Taking Exception

AIDS Prevention: What Works?

By Daniel Halperin
Monday, October 22, 2007

Richard Holbrooke should be commended for urging a renewed focus on global HIV prevention [" Still Losing the AIDS Fight," op-ed, Oct. 9]. Although HIV treatment and care programs must be expanded, only by preventing new infections can we ultimately hope to turn back this devastating disease.

Holbrooke is also correct to emphasize the importance of HIV testing, especially for providing the main gateway into treatment. There is, however, little evidence that knowing one's HIV status fundamentally alters behavior. A few studies have found some modest changes in behavior among those who test positive, but most trials unfortunately show that people who discover that they are not infected with the virus continue acting as they did before being tested -- despite the obvious danger to themselves and to others.

The most rigorous study yet conducted, a randomized trial from Zimbabwe published last month in the journal AIDS, found an increased rate of HIV after people underwent testing and counseling compared with those who did not, though the increase was not quite statistically significant. The London-based researchers noted that some other studies similarly have found "disinhibition," or a worsening of behavior, among people who learned they were not infected. While it might seem intuitive that knowing one's HIV status and, ideally, receiving good counseling would lead to behavior change and reduced risk, the real-world evidence for this conventional wisdom is still unclear, especially for the large majority who test negative.

Worse, those who have been recently infected are by far the most infectious, even though they generally do not test positive (as they have yet to develop HIV antibodies). A recent study from Uganda estimated that about half of all HIV transmission there was due to such "window period" cases. And in a 2003 Tanzanian study, more than a quarter of women who tested positive and disclosed this to their male partners suffered negative consequences, including violence and abandonment.

HIV testing, in fact, does not appear to have played a crucial role in the declines in HIV rates observed in a number of countries and settings, including the U.S. gay community in the 1980s; in Thailand and Uganda in the early 1990s; and more recently in Kenya, Zimbabwe, southern India, urban Malawi and Ethiopia. In most of those places, HIV testing facilities were scarce or even nonexistent during the period in which HIV rates fell fastest.

Holbrooke listed a number of other key elements of a viable prevention strategy -- "education and counseling, free condoms, female empowerment, more male circumcision, and abstinence" -- but he neglected to mention the central role that multiple sexual partnerships play in infection rates. Broad interventions are needed to address this factor. While approaches such as condom promotion (particularly in epidemics spread mainly by prostitution, as in much of Asia) are important, various studies have shown that virtually everywhere in Africa where HIV rates have fallen, these declines were preceded by steep declines in multiple partnerships. As Helen Epstein explores in her recent book on AIDS in Africa, such profound shifts in behavior have usually been accompanied by broad transformations in sexual and societal norms, such as the homegrown "Zero Grazing" campaign that took root in Uganda in the late 1980s.

One of the biggest challenges for behavior change is the practice common throughout much the continent of multiple "concurrent" partnerships. These relationships, typically involving longer-term, overlapping liaisons, result in closely linked sexual networks whereby HIV can spread rapidly (especially during the dangerous window-period phase) throughout the population -- even though most people do not have many sexual partners.

As Holbrooke noted, circumcision has indisputably been proven to prevent HIV. It reduces the risk of male infection during intercourse by at least 60 percent and, unlike a condom, cannot be forgotten during a moment of passion. Nearly all of 15 studies conducted throughout Africa found that most uncircumcised men would want the service if it were affordable and safe, and even more women prefer it for their partners and children.

Another often neglected aspect of HIV prevention -- one prohibited from funding by the Bush administration's international AIDS program -- involves expanding family planning services, including for HIV-positive women who do not want to conceive. Reducing unintended pregnancies could greatly decrease the number of infected infants as well as the number of children who eventually become orphans.

While approaches such as testing, condom use and abstinence are important, no magic bullet exists for preventing AIDS. The most rigorous evidence suggests that there needs to be a vigorous expansion in Africa of behavior-change programs, for promoting partner reduction in particular, and greatly increased access to safe male circumcision. Of course it would also be useful to expand HIV testing programs, if not primarily for prevention purposes, then at least to help facilitate the care and treatment programs that are also vital to mitigating this pandemic.

The writer is a senior research scientist at the Center for Population and Development Studies at Harvard University's School of Public Health.

© 2007 The Washington Post Company