Through a glass, darkly: data and uncertainty in the AIDS debate.

The HIV/AIDS epidemic is the greatest threat to development in much of Africa. It is already the main cause of death in many countries, especially those in Southern Africa. However there is an absence of solid data on the scale and scope of the disease and how it is evolving. In this article we discuss the data on the epidemic--where it comes from and how it is presented. We note the limitations of the use of antenatal clinic surveys--which provide the bulk of our information. We then turn to the evidence of impact. The paper shows that the long incubation period between infection and illness means that it takes time for HIV infections to turn into AIDS cases, and AIDS cases to translate into deaths with all the consequences of orphaning, poverty and changing population structures. Furthermore it means that once the HIV prevalence has peaked, AIDS impact will take years to work through--this epidemic is a 'long-wave' event. The paper is premised on the view that HIV causes AIDS and AIDS causes death. It notes that insufficient and/or unreliable data have allowed leaderships to deny the scope and scale of the problem and that this is unacceptable. However it is incumbent on all to accept the moral responsibility for and the moral consequences of their work, and this includes those who gather, interpret and use the data.