Ironically — given their attack on my person in response to my argument in noseweek that African HIV and Aids statistics are grossly over-exaggerated — on the major issues, the Treatment Action Campaign’s (TAC) Nathan Geffen agrees with me.
In his response, recently published in the Mail & Guardian (Letters, January 30), Geffen says: “We do not know, never had known and probably never will know the exact number of people infected with HIV in any developing country.”
Well, yes. My point exactly. We do not know and have never known. But this has never stopped the Aids industry from issuing claims that give the opposite impression.
Geffen says: “UNAids annually published, computer-modelled epidemiological data must be treated with caution … Some, maybe many, of their calculations have overestimated the epidemic.” Again, we agree: UNAids overestimates the African epidemic.
Geffen says: “There are some consultants with little knowledge or experience making wild forecasts about the size of the epidemic.”
Couldn’t agree more. I just wish Geffen could bring himself to name these bad apples. They are not marginal players. They are widely respected “experts” whose services (especially to business) cost a fortune and whose utterances are treated with reverence by Aids reporters.
Finally, Geffen says: “Estimates made by the Department of Health and the Actuarial Society of South Africa have for many years been considered more definitive for the South African epidemic than the UNAids ones.”
Again, my point exactly. Local Aids industry insiders have known for years that UNAids estimates are absurd. For proof, one need only turn to the TAC’s press releases for 2003, which invariably claim an Aids-death rate of “600 a day,” or 219 000 a year.
If the TAC had continued to use UNAids estimates, the cited death toll would have been closer to half a million. One presumes the TAC refrained from using the higher United Nations estimates because it knew it would invite ridicule and accusations of wolf-crying.
This is a crucial point, given that South Africa is the only country in Africa with viable death registration machinery. Only in South Africa is it possible to validate the output of Aids computer models against real-life death registrations. Over the past three years South African researchers have been attempting to do just that.
The core of my argument (not even mentioned by Geffen, let alone refuted) is that their work has resulted in several downward revisions of South Africa’s estimated Aids deaths, which are now running at less than 50% of the level projected by UNAids.
The discovery of gremlins in UNAids’s estimation methodology has staggering implications for the rest of Africa, where accurate records are almost non-existent and UNAids estimates are almost all we have to go on.
I believe (and Geffen confirms) that South African insiders have been aware “for many years” that UNAids stimates are risible, but were unwilling to say so in public. Instead they watched in silence while Africa languished in ignorance and donors in Europe and the United States were misled by estimates they knew to be inaccurate.
Why would anyone do such a thing? At a guess — because they believe they were doing good; because high estimates are useful for fund-raising purposes, and because they were scared of committing the deadly, denialist sin of “sowing confusion”.
Turning to Geffen’s uncharitable remarks about my integrity, all I can say is that it is fiendishly difficult to write about Aids statistics in a way that laymen find comprehensible. At every turn you find yourself resorting to brute-force simplifications that antagonists might find misleading. My noseweek article contained several such, and Geffen has seized upon them with glee, loudly shouting that he has found errors. He’s welcome, but I would like to point out that his tirade contains several similar simplifications, one of which is infinitely more misleading than anything I have committed to print.
At issue here is a recent survey conducted in Kenya under the auspices of the US Centres for Disease Control and Prevention (CDC).
As Geffen notes en passant, the CDC found that 6,7% of adult Kenyans were HIV-infected, as opposed to the 9,7% established by earlier studies in pregnancy clinics.
Technically, Geffen’s facts are correct, but there is an elephant in the room here, a monstrous tusker unnamed and unmentioned.
In addition to finding a lower than expected HIV prevalence, the CDC study unmasked deep flaws in the methods UNAids uses to project pregnancy-clinic results into the general population.
As a result, estimated HIV prevalence in Kenya fell overnight from “as many as 4,8-million” (according to the BBC) to as few as one million (according to a personal communication from one of the CDC researchers).
Coming atop similar reversals in six or seven other African countries, the Kenya study is the death knell for the methods traditionally used to produce Africa’s HIV estimates. Indeed, a rethink was under way at World Health Organisation and UNAids even before the Kenya debacle, and all estimates are now up for revision.
Geffen writes that it barely matters if it turns out that only 15-million Africans are HIV-infected, as opposed to the 30-million postulated a year ago. I cannot accept this. Saying truth is irrelevant is an argument in favour of ignorance.
Rian Malan’s original article can be found at www.noseweek.co.za