GLOBAL DIALOGUE Volume 6 ● Number 3–4 ● Summer/Autumn 2004—Africa in Crisis
The HIV/AIDS Pandemic in Southern Africa: Implications for Development
ALAN WHITESIDE
Alan Whiteside is director of the Health Economics and HIV/AIDS Research Division at the University of KwaZulu-Natal. He is a visiting professor at the University of Liverpool and Leverhulme Professor at Southampton University.
The AIDS epidemic, currently the greatest threat to human development, emerged in the past century and will be an issue for many decades to come. Most seriously affected are sub-Saharan Africa and especially the southern half of the continent. This disease is desperately worrying as Africa’s economic, political and social development has, at best, been faltering.
Soon, African countries will begin the cycle of celebrating fifty years since independence from colonial rule (Sudan, in 1955, was the first to achieve independence, followed by Morocco and Tunisia in 1956 and Ghana in 1957). In the 1950s, practitioners of the fledgling new discipline of “development economics” argued that Africa’s potential was unlimited; by contrast, looking at South-East Asia, they shook their heads in despair. But by the end of the 1970s it was evident that in many African countries things had gone seriously wrong.
The closing of the twentieth and dawning of the twenty-first centuries saw a new optimism. The end of apartheid had brought peace to southern Africa by the mid-1990s. There was a growing trend towards democratisation, and a number of countries had had their second free and fair elections. The establishment of NEPAD—the New Partnership for Africa’s Development—and the African Union seemed set to give a new lease of life to co-operation and integration across the continent. In 2000, the World Bank went so far as to publish a report,Can Africa Claim the 21st Century?The answer was a qualified yes.
But in the wings HIV was spreading and AIDS was waiting to take its place as one of the central actors in the development (or underdevelopment) of Africa. There were voices warning of what could happen—but they were largely ignored. Today, there is evidence of an emerging appreciation among some of the African leadership of just what AIDS might mean. For example, President Festus Mogae of Botswana has warned of the threat that AIDS poses to his country’s very existence. In May 2004, shortly after South Africa’s third democratic election, Mangosuthu Gatsha Buthelezi, leader of the largely KwaZulu-Natal–based Inkhata Freedom Party, openly announced that his son had just died of AIDS. United Nations Secretary-General Kofi Annan has established a special Commission on HIV/AIDS and Governance in Africa. Thus, there are grounds for cautious optimism.
This article will assess the scale of the problem and the validity of the data concerning it. I go on to look at the current state of the epidemic and explore why it is so serious. The likely implications of the epidemic for Africa’s development will be examined. I will assess responses to date, touch on the role of leaders such as President Thabo Mbeki of South Africa, note where there have been successes and consider why these have occurred. Finally, I will point to some priority areas for action.
Assessing the Crisis
The figures for the global epidemic are readily available, and indeed there is international consensus on the scale of the problem. At the end of 2003 there were estimated to be between 34 million and 46 million people worldwide living with HIV/AIDS. Sub-Saharan Africa has between 25 million and 28 million people infected. There were about 3 million deaths from AIDS in 2003, of which between 2.2 million and 2.4 million were in Africa.1Already, says the World Health Organisation (WHO), AIDS has killed more than 20 million Africans. It is the leading cause of death and lost years of productive life for Africans aged fifteen to fifty-nine.
The global data on the epidemic are compiled by the Joint United Nations Programme on HIV/AIDS (UNAIDS), but are co-released by UNAIDS and the WHO. The 2003 report noted that the estimates published that year were lower than those in 2002 because better data and understanding had enabled the UNAIDS secretariat and the WHO to arrive at more accurate statistics. The apparent concessions of error from the traditional providers of such data unfortunately coincide with the re-emergence of “dissident” views on AIDS statistics, and are fuelling scepticism about the impact of HIV/AIDS in many quarters, including a number of African capitals.
Scaremongering?
This has been picked up in the popular press. The most florid expression of scepticism has been by South African journalist, Rian Malan, arguing that “Africa Isn’t Dying of AIDS”.2Others have made the same claim. These researchers, journalists and social commentators have identified data inconsistencies thatappearto suggest anything from flawed projections to willful misinterpretation. The fact that any rational analysis accommodates the variability of modelling seems to have been lost in the emerging debate. Instead, we are left with the uncomfortable implication that the “AIDS industry” has been caught out. How does this match the evidence? And does it matter?
There are two sorts of data that have been called into question: those produced by the international agencies giving global figures, and those for individual countries. We need to understand how these data are collected and constructed. Global data (saying, for example, that in 2003 some 40 million people were living with AIDS and that it had killed a total of 20 million people) are compiled by UNAIDS using whatever national data are available. No international agency goes out and conducts its own surveys. Furthermore, reports give the dates of the data and estimate the range of uncertainty.
A review of the UNAIDS/WHO December 2003AIDS Epidemic Updateshows that there are no estimates of HIV prevalence in Liberia as a whole, and that the last HIV prevalence survey in urban areas there was in 1993. By contrast, for Zimbabwe the survey used in the update was from 2000, carried out in both rural and urban areas. Here, it is estimated that between 1,800,000 and 2,700,000 adults are infected: the median figure is 2,300,000.
It is worth reiterating that international agencies are dependent on country-generated data, and these are very variable in quality. Until recently, the data collected and supplied by countries were mainly gathered by surveillance systems that focused on pregnant women attending antenatal clinics. There were, of course, problems with representativeness, coverage and the fact that men do not attend such clinics. These considerations were factored into the data calculation, and they constitute one reason why a range of estimates was produced.
Recently, national population-based surveys have been carried out in Mali, Zambia, Kenya and South Africa. In the first three countries, these were demographic and health surveys (DHS), which are nationally representative household surveys with large sample sizes of about five thousand households. In South Africa, the survey was conducted by the country’s Human Sciences Research Council (HSRC). Population-based surveys are those in which samples are taken randomly from the population. Provided this is done properly, it should give an idea of prevalence across the population. In Kenya and South Africa, the results were lower than those obtained from antenatal surveys. Consequently, they were seized on by the press to suggest that national HIV prevalence has been overestimated.
The UNAIDS figure for Kenya was that 9.4 per cent of all Kenyans were living with HIV/AIDS, whereas the DHS estimated that 6.7 per cent of Kenyans were infected. In South Africa, the HSRC’s estimated prevalence was 15.6 per cent for those aged between fifteen and forty-nine. The UNAIDS estimate is 20.1 per cent for the same age group.
Do we then, as some would suggest, dismiss antenatal surveys and the data produced by UNAIDS as being too pessimistic? A hard look at the population surveys shows that they, too, are not without problems. In both Kenya and South Africa there were high refusal rates—people who were not contactable, or who would not be interviewed or provide specimens. In South Africa, of the 13,518 individuals selected and contacted for the survey, 73.7 per cent agreed to be interviewed, and 65.4 per cent agreed to give a specimen for an HIV test. In Kenya, 70 per cent of those eligible agreed to give blood samples. Epidemiologists become concerned when participation rates fall below 80 per cent.
Statistically, the South African antenatal clinic and HSRC “confidence intervals” overlap, that is to say, the lowest levels recorded by the former survey overlap with the highest levels of the latter, which means that either result could be right. In Kenya, the full DHS results have not been released so detailed comment is impossible, but UNAIDS notes that when the results are broken down by gender, the HIV prevalence of 8.7 per cent among women is in the same range as the 9.4 per cent prevalence estimated by UNAIDS and the WHO. What this points to is that data need to be used cautiously and open-mindedly.
An additional problem is the lax way in which numbers are thrown around by both the press and many AIDS activists. For example, Swaziland currently has the world’s highest HIV prevalence rate among attenders of antenatal clinics. Swaziland’s official 2002 antenatal survey found a prevalence of 38.6 per cent among those tested. But this is then presented as though 38 per cent of adults are infected, or even as though 38 per cent of the whole population has HIV. The result is complete confusion or deep despair—or both.
A second, previously noted difficulty is that in order to arrive at the number of men and infants infected, we are generally obliged to use antenatal clinic data. For example, in South Africa, according to an antenatal survey conducted in October 2002 by the ministry of health, 26.5 per cent of pregnant women were HIV-positive. Based on these results, and using a model developed by the Department of Health, an estimated 5.3 million South Africans are infected (2.95 million women and 2.3 million men aged fifteen to forty-nine, and 91,271 babies infected through mother-to-child transmission). Are there four million or six million infections? The answer is that we cannot know with complete accuracy; we have to work within a band of estimates. The reality is that there are millions of South Africans infected, and we will have to deal with the consequences.
Facing Facts
Finally, we are measuring HIV infections not AIDS cases, and it is the full-blown illness and deaths that will elicit a response from planners and policymakers. Because HIV is hidden—we do not know who is infected—it can be denied. An extreme example of such denial is the September 2003 interview in which President Mbeki said that no one close to him had died of the disease, nor did he know anyone infected with HIV.
It must be admitted that the impact of AIDS has not been as rapid as anticipated ten years ago. This is no surprise to science, given the nature of the pandemic and of the research process, but it is grist to the mill of those who have found a media niche and an audience anxious to have its doubts and nervousness assuaged. Sadly, this includes many senior policymakers. But it ignores the fact that we arestillfaced with a full-blown development crisis.
We have to keep reminding ourselves that slightly reduced numbers do not equate to “good news”: whether you drown in six inches or six feet of water doesn’t change the fact that you are dead. The great danger of this data debate is that it provides the perfect excuse for inaction. By contrast, the effective prediction and warning of impact can mobilise a response which, theoretically at least, may avert the impact. It is rather like the difference between measuring the damageaftera flood or predicting the potential damagebeforeit: if the householders fill sandbags and move the furniture upstairs, and thus reduce the damage, was the prediction wrong?
John Maynard Keynes reportedly said: “When the facts change, I change my mind. What do you do, sir?” We need to be honest and recognise when the facts have changed, but equally that we are not in any way suggesting there is no crisis—just that we need better to understand its scale, scope and timing.
HIV/AIDS in Africa
So how bad is the epidemic in Africa? The first and perhaps most important point is to recognise that there is notanAfrican epidemic: there are many epidemics.
In north Africa and parts of the Sahel there is little evidence of an outbreak. Some countries do not report any cases and where there are data the prevalence rate is below 0.25 per cent among adults. There have been some signs of outbreaks among drug users, although sexual intercourse remains the dominant form of transmission. The intense stigma attached to HIV/AIDS, and the discrimination against those infected with it, mean the epidemic may remain hidden. Although HIV/AIDS is nascent, there is concern about levels of conflict, poverty and unemployment, which may speed its spread.
In eastern Africa, the epidemic appears to be stable at levels of between 5 and 15 per cent among adults. This translates into horrific numbers, though: over 2 million infected Ethiopians, 2.3 million Kenyans, and 1 million Mozambicans.
The 2002 UNAIDS update notes that in west and central Africa, the relatively low adult HIV prevalence rates in countries such as Senegal (under 1 per cent) and Mali (1.7 per cent) are shadowed by more ominous patterns of growth.
HIV prevalence is estimated to exceed 5% in eight other countries of west and central Africa, including Cameroon (11.8%), Central African Republic (12.9%), Côte d’Ivoire (9.7%) and Nigeria (5.8%)—sobering reminders that no country or region is shielded from the epidemic. The sharp rise in HIV prevalence among pregnant women in Cameroon (more than doubling to over 11% among those aged 20–24 between 1998 and 2000), shows how suddenly the epidemic can surge.3