Scientific methodology of council's report is questioned

By Pat Sidley

Business Day (SA) 12 Oct. 2001

Johannesburg -- The simmering row between Statistics SA, government and the Medical Research Council (MRC) bubbled over yesterday with an hourlong attack by the state on the MRC's findings that 40% of deaths of people between the ages of 15 and 49 were related to AIDS.

Stats SA launched the attack on the MRC report yesterday, questioning the scientific methodology and demographic modelling of the report and criticising its alleged lack of empirical evidence.

In a presentation at the offices of the Government Communication and Information Service, devoted to questioning the MRC's as yet unreleased report, Statistician-General Pali Lehohla damned the report with faint praise, saying it was a valuable contribution to the debate.

MRC president Malegapuru Makgoba said yesterday he would respond formally next week, when the report was formally released.

But in a speech to the Black Management Forum in Sandton, he tore into Stats SA without mentioning the organisation, referring instead to experts who gained "authority overnight". He reminded the influential audience of would-be experts such as former political consultant Eugene Nyati, who proved to have no credentials at all.

"(The) controversy and innuendo around the report was a measure of the degree of the level of denial (of AIDS) in this society," he said.

The MRC report, to be formally released on Tuesday, says that 40% of deaths of people between 15 and 49 years of age last year were caused by AIDS and 25% of all deaths last year were caused by AIDS.

Yesterday's attack follows leaks of the MRC report and a cabinet decision to approve its release. The cabinet's announcement on Wednesday was accompanied by a statement from Stats SA outlining its problems with the report.

Ros Hirschowitz of Stats SA said the MRC findings did not flow from empirical evidence, that demographic modelling was problematic and that decisions on whether excess deaths were caused by AIDS were "difficult in the absence of accurate information" on the cause of death.

She also questioned the assumptions on which the MRC made its findings and raised the issue of the "probability of transmission" or how likely it is that HIV would be transmitted during sex.

Stats SA said empirical evidence was needed and proposed a survey using its household informationgathering skills and a sample of 60000 people.