Generalised epidemics – condoms are not enough
There has been controversy in some African countries about the decisions by the US government to downgrade spending on generalised condom promotion, but evidence from Uganda shows that a much more targeted approach may be required.
David Apuuli of Uganda’s AIDS Commission said in an accompanying presentation that while the majority of new infections were occurring within stable partnerships in Uganda, only 14% of new infections could be attributed to casual sexual partnerships. Both mother to child transmission and sex with commercial sex workers contributed higher proportions of new infections, clearly making the case for a more targeted approach in Uganda.
And as for targeting young people, “Whereas four years ago, the main new infections used to occur between 20 and 24, we now found the biggest new infections were occurring between 30 and 40,” said Apuuli.
“We found that discordance had gone up to about 48% among couples from just over 22% in four years.”
HIV prevention campaigns in Africa and Asia are often tilting at the wrong target and wasting money because of a basic lack of information about who is becoming infected in a country, David Wilson of the World Bank told the 2007 HIV Implementers’ meeting in Kigali, Rwanda, last month.
“The ABC wars have largely grown from our failure to know our epidemics,” he told the conference, which was attended by around 2000 people responsible for implementing HIV treatment, care and prevention programmes in developing countries. The ABC wars refer to controversies over prevention approaches that emphasise abstinence and being faithful to partners in preference to promotion of condom use.
Prevention programmes needed to recognise the difference between a concentrated epidemic and a generalised epidemic, he said.
“The conventional 1-percent [of the population] definition of generalised epidemics has hindered us knowing our epidemics”. The more relevant question he argued, is: “Where did our last 1,000 HIV infections come from?”
He highlighted the contrast between Uganda, where 65% of HIV infections are now estimated to take place within marriages, and Ghana, where 76% of adult male infections are linked to commercial sex workers, yet only 1% of World Bank prevention funding in Ghana goes towards work with sex workers.
He also highlighted the case of Zambia, where the epidemic has moved away from the traditional vulnerable groups of sex workers, their clients, truckers and uniformed services.
“A careful study by Mark Shields suggests that only seven percent of Zambia’s infections today were occurring among traditional vulnerable groups,” he pointed out.
“We need improved surveillance and analysis. We’re not using the data we have well enough to understand our epidemics, and we have to respond more rapidly to new evidence.”
More than behaviour change needed
Why is HIV incidence not falling in southern Africa despite high death rates, and why is high-risk behaviour resurging in countries like Uganda?
The reasons, David Wilson argued, are more to do with social environment than the sexual behaviour of individuals.
In South Africa, for example, young women are at much greater risk of HIV infection than their male peers. One of the chief causes appears to be age-mixing: young women with a partner at least five years older are six times more likely to become infected with HIV, Wilson said. While the prevalence of HIV infection peaks in the 25-29 age group in women, the peak prevalence isn’t reached until the age of 35 in South African men.
In the Rakai district of Uganda, said Wilson, 15% of women have a coercive sexual debut and 35% have ever experienced sexual coercion, and HIV incidence is two-fold higher among women with a coercive sexual debut. It’s a pattern repeated across southern and eastern Africa.
Also in Rakai, alcohol abuse is associated with a 50-80% increased risk of HIV infection.
Finally, concurrent sexual partnerships provide opportunities for multiple transmissions of HIV during primary HIV infection, when viral load is high. In Botswana, the country with one of the highest HIV prevalences in southern Africa, said Wilson, 43% of young men aged 15-24 reported concurrent sexual partnerships.
“What are the major lessons for generalised epidemics? Well, we have to reduce concurrency, intergenerational sex, sexual coercion, cultures of alcohol abuse and the vulnerability of couples," Wilson said.
“Generalised epidemics need fundamental community change and safer sexual environments. But globally, they are the exception and will always remain the exception. And the good news is that we know what to do in concentrated epidemics.”
Concentrated epidemics need condoms and clean needles
“I hope 2007 is the year we finally recognise Asia’s epidemics are concentrated and focus appropriately,” Wilson told the conference.
“They’re epidemics that are driven by vulnerable groups and initiated by sex or drugs. And the mathematics of Asian epidemics are extraordinarily robust.”
“They are initiated by sex if we have uncircumcised men, many of whom - about 10% - routinely visit sex workers. And sex workers have large numbers of clients, typically, about 20 a week or more. And so the first wave of Asian epidemics in Thailand, Cambodia and India were sexually initiated.”
“But elsewhere in second-wave Asian epidemics, injecting drug use has been the spark that’s created sexual transmission, then sex work the motor that maintains it.“
He highlighted recent data from Haiphong in northern Vietnam, where HIV prevalence is 2% among sex workers who don’t inject drugs and 55% among sex workers who do inject drugs.
“If we dissect the figures [the wrong] way, we could easily mistake a primarily injecting epidemic for a primarily sexual epidemic,” he said.
What’s needed in this situation is clean needles for drug injectors and condom promotion to commercial sex workers and their clients.
“But as epidemics mature and behaviour change occurs, new infections occur increasingly between infected individuals and their stable partners. And so we move from primary to secondary prevention with a far great emphasis on testing and counselling, especially among couples.”
“It’s extraordinarily difficult to get high rates of condom use in stable unions,” he said.
Summing up the approach that is needed, Wilson said: "In generalised epidemics we lead with partner reduction supported by testing and counselling particularly in mature epidemics. In concentrated ones, we lead with condoms and clean needles. But many epidemics are mixed enough to need both, based on understanding of our last 1,000 infections."