not down. All the co-factors that spread AIDS are there: internal displacement, violence, unemployment, hunger and poverty.

I cannot ask my Zimbabwean AIDS contacts what happened because they are all in Bangkok having massage and Thai food, and yes, attending the AIDS conference. I hope they are all practicing safe sex and having their stats straight, no chaser.

Read Alex de Waal's response, VIEWPOINT: AIDS is our adversary, however resourceful Africans are

Any views expressed in this article are those of the writer and not of Reuters.