Durban: January 31, 2005

Dr Ruth Rabinowitz IFP Health Spokesperson

The debate around AIDS statistics preoccupying the minds of journalists, academics and commentators like Riaan Malan, amounts to a lot of hot air. We would do better to open our eyes to the reality that surrounds us daily. HIV and AIDS are steadily on the increase and more energy should be spent changing government will to end denial and change behaviour. All our stats in South Africa will be imperfect in view of the paucity of tests that are performed, the format of the Death Certificate and the inadequacy of our statistical records.

The B1 -1663 Notification of Death form, in Part G asks doctors to fill in a) IMMEDIATE CAUSE (Final condition resulting in death) and under the line states (Due to or as a consequence of), with three more lines provided to give the sequential cause of the last cause of death. The instruction is to "Sequentially list conditions, if any, leading to immediate cause. Enter underlying cause last (disease or injury that initiated events resulting in death)".

In the case of AIDS a physician might write Pneumonia as the immediate cause, followed by "due to TB, due to HIV". But this would necessitate a positive HIV test being in that physician's possession or recorded on a data-base to which the doctor has access. Given the secrecy and stigma surrounding HIV/AIDS and the limited testing that is done routinely, few doctors would hazard a guess that the patient died of HIV/AIDS. This is particularly true of rural areas, where the greatest number of deaths occur and where the hospitals have insufficient resources to treat patients, let alone perform HIV tests on all of them.

A separate research project is necessary to establish what percentage of doctors fill in the underlying cause of death as HIV/AIDS. Each time another layer of research is added to the extrapolation, there is further room for error. Figures for 2001 vary as much as three times between Home Affairs statistics and MRC stats. A similar chasm exists between South African estimates and UNAIDS estimates.

South African studies estimate the prevalence of HIV as 11.4% of the population in 2003 while UNAIDS statistics estimate the figure to be double that at 21.5%. These large discrepancies suggest that the statistical debate is currently meaningless.

Another factor rendering the statistics virtually useless is the time it takes for them to be published. It is now 2005 and we are arguing about statistics from 2001. What relevance do these have to the current situation?

Furthermore both the government and the MRC stats are based on limited samples.

The only ways statistic will start to mean something is when the approach to HIV changes from one of academic interest to one befitting a national emergency. Then testing will be done at all clinics, hospitals and medical practices, tests will be provided free by government and a data base will be kept current, listing numbers of people who test positive and their demographic profile.

Confidentiality can still be honoured, even if AIDS is made notifiable, while knowledge by participants of their HIV status is of paramount importance. Until now the focus has been on secrecy, one on one pre test counselling and the right of participants NOT to be told their HIV status. This politically correct approach amounts to damaging populism that regards privacy as more important than death.

The antiretroviral roll out, although it has not reached desired levels, is increasing and will in itself encourage more people to be open. It would be informative to link statistics from testing and treatment sites of government and NGO's funded by President Bush's Emergency fund or others.

These facts (not projections) would be current and provide some insight into the progress of the pandemic and its management. The current debate is academic and superfluous.

Dr. Ruth Rabinowitz MBBCh (MP)
Mobile: +27 (0)82 579 3698
Home/Office: +27(0)11 802 1826
Fax: +27(0)11 804 4221