You are hereRian Malan still getting AIDS stats wrong

Rian Malan still getting AIDS stats wrong

01 December 2009

Printer-friendly version

by Nathan Geffen, 1 December 2009

Substantially updated by the author on 7 December 2009

In a piece published in Rapport newspaper and on politicsweb, Rian Malan claims:

[D]on't trust anything the Aids bwanas say - especially not Nathan Geffen of TAC. Earlier this week, he informed the world that Zuma's mistake "was of little consequence," because other data showed that SA's death rate has doubled since l997. Hmmm. It is true that annual death registrations rose from 316,000 in 1997 to around 600,000 in 2007, but it is absurd to claim, as Geffen did, that this was almost entirely the result of Aids.

Over the same period, completeness of registration rose from around 67 percent to 81 percent, according to Stats SA, while our population rose by close on seven million. If you adjust the raw numbers accordingly, Geffen's apparent doubling shrinks to an increase of around 15 to 20 percent.

He then published a correction:

Correction: I am informed that my mathematical skills leave much to be desired. If you run the numbers in the penultimate paragraph correctly, the real increase in SA death registrations since l997 comes out at around 30 percent - still a tragedy by any reckoning, but still way short of the doubling claimed by Geffen.

First let's deal with the numbers. Malan continues to get it wrong, even in his correction. AIDS deaths have conservatively increased 6-fold since 1997, from approximately 50,000 to well over 300,000 in 2006 as well as 2007 and maybe even over 350,000, meaning the real increase in deaths -which is way more than 30%- recorded or otherwise, was mainly due to AIDS. The ASSA2003 interventions model calculates under 150,000 AIDS deaths in 2000 and about 360,000 in 2007. Approximately 45% of deaths in 2006 and 2007 were due to AIDS. The model probably overestimates AIDS deaths but not substantially enough to give any material support to Malan's argument. But even a 30% increase in mortality, as Malan acknowledges, is a tragedy.

Far more knowledgeable people than either Malan or myself work on the ASSA model. They take into account all the available data: recorded deaths, improved registration, population growth and much else. Of course, as with any model, there is a good deal of uncertainty, but it is the best we have to go on - and far better than Malan, who admits his mathematical skills leave much to be desired.

In Malan's Rolling Stone article in 2001, his arguments were based on outright AIDS denialism, in which he confused the different types of testing algorithms needed for diagnosis of an individual patient versus epidemiological surveys. His articles in 2003 in Spectator and Noseweek continued in the same vein. But with each subsequent article he has come closer and closer to admitting the massive scale of the HIV epidemic. His latest acknowledgement that AIDS is responsible for a 30% increase in mortality is almost mainstream. Yet his writing style remains unrepentant and he brushes off as a triviality the realisation that he is not mathematically competent to do this work.

He also misrepresents me. Nowhere did I write that real deaths had doubled in my recent article that he appears to be referring to. I wrote, accurately, that recorded deaths increased over 90% in a decade. I also wrote, "Improved death registration and population growth can account for only a small portion of this increase. The vast majority of additional deaths are due to the HIV epidemic."

This was also accurate. Only a careless reading of my wording would imply that total deaths (i.e. recorded plus unrecorded) have doubled due to AIDS. This might seem a minor "He said, I said" spat, but it demonstrates a lack of integrity in public engagement. Despite our overall success, I am sure there is much fair criticism that can be directed at TAC about our actions over the last decade, but manipulating AIDS statistics is not one of them.

Without any sense of irony, Malan concludes, "we should just ignore those who try to manipulate us with numbers and support Zuma's common-sense plan to stamp out the disease." Actually, it is because of the people who Malan falsely accuses of manipulation that we finally have a common-sense AIDS plan. While we were fighting for it, Malan was supporting Mbeki on AIDS.

Some of my colleagues have been weary about me responding to Malan. They have warned me that he is being a contrarian so that he can promote his new book and that a response is exactly what he wants. But I think Malan is a talented wordsmith whose writing style convinces some people; it is them I aim my articles at. But being clever with words does not imply competence and Malan's articles on AIDS are littered with errors.

According to Wikipedia Malan stated, "I get a kick out of it when the Treatment Action Campaign attacks me; it's like sport." I do not know if he really said this, but it does appear to be sport for him.

For TAC and me, it is a waste of precious time and aggravating. Many people find AIDS statistics impersonal. But I do not; they remind me of Christopher Moraka, who testified before Parliament in 2000 that he could not get medicines to treat his systemic thrush. He died a couple of months later. Or Edward Mabunda, TAC's late firebrand poet. And Ronald Louw, a close friend and brilliant lawyer based at UKZN who died a few years ago of AIDS. No, this debate is not sport for TAC or me.