HIV/AIDS Statistics: Misleading?

Published on 11th July 2006

Gross anomalies in HIV figures see a drop from 14 to seven per cent as other African countries revise their prevalence rates to secure donor funding.

You wake up one morning to news that 50 people have died in a road accident. Later in the day, it is clarified that only 10 perished, while the others were treated in hospital and discharged. Should the news source take credit for the 80 per cent reduction in fatality rate? Put another way, if the initial figures released on Aids last decade were erroneous, should they still be used to estimate the current prevalence and mortality rates?

Kenya and 15 other African countries have lately recorded a dramatic decline in Aids and HIV cases — with Rwanda, for instance, reporting a paltry prevalence rate of one per cent from a 30 per cent high 10 years ago. The rate in Kenya halved dramatically from 14 per cent in the Nineties to 6.7 per cent in 2003. Various Government departments and officials have publicly focused on the “success”. During the World Aids Day on December 1, 2003, President Mwai Kibaki said: “The prevalence rate has gone down from 13.1 per cent in 2001 to 10.2 per cent. But we should not celebrate these modest gains.” 

The National Aids Control Council (NACC), in a paid-up advertorial in the Press last month, quoted as evidence of its success the “declining HIV prevalence rate” from a peak of 10 per cent in the Nineties to seven per cent in 2003. The same statistics have been presented to UNAids, which has termed Kenya a success story in the fight against the pandemic.

In the Nineties, public officials warned on Aids, which was said to be claiming 700 lives daily. For a while, the “700 deaths daily” were synonymous with the disease. In his 2003 speech, the President said deaths had declined from 700 to 300 a day. Yet even a casual look at the figures show that at no time did the syndrome kill 700 people every day. Neither did the prevalence rate reach 14 per cent.

The discrepancies came to the fore in 2003 when the Kenya Demographic and Health Survey conducted a population wide random testing — a much more reliable measure than the previous ones. This resulted in a revision of the national HIV prevalence rate to 6.7 per cent. Gradually, the terms “revision” and “correction” were swapped for “decline” and “reduction”, with the NACC claiming credit for the feat.

Just before NACC officials flew to New York for a United Nations forum on Aids last month, the acting Director, Prof Aloyse Orago, told of an indicated decline in new infections ... even as the deaths per year rose.

The mortality rate is at its highest today (300 people every day), owing to maturation of the disease among people infected as far back as 10 to 15 years ago. Questions have also been raised over the modest decline from 6.7 prevalence rate to 6.1 in 2004. Was it a result of natural progression or deliberate control efforts? 

“As the pandemic matures to claim a greater number of infected people, the rates can be higher than those of new infections. Thus the number of people living with HIV would decrease (as would the prevalence ratio) but not, necessarily, that of new infections,” says NACC in its 2006 country report.

Use of inflated Aids figures is not unique to Kenya. It is an African phenomena, with many countries adjusting their initial estimates. Some seem to be saying, “sorry, we got it wrong,” while a few like Kenya are stating, “we have worked very hard on the matter; give us more money to do even more.”

Debate on the new statistics was spurred in April by the Washington Post following release of this year’s UNAids annual statistics in a story titled: How Aids in Africa was overstated. An attendant editorial had the headline: The Government’s wildly exaggerated African Aids pandemic.

New studies like the Kenya Demographic and Health Survey rely on random testing across entire countries, rather than among selected groups like pregnant women. They require two forms of blood testing to guard against the false positive results that inflated early estimates of the disease. These studies are far more effective at measuring variations in infection rates between rural and urban people and between men and women.

The new data suggests that the rate in Rwanda neither reached the 30 per cent estimated by early researchers, nor the 13 per cent given by the UN in 1998. The study and similar ones in 15 countries shed light on the disease across Africa, portraying lower rates in East Africa, and an epidemic that never was in most of West Africa — but which was a big problem in southern Africa.

In Ghana for example, says the Washington Post, the overall infection rate is 2.2 per cent. But in Botswana, the national infection rate is 34.9 per cent.  Most of the studies were by ORC Macro, a research corporation based in Calverton, Md. with funding from USAID and other international donors and concerned national governments.

Taken together, they raise questions on monitoring by UNAids, which for years overestimated the extent of infections in East and West Africa and by a smaller margin, in southern Africa. The new studies show that earlier estimates were skewed in favour of young, sexually active women in urban areas that had prenatal clinics. It is a fact that HIV rate among these women is higher than that among the general population.

UNAids produced its first global snapshot of the disease in 1998. Every year since, the United Nations has issued increasingly dire assessments: UNAids estimated that 36 million people worldwide were infected in 2000, including 25 million in Africa. In 2002, the numbers were 42 million globally, with 29 million in Africa.

But by 2002, disparities were already emerging. A national study in Zambia found a rate of 15.6 per cent — significantly lower than the UN rate of 21.5 per cent. In Burundi, a national study established a rate of 5.4 per cent, and not the 8.3 per cent estimated by UNAids. In West Africa, Sierra Leone had a 1 per cent rate — compared with an estimated UN rate of 7 per cent. In an interview with Craig Timberg of the Washington Post, Paul Bennell, a British economist whose studies on the impact of Aids on African school systems have shown mortality far below UNAids predictions, said: “From a research point of view, they’ve done a pathetic job. They were not predisposed, let’s put it that way, to weigh the counter-evidence. They wanted to generate big bucks.”

The United Nations started to revise its estimates in light of the new studies in its 2004 report, reducing the number of infections in Africa by 4.4 million. It also gradually decreased the overall infection rate for working-age adults in sub-Saharan Africa from nine per cent in a 2002 report to 7.2 per cent in its latest report.

Peter Ghys, an epidemiologist who has worked for UNAids since 1999, acknowledged in an interview with Timberg from his office in Geneva that HIV projections several years ago were too high since they relied on data from prenatal clinics. The warning on reliance on ambiguous statistics was raised much earlier by the Kenya Aids Watch Institute. In its website, it argues: “Listen carefully to the statisticians, who always insist that it is impossible to know the exact number of people living with HIV and Aids, and that the best use for surveillance statistics is identification of trends over time, rather than ‘correction’ of prevalence levels.”  

“The war on Aids must move from rhetoric over erroneous prevalence figures. A real solution will be built on the solid foundation of truth,” says KAWI director Francis Kajumo.

Some experts say that tallying HIV cases is not as important as finding the resources to fight the disease. But to researchers who drive Aids policies, differences in infection rates are not merely academic. Programmes deemed successful are urged on and funded lavishly by international donors, often to the exclusion of other projects. Hence the need to be seen to be succeeding.

This article appeared in the Daily Nation on July 6 2006

This article has been read 2,639 times